NT will kill Unix in a year, OS war & Web revolution
Michael Hoffman
relationship between the OS war and the Web revolution. I should
probably differentiate these issues into 2 or 3 threads, but they are
interconnected in my mind. I hope this combination of concerns is
interesting to you. If the thread explodes or wastes bandwidth, maybe
someone out there can extinguish it.
I am becoming a professional Web developer. I have to decide what
tools to learn and what skills to develop. I'm pretty much at the
start of my career, and I want to learn systems that will be relevant
in the future.
Currently, most Web servers run Unix. But NT and the free MS server
software have been getting favorable results. If I just looked at a
static snapshot of today's dominant platforms, I would say that
learning the Unix environment for Web development is a much better
idea than committing to an NT environment. However, looking at the
*direction* that things are going, looking at the latest *trends*, I
think NT has a much better future than Unix.
I think that technical arguments can miss the point. Superiority of a
technology does not correspond with dominance of that technology.
I am confident that NT will quickly become the dominant operating
system for the next few years, whether or not it is technologically
superior. I think it's a better career move to align myself with
NT-based Web development, than Unix-based Web development.
I am facing some hard decisions now. Currently, the recommended way
to write CGI scripts is to code in Perl on Unix. But I think there
will be greater demand for NT-based CGI programming, over the next few
years.
I'm facing a similar hard choice between Java and Visual Basic Script.
I am sure that I want to become a specialist in Microsoft solutions,
because I like Microsoft products and they are going to become the
most popular. Microsoft skills will have the greatest demand.
I want to learn the most appropriate Microsoft-oriented tools and
environments for Web development. I don't want to be ignorant of
Unix, Perl, and Java, but I find I have to resist jumping into the
Unix/Perl camp, which is currently dominant and tries to pull me in.
Unix is yesterday's revolution and I'd rather minimize my involvement
with it. NT is far from perfect, but the NT environment and mentality
is far more up-to-date than Unix, which has its roots in the early
70s. The Unix mentality is obsolete, never mind the technical
details.
While Unix will live on, it will be killed in the sense of being
overwhelmed by NT, as everyone in the world becomes involved in
computers. The 90% of people who have not yet touched a computer will
not buy into Unix; they will buy into NT.
Internet terminals will shield the masses from dirtying their hands
with either NT or Unix. These terminals will be served by Unix
servers at first, but soon, the entire business world will switch to
NT. Technical issues aside, Unix has no future; Microsoft is rapidly
becoming the OS monopoly. OS/2 is completely out of the picture.
Technical people might hate this trend, but based on reading the
technology and business magazines and reports, NT and Microsoft are
unstoppable. Whether or not this is desirable, it's happening, and
I'm going along with it.
_________________
Microsoft is the right balance of price, consistency, and popularity.
I've always hated Apple because they were overpriced when I bought my
first computer in 1988.
I've always hated Unix because it's essentially an antiquated, chaotic
character-mode environment with no serious GUI and poor compatibility
among the zillion flavors.
I've always hated OS/2 because it's made by IBM.
I've always liked Windows because it's affordable and dedicated to
consistency of user interface.
________________
Windows may suck in some ways technically. But technological
superiority is not the driving factor for popularity. The important
thing is the balance of price, popularity, and ease-of-use, as well as
technically being good enough.
Windows is technically good *enough*.
In choosing and advocating an operating environment, find the best
*balance* of:
o ease-of-use
o price
o popularity/ubiquity
o technical performance
Linux or OS/2 or NextStep may be in some ways technically better than
NT. Macintosh may in some ways have better ease-of-use than NT.
Amiga may in some ways have better price/performance than NT (or 95).
But the real deciding factor is *balance*.
When you consider the *balance* of factors, Microsoft Windows and NT
has trounced the others and deserves to become predominant.
Technological superiority of Linux or OS/2 is merely one factor to
consider. I respect technological considerations, but they are just
one consideration among several.
I've never had to think deeply about justifying my commitment to
Windows until the opportunity to become a Web developer came along.
Now I must decide whether to learn Java or Visual Basic Script, Unix
or NT, and coding CGI in Perl or Basic. I concretely faced these
decisions when standing in the largest bookstore in the region -- the
Stanford campus bookstore, which currently has many more books about
the Unix operating system than on the Windows operating system. Each
substantial book costs $50. Should I buy a $50 Unix book, or a $50 NT
book? A $50 Visual Basic book, or a $50 Perl&CGI book? I still am
unclear about which tools and skills to master to specialize in Web
development for Windows, and the problem is exploding as the entire
world realigns itself around the Web.
I would appreciate any insight you can give me, especially regarding
Unix vs. NT trends relating to a Web development career. I hope the
responses, if any, contain *substantial* insights or criticism, not
just empty flames that waste everyone's time.
The OS wars are being affected in several ways by the Web revolution,
including the Internet terminal concept and Java, which supposedly
render the OS war irrelevant.
-- Michael Hoffman, EE, technical writer, Web developer,
techno-trendmonger
Erik Blomquist
news1.radix.net!news.cais.net!primus.ac.net!imci4!newsfeed.internetmci.com!news.exodus.net!news1.best.com!shellx.best.com!usenet
From: mic...@cybtrans.com (Michael Hoffman)
Newsgroups:
comp.os.ms-windows.nt.advocacy,comp.os.ms-windows.advocacy,comp.infosystems.www.advocacy,comp.os.linux.advocacy,comp.os.os2.advocacy,comp.os.amiga.advocacy,comp.sys.mac.advocacy,comp.unix.advocacy,comp.sys.next.advocacy
Subject: NT will kill Unix in a year, OS war & Web revolution
Date: Sun, 25 Feb 1996 00:21:57 GMT
Organization: cybtrans.com
Lines: 137
Message-ID: <4goa2q$9...@shellx.best.com>
Reply-To: mic...@cybtrans.com
NNTP-Posting-Host: pdr.vip.best.com
X-Newsreader: Forte Free Agent 1.0.82
Xref: news1.radix.net comp.os.ms-windows.nt.advocacy:15945
comp.os.ms-windows.advocacy:134080 comp.infosystems.www.advocacy:2655
comp.os.linux.advocacy:47085 comp.os.os2.advocacy:208937
comp.sys.mac.advocacy:104067 comp.unix.advocacy:17154
comp.sys.next.advocacy:38821
Now this is just sad. It's obvious that you put a lot of time and effort
into composing this post (137 lines!), but you'll never be able to
successfully troll a newsgroup until you learn how to properly forge a
newsgroup header.
-Erik
Michael Hoffman
>into composing this post (137 lines!), but you'll never be able to
>successfully troll a newsgroup until you learn how to properly forge a
>newsgroup header.
>-Erik
So what do you think about the impact of the cross-platform Web, Java,
and Internet terminals on the OS war?
Do you think it is likely that NT will very rapidly gain acceptance
over Unix for corporate intranets? Will the Unix-to-NT Web server
ratio suddenly flip-flop?
Michael R. Delahoz
Hoffman) writes:
:>
:>Sorry if I'm excessively cross-posting here. I'm exploring the
:>relationship between the OS war and the Web revolution. I should
:>with either NT or Unix. These terminals will be served by Unix
:>servers at first, but soon, the entire business world will switch to
:>NT. Technical issues aside, Unix has no future; Microsoft is rapidly
:>becoming the OS monopoly. OS/2 is completely out of the picture.
So of course I snipped this asshole's bandwith hogging dribble.
:>I've always hated OS/2 because it's made by IBM.
:>
Then get the fuck out of Comp OS OS/2 Advocacy
:>Linux or OS/2 or NextStep may be in some ways technically better than
:>NT. Macintosh may in some ways have better ease-of-use than NT.
:>Amiga may in some ways have better price/performance than NT (or 95).
That's true, but like the moron that you are you come into an OS/2 group and
spam us about an inferior system that is a lie as far as a web soulution goes.
Netscape has already in great detail elaborated on the cost differential
between Netscape's server solution and M$' own bullshit "solution."
To completely implement the Netscape version an IS shop has to invest about
$1,700.00 more or less, to do with M$' "free" solution which BTW only runs on
NT Server, the projected costs are in the vicinity of $6,000.00
Get a clue asshole, but until you do, do c.o.o.a a favor and spam elsewhere.
BB
Michael Hoffman
>In message <4goa2q$9...@shellx.best.com> - mic...@cybtrans.com (Michael
>Hoffman) writes:
>:>Linux or OS/2 or NextStep may be in some ways technically better than
>:>NT. Macintosh may in some ways have better ease-of-use than NT.
>:>Amiga may in some ways have better price/performance than NT (or 95).
>Netscape has already in great detail elaborated on the cost differential
>between Netscape's server solution and M$' own bullshit "solution."
Yes, I read that. Very interesting. But somehow unconvincing, or
beside the real point.
>To completely implement the Netscape version an IS shop has to invest about
>$1,700.00 more or less, to do with M$' "free" solution which BTW only runs on
>NT Server, the projected costs are in the vicinity of $6,000.00
>BB
That's simple logic. I get the impression that large corporations are
yearning to move to NT, and MS's giveaway of the server software is
the excuse they will use to choose NT servers rather than Unix servers
running Netscape. It's not logical, it's twisted motives. Many
corporations *want* a monolithic Microsoft solution. This is the buzz
I'm picking up from reading various magazines and online news.
I'm arguing from market perception, rather than from plain economics
or technical points. I've seen several references to NTs
"familiarity" and "ease of use" over Unix. This translates to lower
maintenance costs.
There is a very large chance that Microsoft is about to become a
software monopoly far more than they are now, very soon. They've
passed a critical threshold of mindshare.
I propose that all future software sales will come from Microsoft, as
the nontechnical 90% of the world suddenly starts using computers this
year because of the Web. It's going to happen instantaneously. This
is a prediction about the most major discontinuous change.
Microsoft will become a software monopoly for all the people who are
about to explode onto the Web. The Web will make the OS war
unimportant at the same time as Microsoft's OS captures the billions
of people, the vast majority of people who have never given a thought
about NeXT vs Amiga vs OS/2 vs Windows vs Mac vs Linux. The vast
majority of people are completely unaware of the OS war, and they are
about to start surfing the Web. If ever they have to pick an OS, they
will choose Microsoft, because it has the marketing mindshare and the
desire to capture these people. Windows is positioned as everyman's
OS; it is in no way elite. It is humble and "good enough". It is
very Bob.
If most people end up using cheap Internet terminals, they might find
themselves using Sun hardware (Java chips) or accessing Web pages that
happen to reside on Unix servers. Or maybe mainframes or NT servers.
I sure haven't read much about OS/2 Web servers yet.
This year, there will be a huge explosion as the majority of people
all over the world, rather than a very small minority, rush onto the
Web. The market for computing power -- whether desktop computers,
servers, or terminals -- is going to grow more this year than it has
in all past years, combined.
What hardware and OSs will predominate a year from now, after this
remarkable event that changes the whole scope of computing? Watch out
-- we're all about to be trampled by discontinuous change.
John Cathie
:>system for the next few years, whether or not it is technologically
:>superior. I think it's a better career move to align myself with
I am confident you are very wrong.
:>Microsoft is the right balance of price, consistency, and popularity.
What consistency? APIs that change more frequently than my underwear?
What popularity? About a million copies sold in 3 years? I think not.
Price? Pay per server, pay per seat. They get you coming both ways.
:>Windows may suck in some ways technically. But technological
:>superiority is not the driving factor for popularity. The important
:>thing is the balance of price, popularity, and ease-of-use, as well as
:>technically being good enough.
:>
:>Windows is technically good *enough*.
Not for me it ain't. It failed to deliver the capabilities I required with
version 3.0. It still fails to meet my standards. And all I want is a stable,
reliable multitasking operating system.
:>When you consider the *balance* of factors, Microsoft Windows and NT
:>has trounced the others and deserves to become predominant.
ROTFLMAO
:>I would appreciate any insight you can give me, especially regarding
:>Unix vs. NT trends relating to a Web development career. I hope the
:>responses, if any, contain *substantial* insights or criticism, not
:>just empty flames that waste everyone's time.
Re-evaluate the parameters and the information you have assimilated to this
point. Some of the data is plain *wrong*.
UNIX is going to be around for a long, long time to come.
And as for hating OS/2 because IBM sells it; buy OS/2 1.3 from M$ if you feel
that strongly. It's rock solid. But I have to empathize with you, It's likely
that I despise M$ as much as you hate IBM (for reasons I'll pass on in email
if you really need to know). But if M$ ever came up with something superior
to OS/2 (right, and pigs will fly), I'd try it, and maybe even use it. But it
would take a lot to get me away from OS/2 to *any* other OS.
I believe you are shortchanging yourself by dismissing OS/2 (and IBM in
general).
Good luck whatever your decision.
/////////////////////////////////////////////////////////////
//
// John Cathie jca...@mindlink.bc.ca
// Team OS/2
//
// Certified IBM OS/2 LAN Server Software Specialist
// IBM BESTeam member
//
William Unruh
>I am becoming a professional Web developer. I have to decide what
>tools to learn and what skills to develop. I'm pretty much at the
>start of my career, and I want to learn systems that will be relevant
>in the future.
>Currently, most Web servers run Unix. But NT and the free MS server
>software have been getting favorable results. If I just looked at a
>static snapshot of today's dominant platforms, I would say that
>learning the Unix environment for Web development is a much better
>idea than committing to an NT environment. However, looking at the
>*direction* that things are going, looking at the latest *trends*, I
>think NT has a much better future than Unix.
Lets see. Unix has had free Server software since day 1. Microsoft
releases a free server and they will win. NT after 3 years has less than
1M purchasers and it is the wave of the future. Unix has more free and
powerfull tools than any other system, but NT is the latest trend.
You know, I think you should lie down a while, switch off your TV, stop
your subscriptions to advertising magazines and then think a while.
a) WinNT is not moving anywhere right now.
b) Win95 aint a Web Server platform.
(Note despite that common first name the above two are not identical)
c) Unix is here, is established and is running most servers.
d) OS2 is growing and with connect ans server will have a Web prsence.
However IBM's server costs are a bit steep. (What in the world makes
them think server software is worth >1000 )
So, go out, buy that Perl book. If your fears of NT are correct, someone
might well have ported Perl to NT anyway.(They already have to OS2)
And the stuff you learn will be pretty portable even if you have to
learn a new computer language.
>I'm facing a similar hard choice between Java and Visual Basic Script.
Uh, you might not have heard, but MS is licensing JAVA. They have
themselves thrown in the towel.
>70s. The Unix mentality is obsolete, never mind the technical
>details.
What mentality was that? Getting work done? People developing masses of
software for free? Oh, you mean command line! Funny, I thought most
programs were still "line oriented".
>While Unix will live on, it will be killed in the sense of being
>overwhelmed by NT, as everyone in the world becomes involved in
>computers. The 90% of people who have not yet touched a computer will
>not buy into Unix; they will buy into NT.
Uh, but most of that 90% will not be running servers. They will be using
ISP's to provide them with the path onto the internet. I have never been
able to tell what machine it was that served me my web pages. Have you?
So what if 90% use Win95 or even Dos? (and by the way that 90% sure
won't be buying NT either).
>Internet terminals will shield the masses from dirtying their hands
>with either NT or Unix. These terminals will be served by Unix
>servers at first, but soon, the entire business world will switch to
>NT. Technical issues aside, Unix has no future; Microsoft is rapidly
>becoming the OS monopoly. OS/2 is completely out of the picture.
Dear me. You really do seem to have been brainwashed haven;t you. I
think you should rethink your srategy. There is this great system called
Bob. I think you should abandon NT and go for BOB. It will take over all
computers I thought I had heard.
...........
--
Bill Unruh
un...@physics.ubc.ca
Joseph Perkowski
Joe
Ian S. Nelson
choice.
good luck,
ian nelson
Yusseri Yusoff
>
<snip>
<extremely long diatribe about NT, Unix, OS/2, etc..>
>
> -- Michael Hoffman, EE, technical writer, Web developer,
> techno-trendmonger
Wow! What a great bait!
--
|Yusseri Md. Yusoff |
|Dept. of Electronic Systems Engineering | email:ymy...@essex.ac.uk
|University of Essex |
BOWEN JASON MICHAEL
and Comp. Sci. friends that I have, all laugh when the idea of NT
replacing UNIX comes up. The only people who would buy the
self-prophesising dribble in most industry rags are those who can't
think and evaluate for themselves.
Jason
Michael Hoffman (mic...@cybtrans.com)
wrote: : Sorry if I'm excessively cross-posting here. I'm exploring the
Bob Madden
>bad...@seanet.com (Michael R. Delahoz) wrote:
>>In message <4goa2q$9...@shellx.best.com> - mic...@cybtrans.com (Michael
>>Hoffman) writes:
>>:>Linux or OS/2 or NextStep may be in some ways technically better than
>>:>NT. Macintosh may in some ways have better ease-of-use than NT.
>>:>Amiga may in some ways have better price/performance than NT (or 95).
>>Netscape has already in great detail elaborated on the cost differential
>>between Netscape's server solution and M$' own bullshit "solution."
>Yes, I read that. Very interesting. But somehow unconvincing, or
>beside the real point.
You need to provide some justification for this conclusion. Just your
saying it, does not necessarily make it so.
>>To completely implement the Netscape version an IS shop has to invest about
>>$1,700.00 more or less, to do with M$' "free" solution which BTW only runs on
>>NT Server, the projected costs are in the vicinity of $6,000.00
>That's simple logic. I get the impression that large corporations are
>yearning to move to NT, and MS's giveaway of the server software is
>the excuse they will use to choose NT servers rather than Unix servers
>running Netscape.
Maybe Microsoft finally woke-up, and started charging what the
software was actually worth.
> It's not logical, it's twisted motives. Many
>corporations *want* a monolithic Microsoft solution. This is the buzz
>I'm picking up from reading various magazines and online news.
These same "various magazines" have been declaring OS/2's demise for
over 3 years, that I know of, and it still hasn't happened. Could it
be that they are wrong about a few other things as well? Don't just
take everything you read as "Gospel".
>I'm arguing from market perception, rather than from plain economics
>or technical points. I've seen several references to NTs
>"familiarity" and "ease of use" over Unix. This translates to lower
>maintenance costs.
That is generally what Microsoft does quite well, since it is so
difficult to argue in favor of their products on technical grounds
alone.
>There is a very large chance that Microsoft is about to become a
>software monopoly far more than they are now, very soon. They've
>passed a critical threshold of mindshare.
I must not have been paying attention again... Just when did this
happen? And WHY didn't somebody TELL ME?
>I propose that all future software sales will come from Microsoft, as
>the nontechnical 90% of the world suddenly starts using computers this
>year because of the Web. It's going to happen instantaneously. This
>is a prediction about the most major discontinuous change.
<ROFL> Those are some damned good drugs you're doing!
>Microsoft will become a software monopoly for all the people who are
>about to explode onto the Web. The Web will make the OS war
>unimportant at the same time as Microsoft's OS captures the billions
>of people, the vast majority of people who have never given a thought
>about NeXT vs Amiga vs OS/2 vs Windows vs Mac vs Linux. The vast
>majority of people are completely unaware of the OS war, and they are
>about to start surfing the Web. If ever they have to pick an OS, they
>will choose Microsoft, because it has the marketing mindshare and the
>desire to capture these people. Windows is positioned as everyman's
>OS; it is in no way elite. It is humble and "good enough". It is
>very Bob.
<GAG!><CHOKE!>
>If most people end up using cheap Internet terminals, they might find
>themselves using Sun hardware (Java chips) or accessing Web pages that
>happen to reside on Unix servers. Or maybe mainframes or NT servers.
>I sure haven't read much about OS/2 Web servers yet.
>This year, there will be a huge explosion as the majority of people
>all over the world, rather than a very small minority, rush onto the
>Web. The market for computing power -- whether desktop computers,
>servers, or terminals -- is going to grow more this year than it has
>in all past years, combined.
You keep proceeding from what are at best... "faulty assumptions".
Before the "rush" you are talking about, can really happen, the
low-cost, high-bandwidth connections will need to become reality.
>What hardware and OSs will predominate a year from now, after this
>remarkable event that changes the whole scope of computing? Watch out
>-- we're all about to be trampled by discontinuous change.
So... Now you've pontificated for several paragraphs, and all you can
come up with is that things are going to change... a LOT?
Golly! You must be one of those "Rocket Scientists" I keep hearing
about.
Bob M.
Michael Hoffman
>good luck,
>ian nelson
What's your prediction for the most likely future scenario regarding
the Web's impact on the various OS markets? Servers, desktop OSs, and
Internet terminals might have very different market shares. For
example, it's highly possible that most servers will be Unix, while
most desktop computers will run some form of Windows. Your scenario
could affect my decision of how much to learn about the various web
tools and environments.
Michael Hoffman
>Michael Hoffman wrote:
>>
> <snip>
> <extremely long diatribe about NT, Unix, OS/2, etc..>
>>
>> -- Michael Hoffman, EE, technical writer, Web developer,
>> techno-trendmonger
>Wow! What a great bait!
>--
>|Yusseri Md. Yusoff |
>|Dept. of Electronic Systems Engineering | email:ymy...@essex.ac.uk
>|University of Essex |
What do you think, will the server market be split 50/50 between NT
and Unix a year from now? If Microsoft gets its head together and
merges Win95 and NT, is it guaranteed that NT will explode and
overshadow the Unix server market that is relatively large today?
raster
>Now this is just sad. It's obvious that you put a lot of time and effort
>into composing this post (137 lines!), but you'll never be able to
>successfully troll a newsgroup until you learn how to properly forge a
>newsgroup header.
>-Erik
Come on, nothing wrong with a good debate, isn't that what *.advocacy groups are
about. He brought up some points for discussion, tact may not be his strong
suit, but if people bruise easily, they need to find another venue for their
expression. Besides every thread has to start somewhere.
With Respect.
Raster
Michael Hoffman
>I read this and it gives me a good chuckle. Of all the graduate E.E.
>and Comp. Sci. friends that I have, all laugh when the idea of NT
>replacing UNIX comes up. The only people who would buy the
>self-prophesising dribble in most industry rags are those who can't
>think and evaluate for themselves.
>Jason
Whether or not they can think and evaluate for themselves, a lot of
people are eager to buy NT, *if* certain changes are made. A lot of
people buy Unix servers reluctantly and wish that NT were a little
more mature.
Graduate EE and CS students can be blind to factors that are as
important as the technical pros and cons. If technical factors were
so important, NeXT would have taken off. Marketing, mindshare,
pricing, and industry attitudes are in some ways more influential than
technical factors. When the NeXT came out, who would have thought
that developers would be forced to abandon it in favor of Windows,
which was in an ugly, early version at the time? The NeXT watchers
laughed at Windows, but nevertheless were forced to accept it.
Pricing strategy is crucial. Microsoft has been very careful to never
ask for too much money all at once. The NeXT and its pricey
applications required a large lump sum. This demand for a financial
sacrifice all at once prohibited OS/2, Mac, Unix, and NeXT from taking
off. The MS solution is not cheap when you total the costs, but the
ability to incrementally upgrade PCs using building blocks from a vast
arena of suppliers has enabled people to make baby steps toward
ever-better versions of Windows. The next step is for the Win95 crowd
to inch toward NT Workstation. Then, look out.
NT is currently a joke compared technically to Unix.
Windows was a joke compared technically to NeXT.
Windows continually improved and is thriving, while NeXT is dead.
The grad geeks can and should laugh at the idea of NT replacing Unix.
This highly possible outcome is as ironic as Windows racing past NeXT.
Michael Hermann
|> Sorry if I'm excessively cross-posting here. I'm exploring the
|> relationship between the OS war and the Web revolution. I should
|> probably differentiate these issues into 2 or 3 threads, but they are
|> interconnected in my mind. I hope this combination of concerns is
|> interesting to you. If the thread explodes or wastes bandwidth, maybe
|> someone out there can extinguish it.
|>
|> I am becoming a professional Web developer. I have to decide what
|> tools to learn and what skills to develop. I'm pretty much at the
|> start of my career, and I want to learn systems that will be relevant
|> in the future.
|>
|> Currently, most Web servers run Unix. But NT and the free MS server
|> software have been getting favorable results. If I just looked at a
|> static snapshot of today's dominant platforms, I would say that
|> learning the Unix environment for Web development is a much better
|> idea than committing to an NT environment. However, looking at the
|> *direction* that things are going, looking at the latest *trends*, I
|> think NT has a much better future than Unix.
Given that the trend is that NT had no market share and now has a very
small one you call this rising market share while in reality it means
having ang gaining little marketshare.
|> I think that technical arguments can miss the point. Superiority of a
|> technology does not correspond with dominance of that technology.
Correct
|> I am confident that NT will quickly become the dominant operating
|> system for the next few years, whether or not it is technologically
|> superior. I think it's a better career move to align myself with
|> NT-based Web development, than Unix-based Web development.
NT will get a piece of the cake, a bigger one than it has now but it
won't become as dominant as Win3.1 was. NT can be lucky to get something
like 20-30% IMO.
|> I am facing some hard decisions now. Currently, the recommended way
|> to write CGI scripts is to code in Perl on Unix. But I think there
|> will be greater demand for NT-based CGI programming, over the next few
|> years.
|>
|> I'm facing a similar hard choice between Java and Visual Basic Script.
|> I am sure that I want to become a specialist in Microsoft solutions,
|> because I like Microsoft products and they are going to become the
|> most popular. Microsoft skills will have the greatest demand.
Go with Java, VBS won't be on par
|> I want to learn the most appropriate Microsoft-oriented tools and
|> environments for Web development. I don't want to be ignorant of
|> Unix, Perl, and Java, but I find I have to resist jumping into the
|> Unix/Perl camp, which is currently dominant and tries to pull me in.
|> Unix is yesterday's revolution and I'd rather minimize my involvement
|> with it. NT is far from perfect, but the NT environment and mentality
|> is far more up-to-date than Unix, which has its roots in the early
|> 70s. The Unix mentality is obsolete, never mind the technical
|> details.
The one OS fits all mentality is obsolete, Unix is king of the hill
and will remain it for quite a while, NT won't even come close to
replacing it because it just can't handle it.
|> While Unix will live on, it will be killed in the sense of being
|> overwhelmed by NT, as everyone in the world becomes involved in
|> computers. The 90% of people who have not yet touched a computer will
|> not buy into Unix; they will buy into NT.
Forunately those 90% don't make decisions for other peoples PCs.
I don't see why I should go to the uninformmed for advice as you seem
to be doing.
|> Internet terminals will shield the masses from dirtying their hands
|> with either NT or Unix. These terminals will be served by Unix
|> servers at first, but soon, the entire business world will switch to
|> NT. Technical issues aside, Unix has no future; Microsoft is rapidly
|> becoming the OS monopoly. OS/2 is completely out of the picture.
Say, why do you repeat this all over ? Do you have to convince yourself
of it ?
MS has been the OS monopoly, as of now they are starting to lose it.
|> Technical people might hate this trend, but based on reading the
|> technology and business magazines and reports, NT and Microsoft are
|> unstoppable. Whether or not this is desirable, it's happening, and
|> I'm going along with it.
Reading the mags MS may seem to be unstoppable, but that doesn't
reflect reality. MS is hard pressed to even stay in the position
they already have, much less replace Unix.
|> Microsoft is the right balance of price, consistency, and popularity.
Price and popularity yes, but nothing else (what do you mean by
consistence ? constantly making inferior products ?)
|> I've always hated Apple because they were overpriced when I bought my
|> first computer in 1988.
I never liked Apple either, never bought one either. Why did you buy one ?
|> I've always hated Unix because it's essentially an antiquated, chaotic
|> character-mode environment with no serious GUI and poor compatibility
|> among the zillion flavors.
It's character mode and has several GUIs and is higly compatible.
It is far from antiquated, NT has a lot to learn to get to where
Unix is now (and I don't mean market size..)
|> I've always hated OS/2 because it's made by IBM.
Fine, whith decisions that sound and reasonable I cant' see why
you should have any problems with the outcome..
|> I've always liked Windows because it's affordable and dedicated to
|> consistency of user interface.
ha, it's a lot less consistent that most other GUIs. It is affordable
though but that is the only positive thing I can associate with it.
|> Windows may suck in some ways technically. But technological
|> superiority is not the driving factor for popularity. The important
|> thing is the balance of price, popularity, and ease-of-use, as well as
|> technically being good enough.
again this crap, the important factor is preloads on the PC,
superiority on other machines
|> Windows is technically good *enough*.
no, never was.
|> In choosing and advocating an operating environment, find the best
|> *balance* of:
|>
|> o ease-of-use
|> o price
|> o popularity/ubiquity
|> o technical performance
|>
|> Linux or OS/2 or NextStep may be in some ways technically better than
|> NT. Macintosh may in some ways have better ease-of-use than NT.
|> Amiga may in some ways have better price/performance than NT (or 95).
|>
|> But the real deciding factor is *balance*.
Why do you believe that everyone needs the same thing you do or has the
same balance between the above. IMO most people (except for PC buyers)
buy the OS that fits their needs best, the PC people buy preloads..
|> When you consider the *balance* of factors, Microsoft Windows and NT
|> has trounced the others and deserves to become predominant.
|> Technological superiority of Linux or OS/2 is merely one factor to
|> consider. I respect technological considerations, but they are just
|> one consideration among several.
Where does it have any useful balance, it's easy to use has a laughable
GUI (both 3.1 and '95) and lacks power when copmpared to Unix, needs
to many resources when comapred to PC OSes.
|> I've never had to think deeply about justifying my commitment to
|> Windows until the opportunity to become a Web developer came along.
|> Now I must decide whether to learn Java or Visual Basic Script, Unix
|> or NT, and coding CGI in Perl or Basic. I concretely faced these
|> decisions when standing in the largest bookstore in the region -- the
|> Stanford campus bookstore, which currently has many more books about
|> the Unix operating system than on the Windows operating system. Each
|> substantial book costs $50. Should I buy a $50 Unix book, or a $50 NT
|> book? A $50 Visual Basic book, or a $50 Perl&CGI book? I still am
|> unclear about which tools and skills to master to specialize in Web
|> development for Windows, and the problem is exploding as the entire
|> world realigns itself around the Web.
Web development, buy a book about the tools you want to use.
If I were to buy them I would buy one about NT and Java (or
Unix and java depending on what I intend to run it on)
|> I would appreciate any insight you can give me, especially regarding
|> Unix vs. NT trends relating to a Web development career. I hope the
|> responses, if any, contain *substantial* insights or criticism, not
|> just empty flames that waste everyone's time.
What do you mean by web development ? Creating WWW pages or setting
up a WWW site ?
|> The OS wars are being affected in several ways by the Web revolution,
|> including the Internet terminal concept and Java, which supposedly
|> render the OS war irrelevant.
I don't think the internet terminal will take off. I don't think VB Script
will fly either. Java OTOH is already close to unstoppable so that's the
way I would go, as to which OS, take what you want, it won't matter
much.
-Mike
Michael Hoffman
>>bad...@seanet.com (Michael R. Delahoz) wrote:
>>>In message <4goa2q$9...@shellx.best.com> - mic...@cybtrans.com (Michael
>>>Hoffman) writes:
>>>:>Linux or OS/2 or NextStep may be in some ways technically better than
>>>:>NT. Macintosh may in some ways have better ease-of-use than NT.
>>>:>Amiga may in some ways have better price/performance than NT (or 95).
>>>Netscape has already in great detail elaborated on the cost differential
>>>between Netscape's server solution and M$' own bullshit "solution."
>>Yes, I read that. Very interesting. But somehow unconvincing, or
>>beside the real point.
>You need to provide some justification for this conclusion. Just your
>saying it, does not necessarily make it so.
The numbers are valid, so far as they go. But I don't think that
calculating the total cost of server hardware and software is that
important. There are maintenance costs and compatibility issues. A
lot of companies that are used to buying all IBM equipment would like
to buy all Windows equipment, rather than having Unix servers and
Windows on the desktop. Currently they are refraining from buying NT
servers, but as soon as NT gets a little better, there could very well
be a stampede towards NT. One such crucial change for this to happen
is bringing the Win95 users closer to NT. I'd like to see Microsoft's
latest OS strategy (it's hard to keep track of). The Win3x/Win95/NT
split is the only thing holding MS back from really taking off.
>>>To completely implement the Netscape version an IS shop has to invest about
>>>$1,700.00 more or less, to do with M$' "free" solution which BTW only runs on
>>>NT Server, the projected costs are in the vicinity of $6,000.00
>>That's simple logic. I get the impression that large corporations are
>>yearning to move to NT, and MS's giveaway of the server software is
>>the excuse they will use to choose NT servers rather than Unix servers
>>running Netscape.
>Maybe Microsoft finally woke-up, and started charging what the
>software was actually worth.
I thought IIS was pretty much brand new.
>> It's not logical, it's twisted motives. Many
>>corporations *want* a monolithic Microsoft solution. This is the buzz
>>I'm picking up from reading various magazines and online news.
>These same "various magazines" have been declaring OS/2's demise for
>over 3 years, that I know of, and it still hasn't happened. Could it
>be that they are wrong about a few other things as well? Don't just
>take everything you read as "Gospel".
I've always thought the columnists were biased *towards* OS/2. I sure
don't take the magazines as gospel. When Win3.0 came out, I instantly
knew it would take off like mad, but the first articles underestimated
the rapidity of its acceptance. The worst problem in the desktop OS
industry at the time was poor interoperability. Windows provided a
badly needed common environment to bring together all the PC software
and hardware. I was more intensely aware of this need than the
magazines.
I have often spotted a trend before the columnists, or violently
disagreed with them. There are still a few lingering articles by the
Internet contrarians -- "conservative analysts" who think the Internet
is a fad because it's (CURRENTLY) insecure and filled with 32 year old
white computer geeks. And articles that prove that the Internet
won't happen, because marketing analysis shows that only a certain
type of person is (CURRENTLY) on the Internet.
My predictions about the likely explosion of NT are based on what I
read in the tabloids, but I read critically.
>>I'm arguing from market perception, rather than from plain economics
>>or technical points. I've seen several references to NTs
>>"familiarity" and "ease of use" over Unix. This translates to lower
>>maintenance costs.
>That is generally what Microsoft does quite well, since it is so
>difficult to argue in favor of their products on technical grounds
>alone.
>> Windows is positioned as everyman's
>>OS; it is in no way elite. It is humble and "good enough". It is
>>very Bob.
><GAG!><CHOKE!>
*Because* of Microsoft's cutthroat strategy and marketing skill, they
are very well positioned to promote NT even if it is in some ways
technically inferior to Unix. Marketing is more important than
technology, to promote a product. The Bob attitude is disgusting, and
successful.
Unix is good at one thing only: networking. Microsoft is good at
marketing, usability, pricing, and incremental evolution, and is
getting better with networking.
>>There is a very large chance that Microsoft is about to become a
>>software monopoly far more than they are now, very soon. They've
>>passed a critical threshold of mindshare.
>I must not have been paying attention again... Just when did this
>happen? And WHY didn't somebody TELL ME?
It first happened when Windows 3.0 came out. It then happened during
the Win95 campaign, followed quickly by the Internet announcements and
demonstrations that showed the Microsoft was definitely committed to
the Internet. The current cover of _Internet World_ is "Microsoft
Declares War", showing general Bill commanding his tanks flying the
Windows flag. Windows 95 and Microsoft's Internet strategy are making
headlines everywhere, which have been telling you about the critical
threshold of mindshare that Microsoft surpassed between August '95 and
February '96.
>>I propose that all future software sales will come from Microsoft, as
>>the nontechnical 90% of the world suddenly starts using computers this
>>year because of the Web. It's going to happen instantaneously. This
>>is a prediction about the most major discontinuous change.
><ROFL> Those are some damned good drugs you're doing!
You think it will be business as usual in the computer industry, with
growth rates the same as they have been? The Web changes everything.
>>Microsoft will become a software monopoly for all the people who are
>>about to explode onto the Web. The Web will make the OS war
>>unimportant at the same time as Microsoft's OS captures the billions
>>of people, the vast majority of people who have never given a thought
>>about NeXT vs Amiga vs OS/2 vs Windows vs Mac vs Linux. The vast
>>majority of people are completely unaware of the OS war, and they are
>>about to start surfing the Web. If ever they have to pick an OS, they
>>will choose Microsoft, because it has the marketing mindshare and the
>>If most people end up using cheap Internet terminals, they might find
>>themselves using Sun hardware (Java chips) or accessing Web pages that
>>happen to reside on Unix servers. Or maybe mainframes or NT servers.
>>I sure haven't read much about OS/2 Web servers yet.
>>This year, there will be a huge explosion as the majority of people
>>all over the world, rather than a very small minority, rush onto the
>>Web. The market for computing power -- whether desktop computers,
>>servers, or terminals -- is going to grow more this year than it has
>>in all past years, combined.
>You keep proceeding from what are at best... "faulty assumptions".
>Before the "rush" you are talking about, can really happen, the
>low-cost, high-bandwidth connections will need to become reality.
All the pieces are about to fall into place to create discontinuous
change. A month ago, people doubted that agreements could be reached
soon. But the telecomm bill suddenly passed, and Visa and Master Card
suddenly came to an agreement. The race is on, to get on the Web, and
no one has time for deadlocks. The Web is now recognized as the
digital Manifest Destiny, even though computer old-timers might be
inclined to dismiss the hype as unwarranted and fail to see the
explosion coming so soon. All the projections I've seen are
reasonable and conservative. I predict that universal Web access will
happen much sooner than people expect. The problems will be solved
quickly, now that everyone has simply identified the shared goal and
grasped the possibility.
@Home http://www.home.net/
@Home is about to make 10 megabits per second available in Sunnyvale
for about $40 a month. That's 350 times faster than 28.8 kbaud.
Secure online transactions are about to become available.
>>What hardware and OSs will predominate a year from now, after this
>>remarkable event that changes the whole scope of computing? Watch out
>>-- we're all about to be trampled by discontinuous change.
>So... Now you've pontificated for several paragraphs, and all you can
>come up with is that things are going to change... a LOT?
I've predicted that NT servers will grow much faster than Unix
servers, but that the OS war will become less important. New
operating environments will have a chance, because they only have to
make themselves compatible with the Web, Java, and HTML or SGML. The
idea of a "platform" as an isolated island will be replaced by the
idea of flavors of environment, all interoperable. And I have asked,
what predictions do you have, for the impact of the Web on the OS war?
This investigation is driven by my need to decide which Web
development skills and environments to learn.
>Bob M.
Michael Hoffman
>er...@radix.net (Erik Blomquist) wrote:
>>-Erik
>With Respect.
>Raster
Debate? On Usenet?
Jane, you ignorant slut:
Allan S. MacKinnon
> I read this and it gives me a good chuckle. Of all the graduate E.E.
> and Comp. Sci. friends that I have, all laugh when the idea of NT
> replacing UNIX comes up. The only people who would buy the
> self-prophesising dribble in most industry rags are those who can't
> think and evaluate for themselves.
> Jason
>
Uhh...
Sorry to be the one to break the news to you, but
NT is replacing Unix everyday.
ASM
--
--
Allan MacKinnon
alla...@blueprint.com Boston, MA
NeXT / MIME Mail Welcome (617) 424-0615
Sangria
>Netscape has already in great detail elaborated on the cost differential
>between Netscape's server solution and M$' own bullshit "solution."
Ha. HA!
That was one of the better pile of stinky I've read in a long while.
Gotta love that funky cost break down where they equated Livewire Pro
to SQL Server, VB Pro and IDC. Right, Livewire Pro is a full fledge
SQL based backend server and also contains a language that directly
supports OCX, VBX, and thousands of third party plugins to do pretty
much anything a full blown development language can do...
Futhermore, they're quite right, ODBC support under NT with Netsite
is so much better than ODBC support under NT with IIS. Really interesting
line of argument there... =*-\
Glad to see other's are just as capable of FUDding as MS is... :-)
-- Sang.
*************************************************************
* Sang K. Choe san...@inlink.com *
* http://www.inlink.com/~sangria/index.html *
* Or http://sangria.inlink.com/index.html *
* finger: sa...@sangria.inlink.com *
*************************************************************
Christopher Robato
:>
:>Sorry if I'm excessively cross-posting here. I'm exploring the
:>relationship between the OS war and the Web revolution. I should
:>probably differentiate these issues into 2 or 3 threads, but they are
:>interconnected in my mind. I hope this combination of concerns is
:>interesting to you. If the thread explodes or wastes bandwidth, maybe
:>someone out there can extinguish it.
:>
:>
:>I am becoming a professional Web developer. I have to decide what
:>tools to learn and what skills to develop. I'm pretty much at the
:>start of my career, and I want to learn systems that will be relevant
:>in the future.
:>
:>Currently, most Web servers run Unix. But NT and the free MS server
:>software have been getting favorable results. If I just looked at a
:>static snapshot of today's dominant platforms, I would say that
:>learning the Unix environment for Web development is a much better
:>idea than committing to an NT environment. However, looking at the
:>*direction* that things are going, looking at the latest *trends*, I
:>think NT has a much better future than Unix.
This has been said well before NT was released, that Unix will be dead when
NT come out way back in 1992. (It was delayed and came out in 1993.)
If I were to take the quality of NT predictions from the very beginning, I
would say Unix should have a bright future instead.
:>
:>I think that technical arguments can miss the point. Superiority of a
:>technology does not correspond with dominance of that technology.
:>
:>I am confident that NT will quickly become the dominant operating
:>system for the next few years, whether or not it is technologically
:>superior. I think it's a better career move to align myself with
:>NT-based Web development, than Unix-based Web development.
Technological superiority is what gets things done, and not better
marketing.
It's like saying, we should all eat McDonald's because McDonald's is better
marketed, better supported, and better organized than mother's own home
cooking, even though home cooked meals are probably superior nutritionally
over a Big Mac.
:>
:>I am facing some hard decisions now. Currently, the recommended way
:>to write CGI scripts is to code in Perl on Unix. But I think there
:>will be greater demand for NT-based CGI programming, over the next few
:>years.
:>
:>I'm facing a similar hard choice between Java and Visual Basic Script.
:>I am sure that I want to become a specialist in Microsoft solutions,
:>because I like Microsoft products and they are going to become the
:>most popular. Microsoft skills will have the greatest demand.
*Will* that's the key word. It hasn't happened, and may not happen.
The consensus of the industry is that JAVA and Unix, led by Sun and Netscape,
won the Internet war, at least the first round (if you consider the first
round decisive.)
I thought that Microsoft has conceded, licensed JAVA, and within weeks,
scrapped Blackbird.
Your decisions is not the least of the "hardships" you will face now. The
greatest hardship happens when you found out you entered the wrong fork of
the road.
:>
:>I want to learn the most appropriate Microsoft-oriented tools and
:>environments for Web development. I don't want to be ignorant of
:>Unix, Perl, and Java, but I find I have to resist jumping into the
:>Unix/Perl camp, which is currently dominant and tries to pull me in.
:>Unix is yesterday's revolution and I'd rather minimize my involvement
:>with it. NT is far from perfect, but the NT environment and mentality
:>is far more up-to-date than Unix, which has its roots in the early
:>70s. The Unix mentality is obsolete, never mind the technical
:>details.
:>
:>While Unix will live on, it will be killed in the sense of being
:>overwhelmed by NT, as everyone in the world becomes involved in
:>computers. The 90% of people who have not yet touched a computer will
:>not buy into Unix; they will buy into NT.
:>
Another puppet marketing Microsoft's inevitability, regardless whether it's
good for customers and the industry in general for the long run.
:>Internet terminals will shield the masses from dirtying their hands
:>with either NT or Unix. These terminals will be served by Unix
:>servers at first, but soon, the entire business world will switch to
:>NT. Technical issues aside, Unix has no future; Microsoft is rapidly
:>becoming the OS monopoly. OS/2 is completely out of the picture.
:>
Internet terminals are going to have Netscape/JAVA interpreters in ROM, and
not exactly Unix. It's back to the late Seventies/early eighties home
computers, where the OSes and Basic interpreters are packed in built in ROMs.
Lately, I noticed an increasing number of providers and web servers using
Linux. I think that's going to be a factor. Linux can be had at almost no
cost and without any restrictive user licensing of sorts. Linux gestated out
of the Internet, and Linux is helping in turn, to build the Internet.
:>Technical people might hate this trend, but based on reading the
:>technology and business magazines and reports, NT and Microsoft are
:>unstoppable. Whether or not this is desirable, it's happening, and
:>I'm going along with it.
It's not desirable, and it's not happening, and you're just reading hype.
:>
:>
:>_________________
:>
:>Microsoft is the right balance of price, consistency, and popularity.
Consistency, hah. Windows 95 isn't even consistent with NT. The Gates Road
Ahead CD ROM, the Internet Explorer and even MSN, won't run in NT. That's
inconsistency on my book.
:>
:>I've always hated Apple because they were overpriced when I bought my
:>first computer in 1988.
:>
I never found Macintoshes more expensive than comparable *high quality* brand
namePCs. Junk PCs are always cheaper, undoubtedly, but anything with class
always have some more cost margin.
:>I've always hated Unix because it's essentially an antiquated, chaotic
:>character-mode environment with no serious GUI and poor compatibility
:>among the zillion flavors.
Woow. Some of the Unix GUIs are among the nicest I have seen visually.
NeXTSTEP looks impressive, followed by HP-UX's OpenVUE interface and
derivations of it like AIX's CDE. Then there is SGI's IRIX, which is said to
be a knockout.
:>
:>I've always hated OS/2 because it's made by IBM.
:>
What kind of logic is that? I should just say I've always hated Windows
and NT because it's made by Microsoft.
:>I've always liked Windows because it's affordable and dedicated to
:>consistency of user interface.
Affordable yes, only for Windows 95. NT isn't for most people.
Consistency of User Interface? The only consistency in the Windows
Interface is that it sucks.
There is never anything consistent with Windows interfaces, past, present and
future.
If anything, the Windows interface, in it's various iterations, has been the
most criticized, and I can go very long on the details.
:>________________
:>
:>
:>Windows may suck in some ways technically. But technological
:>superiority is not the driving factor for popularity. The important
:>thing is the balance of price, popularity, and ease-of-use, as well as
:>technically being good enough.
:>
:>Windows is technically good *enough*.
:>
:>In choosing and advocating an operating environment, find the best
:>*balance* of:
:>
:>o ease-of-use
Macintosh.
:>o price
Linux.
:>o popularity/ubiquity
Windows
:>o technical performance
Unix.
:>
:>Linux or OS/2 or NextStep may be in some ways technically better than
:>NT. Macintosh may in some ways have better ease-of-use than NT.
:>Amiga may in some ways have better price/performance than NT (or 95).
:>
:>But the real deciding factor is *balance*.
If performance/price is concerned, Linux takes the cake, followed by OS/2.
Both in numbers are more popular than NT. NT only outperforms both when
sufficient (32mb) memory is offered. All three are very powerful and
flexible. Linux by virtue of being free, is the cheapest, followed by OS/2
in the Red spline version. NT is not even close in price.
Windows 95 is a different creature. I would regard it inferior technically
to Linux and OS/2, and a tad better than MacOS in performance, but not in
ease of use. Unfortunately, it's the Windows that's closest to Linux and
OS/2 in price.
:>
:>When you consider the *balance* of factors, Microsoft Windows and NT
:>has trounced the others and deserves to become predominant.
I don't think so. It will claim its niche, but it won't become predominant.
:>Technological superiority of Linux or OS/2 is merely one factor to
:>consider. I respect technological considerations, but they are just
:>one consideration among several.
:>
Based on other considerations, Linux and OS/2 will also do quite well.
I would think that heterogenous diversity would be more of the rule of the
day.
It won't matter what OS or machine you are using, so long it's going to have
JAVA.
:>I've never had to think deeply about justifying my commitment to
:>Windows until the opportunity to become a Web developer came along.
:>Now I must decide whether to learn Java or Visual Basic Script, Unix
:>or NT, and coding CGI in Perl or Basic. I concretely faced these
:>decisions when standing in the largest bookstore in the region -- the
:>Stanford campus bookstore, which currently has many more books about
:>the Unix operating system than on the Windows operating system. Each
:>substantial book costs $50. Should I buy a $50 Unix book, or a $50 NT
:>book? A $50 Visual Basic book, or a $50 Perl&CGI book? I still am
:>unclear about which tools and skills to master to specialize in Web
:>development for Windows, and the problem is exploding as the entire
:>world realigns itself around the Web.
Oh, I see. That's the real issue of your post. You can't decide what book
you want to buy.
If you are making money in your development, deciding which $50 book to buy
should not be an issue. If you're a struggling college student, I can see
where it's a problem. A professional would more likely consider the issue of
his/her time and effort to divest in which direction, not the cost of the
books.
I say, buy the Unix one, and the Perl one. That's what the Internet is using
now, and NOW means momentum for the near future, and momemtum just begets
itself, so that what you have -now- is still likely to be the dominant one in
the long term. When you can afford it, buy the NT and VB one later on. If
NT is the future, it can afford to wait, and you can afford to save money to
buy the NT books later on. Besides, the current NT 3.x interface is about to
be obsoleted for NT 4.0, and books teaching how to use the current NT 3.x
will be just as obsolete. Study Visual Basic if you are interested in doing
business applications---I don't think VB will win over JAVA as the Internet
language. Perl would be for Unix scripting, but for web development,
anything on JAVA and HTML should be it.
As for me, $200 is no object for me, so I would just buy all these books you
mentioned, and study them. The more you learn the better and "richer" you
are inside. I personally got a few hundred dollars worth of OS/2 literature
here.
:>
:>I would appreciate any insight you can give me, especially regarding
:>Unix vs. NT trends relating to a Web development career. I hope the
:>responses, if any, contain *substantial* insights or criticism, not
:>just empty flames that waste everyone's time.
:>
:>The OS wars are being affected in several ways by the Web revolution,
:>including the Internet terminal concept and Java, which supposedly
:>render the OS war irrelevant.
:>
:>
:>-- Michael Hoffman, EE, technical writer, Web developer,
:>techno-trendmonger
:>
Rgds,
Chris
* * * Sailor Moon Joins Team OS/2 * * *
Clouds of FUD covers the city, turning millions into lemmings.
Holding Warp CD ROM up, she shouts, "Warp 32 bit Power, Transform!"
Serena turns into Sailor Moon, and her friends into Sailor Mars,
Sailor Mercury, Sailor Venus and Sailor Jupiter. The Sailor Team OS/2
girls crash the Red Moon Palace. Hordes of lemmings rose to fight them.
Moonlight beaming behind their silhouttes, Sailor Moon threatens
the evil Queen Beryl Gates and her diabolical Windowverse forces.
"In the Name of I-B-Moon, I shall right FUD and that means you!"
>>>> cro...@kuentos.guam.net <<<<
Kim Sommer
: Yusseri Yusoff <ymy...@essex.ac.uk> wrote:
: >Michael Hoffman wrote:
: >>
: > <snip>
: > <extremely long diatribe about NT, Unix, OS/2, etc..>
: >>
: >> -- Michael Hoffman, EE, technical writer, Web developer,
: >> techno-trendmonger
: >Wow! What a great bait!
: What do you think, will the server market be split 50/50 between NT
: and Unix a year from now? If Microsoft gets its head together and
: merges Win95 and NT, is it guaranteed that NT will explode and
: overshadow the Unix server market that is relatively large today?
Wow! You really believe what you said don't you? How adding a poor GUI
to NT can make it more attractive as a server is beyond me. You honestly
think unix boxes are only used for webserving and being something to run
on a PC. There's a lot more that unix is being used for than just webstuff.
Banking transactions, instrumentation and test, telecommunications,
satellite systems, imaging, special effects, traffic control, warfare
systems, weather mapping to name a few. NT has a lot of bases to cover
to even come close to knocking on unix's door. Plus it needs to scale
better and be more open.
I guess after we (hopefully) open your eyes to the scale and folly of
what you propose with the timeframe you've stated, you will then try
to say NT will replace mainframes by next fall.
Cory Hamaski, you're up to bat on that one.
Have a nice day. I did.
Kim
--
Kim A. Sommer - kaso...@intersource.com
-or- kas...@dice.nwscc.sea06.navy.mil
//Team-OS/2, B5, aikido, things that go zoom//
lan...@wwnet.com
>bad...@seanet.com (Michael R. Delahoz) wrote:
>
>>Hoffman) writes:
>
>>:>Linux or OS/2 or NextStep may be in some ways technically better than
>>:>NT. Macintosh may in some ways have better ease-of-use than NT.
>>:>Amiga may in some ways have better price/performance than NT (or 95).
>
>
>
>Yes, I read that. Very interesting. But somehow unconvincing, or
>beside the real point.
Actually, there are several relevant points.
1) software cost of solution: How much $$$ does it take to REALLY implement
the software solution (eg: what else must you purchase?)?
2) hardware cost of solution: Intranets often get clobbered with hits. You dont
want to put substandard PC architecture hardware on a box getting on average
170 hits per second (SGI amd Netscape websites). The physical architecture just
cannot handle those loads. If you are setting up a frequently accessed
intranet, the underlying hardware and size of pipe going into that hardware is
one of the most important things you can imagine. Secondly, the software on
top of that hardware has to be finely tuned to that hardware. One must ask
themselves if NT's HAL allows for the necessary fine tuning of the drivers. As
I understand the data (though I may be wrong) this turns out to be extremely
important in the page serving and throughput results, and tends to explain
why Unix tends to be chosen for the high volume stuff.
3) setup and support costs: The box will need to be maintained. If it is
the front end for a database which is getting its own frequent transactions
per second, that cannot be down for any significant length of time, then
you need a high-availability failure tolerant setup. I am not sure such beasts
exist for NT, though they run nicely on SGI boxes. Back to the point of support,
the same group of people currently maintaining the unix servers in the company
can be allocated to maintaining the web server, with very little training (hell, I
taught myself what I needed in about 1 hour). With NT, these nice unix gurus
who have been maintaining your servers for a number of years either need
retraining, augmentation, or replacement. Now given how little time NT has been
out, and thus how few people have gained any significant understanding of
web service or related NT administration, this prospect is remote.
4) Time to implement: So there you are, you have just installed this nice
FDDI card in your new NT PC to handle the input pipe problem and.... the
driver is flaky. Microsoft tells you to call the card vendor.... The card
vendor tells you to call Microsoft (from a local friend who gave up in
frustration on this exact problem....). The Unix vendor tells you to read the
SYSLOG, describe the hardware and software config (with patches) and
depending upon the number of other problems in the queue, will get a patch out
to you in anywhere between 4 hours to 1 week. This patch will be added
to the source tree for the subsequent versions of the unix so HOPEFULLY it
will not break again (though sometimes the trees get truncated at a certain
point for new OS releases).
So there you are with your Fiber Channel card from Ancor, your motherboard
from Intel, your OS from MS, your display card from number 9 or Matrox, your
CD from Toshiba, your memory SIMMs from Hitachi/NEC....
and you get a weird spurious PCI/DMA error while you are playing
an MPEG. Who do you call? Intel? Microsoft? #9 or Matrox? Toshiba? Which
driver from what company is at fault?
Or you get a unix box (say an SGI). Any of this nonsense happens, you dial
1 800 800 4SGI. Someone will take down the info and open a service call.
If they are busy you might not get a call back for a few hours (2 has been
the most in my previous life outside SGI). If they are not busy, it is about
2 minutes. They walk you through whatever is needed to find the problem
component, even as far as walking you though the booting the sash[] stuff.
Once the problem is isolated they check their database of problems and solutions.
If it is in there, they walk you through the solution (including loading miniroots
for some patch installs). If not, they log a bug and the problem gets escalated.
Which model of service costs the end customer more? The "track down the
problem from a bunch of unconcerned volume sales organizations" possibly
spending several days to weeks in search of an answer/resolution, or the call
one number and ask a question?
sceanario: Netscape server unexpectedly dies on an SGI box. You bought
this box with support, and the netscape server was one of the purchased
options (actually it is bundled these days with the base units, but that is
a different story). Who do you call? Netscape or SGI? The answer is ....
1 800 800 4SGI ....
Amazing how that works. One point of contact. Support of this product
is made considerably easier for the customer.
Now try the same thing with an NT box. Which company do you call?
Microsoft? Netscape?
>>To completely implement the Netscape version an IS shop has to invest about
>>$1,700.00 more or less, to do with M$' "free" solution which BTW only runs on
>>NT Server, the projected costs are in the vicinity of $6,000.00
>
>>BB
>
>
>That's simple logic. I get the impression that large corporations are
>yearning to move to NT, and MS's giveaway of the server software is
>the excuse they will use to choose NT servers rather than Unix servers
>running Netscape. It's not logical, it's twisted motives. Many
Actually, from what I have seen, most corporate types want to
use existing hardware (eg unix boxes) for low hit rate web servers,
as the server software is an incremental upgrade cost for this unit.
On the high end, for high powered intra-nets or external sites, most
of my customers want to be able to handle massive numbers of hits
WITHOUT choking the server. Some even want to run databases that
handle order entry/lead generation/marketing research data collection
(also known as surveys)/etc. For this box, you need one bloody fast
hardware engine with tuned software running above it. Which
invariably means a Unix box.
>corporations *want* a monolithic Microsoft solution. This is the buzz
>I'm picking up from reading various magazines and online news.
I agree that there is a terrific amount of "information" about how
NT is massively overtaking unix for every application. However,
one must be careful when assessing the veracity of that information.
The reality of the situation from where I work, is that NT has been evaluated
and placed for the most part on the back burner, as it is not a viable
hetrogenous solution for existing hardware, nor is it currently a
viable replacement for unix boxes in a wide array of new applications. What is
happening seems to be that NT is offsetting W95 in several locations
where one power user decides that w95 is simply to fragile, and the
IS staff wont sign off on OS/2 or Linix for their work (mostly due to the
"information" to which you refer).
>
>I'm arguing from market perception, rather than from plain economics
>or technical points. I've seen several references to NTs
>"familiarity" and "ease of use" over Unix. This translates to lower
>maintenance costs.
For a new untrained sysadmin coming from a windows world, NT may
be easier to maintain initially. After one week of training, all but the most
painful unix problems can be handled by this same person. OTOH, most
unix sysadmins get their start in school, where they learn on their own,
so there is no shortage of UNIX sysadmins. The "ease of use" argument
for NT over Unix reduces to a non-sequitor. And since the lower maint
costs argument rests on the ease of use argument, well that comes
crashing down as well.
>There is a very large chance that Microsoft is about to become a
>software monopoly far more than they are now, very soon. They've
>passed a critical threshold of mindshare.
eg: they have a remarkably good marketing department. However,
I disagree with your asessment. Microsoft was late to the game on
the internet, with VRML, and with Java. Somehow I get the feeling
that this is going to come to haunt MS in a big way...
I noted an article by Jesse Berst in PC week recently, where he talked
glowingly about how MS was setting the 3d standards on the internet
with its activeVRML or whatever, and the MS proposal for VRML v2.
Amazingly, Jesse neglected to mention that VRML has already been hugely
popular on the web, it was developed in a consortia of SGI, TGS, and some
nice people on the USENET. A VRML 2.0 def was solicited, and amazingly,
Jesse picked up on the only one that was sponsered by Microsoft, as being
the obvious successor. Its a good thing that the Usenet people dont listen
to these pundits, as the MS solution was not selected. The moving worlds
def was ratified by a vast majority of people.
Contrary to popular belief and industry pundits, Microsoft does not set
all the standards, and nor are their propositions held in universally high
regard simply because they are who they are.
>I propose that all future software sales will come from Microsoft, as
>the nontechnical 90% of the world suddenly starts using computers this
>year because of the Web. It's going to happen instantaneously. This
>is a prediction about the most major discontinuous change.
This tends to be called a change-in-phase or a phase transition (these
"instantaneous" changes of state from non-users to users). Lets call this
a web melting phenomenon just for laughs (please allow me this lattitude as
a former physics type). The important point of this web melting is that people
will use machines to access the web to do their work. This is not certain,
though I do beleive it to be occuring even as we speak. Most people will use
the instrumentation handed to them. And they will not really care what
the instrumentation to surf this web is really running on. Whether it is
MS windows, OS/2 Warp, NT/W, a Mac, a Next box, or a Nintendo ultra 64, all they
care about is being able to surf. That is, the vast majority of people do not
really care what underlies their system as long as they can get work done.
So, then what matters is getting the boxes out there with the software to
enable this. This is the web melting medium, the instrumentation by which
people access this information on the web. Whatever it runs, they dont
care. As long as it runs.
As a commodity software house, Microsoft must get its
users to live by the credo "buy preloads and upgrade often" or it
will cease to be an important player in the software industry (just
look at all the rest of the SW giants who died after forgetting this rule).
In order to get the end user to buy preloads, it negotiates contracts. Some of
which have been targetted by the DOJ, though the DOJ has shown itself
to be a bit of a paper tiger, by folding on its investigations.
In order to get the end user to upgrade, the marketing department has
decided to label the OSes with the year date code "to make sure that the
customer knows when their OS needs to be updated"... or in english, to
provide a cash cow to MS.
Now here is the problem. What happens to MSes growth when they
have completely saturated the market with their products? Oh...
they cannot grow anymore.... my my... hence the need for this yearly
"upgrade" of software in order for them to SURVIVE.
Back to the "it runs" thead... Frankly all the customer cares about is
running their web browser or server, without caring what is underneath,
as long as they can do work on this system. So does it matter to them if
it is an SGI box, an OS/2 based PC or a Windows machine? No, as long
as they can get their work done.
Wow.... what is the advantage of going to a pure MS environment again? Oh yeah
its that work bit. But hey... wait a minute.... you can do the same stuff on
the SGI or OS/2 boxes as you can the windows box.
So, Microsoft Marketing went bonkers with promoting the idea that
MSes apps work the best on MSes own platform... and gradually the
journalists caught this bug as well (after having this message pounded into
them by these marketeers). So what do you think they write about?
"How MS will conquer the world" film at 11.
The source of all this "information" on NT acceptance and what corporations
are doing seems to fly in the face of what really is happening. However,
people seem more willing to listen to hype than to reality....
"Men are willing to believe the impossible, but not the improbable"... (forgot who
said this)
>Microsoft will become a software monopoly for all the people who are
>about to explode onto the Web. The Web will make the OS war
>unimportant at the same time as Microsoft's OS captures the billions
>of people, the vast majority of people who have never given a thought
>about NeXT vs Amiga vs OS/2 vs Windows vs Mac vs Linux. The vast
>majority of people are completely unaware of the OS war, and they are
>about to start surfing the Web. If ever they have to pick an OS, they
>will choose Microsoft, because it has the marketing mindshare and the
>desire to capture these people. Windows is positioned as everyman's
>OS; it is in no way elite. It is humble and "good enough". It is
>very Bob.
See above
>If most people end up using cheap Internet terminals, they might find
>themselves using Sun hardware (Java chips) or accessing Web pages that
>happen to reside on Unix servers. Or maybe mainframes or NT servers.
>I sure haven't read much about OS/2 Web servers yet.
Check out http://www.caldera.com for interesting stats on Web servers.
You will see OS/2 Warp in there.
You should note that I am not unbiased, as I am an employee of
a certain Unix hardware vendor, though I am writing this on my
own time and with my own account, on an OS/2 based machine
connected to the web. However, in my position, I deal with
customers on a daily basis, handling problems, proposing solutions,
and seeing their vast networks, and upgrade decisions.
Amazingly, it is almost completely unlike what the pundits say.
I would suggest learning Java/Perl combo. You will not go wrong
there (hint: massive installed base, and most rapidly increasing section
of the base are using this)
--
--------------------------------+------------------------+
--=< Joe Landman >=-- | "Compute for insight, |
lan...@wwnet.com | not numbers" |
Systems Engineer by day | Hamming |
Ph.D Thesis Author by night +------------------------+
--------------------------------+
lan...@wwnet.com
>rma...@oz.net (Bob Madden) wrote:
>
>>>bad...@seanet.com (Michael R. Delahoz) wrote:
>
>>>>In message <4goa2q$9...@shellx.best.com> - mic...@cybtrans.com (Michael
>>>>Hoffman) writes:
>
[...]
>Unix is good at one thing only: networking. Microsoft is good at
>marketing, usability, pricing, and incremental evolution, and is
>getting better with networking.
So I see you have never used an SGI box. Too bad.
talk about ease of use and administration....
Secondly... unix is good for anything requiring high powered hardware
to run reliably, such as compute and file servers, visualization and
CAD/CAM/CAE desktops, scientific and engineering software development
and testing platforms, etc. Amazingly SGI boxes are being used as game
development platforms these days as well, simply because the environment
is so conducive to such work and it is NOT HARD to use.
Pricing.... configure an NT commodity box with the same configs as a base
level Indy. You will find the Indy costing less than the NT box by a fair
amount (10-20%). Indys are not commodity boxes designed for the lowest
common denominator (LCD), and nor is its OS, IRIX. OTOH, NT MUST be designed
PRECISELY for the LCD hardware.
Marketing: Dont know anyone who can compete with MS. However, the
workstation arena has been known for its vituperative rhetoric, so
MS really has its work cut out for it...
Incremental evolution: I wont touch that one.
[...]
>It first happened when Windows 3.0 came out. It then happened during
>the Win95 campaign, followed quickly by the Internet announcements and
>demonstrations that showed the Microsoft was definitely committed to
>the Internet.
Wrong. It showed that MS was BEHIND the other companies that actually
had product on the market, and they needed to catch up or be left behind.
MS doesnt own this new platform (the internet) though it thought it could
leverage some level of control over it using MSN. It failed miserably, and
wonderful tools that we had heard and seen (like blackbird) got yanked,
cause they were going orthogonal to the direction customers were going.
You have to hand it to MS, they were late to the bandwagon, but were
at least smart enough to ditch a losing proposition in a battle they couldnt
win.
>You think it will be business as usual in the computer industry, with
>growth rates the same as they have been? The Web changes everything.
Yes it does. Email or call me for a copy of Silicon Junction Unplugged
CD. I will see if I can free one up. SGI has been doing business at explosive
rates with regard to web service, authoring, content authoring/creation, etc.
(I personally have set up quite a few, including what I refer to as
the "fastest in Michigan" at a customer web site, running over a 100 megabyte
per second (not megabit) HiPPI connection)
Yes the web changes everything. It levels the playing field in a way such that
technologically inferior products just cannot compete.
[...]
>I've predicted that NT servers will grow much faster than Unix
>servers, but that the OS war will become less important. New
As I have pointed out to mathematically inclined here previously,
having a small number of installed base and a small run rate can
lead to "fast growth". Eg: I have say 100k installed base, and a
run rate of 20k per month, then the pundits would say that I have a
200+% per year annualized growth rate. But, when it is all said and done,
and I compare to the 5 million installed base with a 50k per month run rate,
it sure looks like the 100k base is going to outstripe the 5 million installed base
real-soon-now. OTOH, by examining more than just the first derivative
and actually looking at the absolute values as well, you get a completely
different picture.
Which is what I percieve to be your problem. You need to look beyond the
1st derivative (the rate of increase) and into the installed base. There is
some interesting data on this at http://www.caldera.com . Check it out.
Quite interesting, and it does not support your argument, rather it savages it.
Michael Hermann
|> John Cathie wrote:
|> >
|> > :>Microsoft is the right balance of price, consistency, and popularity.
|> >
|>
|> underway once a year or so ?
|>
|> Being so or not - what do you have against change or innovation ? Have
|> you been in the IBM-field for too long ?
This has nothing to do with innovation, it has to do with backwards
compatibility and a consistent direction, both things MS is lacking.
|> > What popularity? About a million copies sold in 3 years? I think not.
|> > Price? Pay per server, pay per seat. They get you coming both ways.
|>
|> Oh - I guess the popularity goes about the Windows program, that everybody
|> knows rather well after almost 6 years. Which incidentally has changed over
|> the years - much to your annoyance.
I am not annoyed at windows changing, just at it changing too slowly
|> > :>Windows is technically good *enough*.
|> >
|> > Not for me it ain't. It failed to deliver the capabilities I required with
|> > version 3.0. It still fails to meet my standards. And all I want is a stable,
|> > reliable multitasking operating system.
|>
|> Well then, why are you opposed to change then ? Windows NT v3.51 is working
|> rather well to my knowledge. If all you want is stability & multitasking, then
|> you shouln't complain about NT.
He isn't opposed to change, why do you assume that being opposed to the
latest MS offering means being opposed to change. Perhaps it means the
changes are not far enough...
|> > :>When you consider the *balance* of factors, Microsoft Windows and NT
|> > :>has trounced the others and deserves to become predominant.
|> >
|> > ROTFLMAO
|> >
|> > :>I would appreciate any insight you can give me, especially regarding
|> > :>Unix vs. NT trends relating to a Web development career. I hope the
|> > :>responses, if any, contain *substantial* insights or criticism, not
|> > :>just empty flames that waste everyone's time.
|> >
|> > Re-evaluate the parameters and the information you have assimilated to this
|> > point. Some of the data is plain *wrong*.
|>
|> I don't think so. They are opinions about futures, hence they cannot be wrong
|> now.
Opinions can be wrong, the question is whether this can already be verified
today.
|> > UNIX is going to be around for a long, long time to come.
|>
|> Maybe so, but not as the most popular platform.
Certainly as the most popular platform, NT will take years to get even close,
if it ever will.
|> > And as for hating OS/2 because IBM sells it; buy OS/2 1.3 from M$ if you feel
|> > that strongly. It's rock solid. But I have to empathize with you, It's likely
|> > that I despise M$ as much as you hate IBM (for reasons I'll pass on in email
|> > if you really need to know). But if M$ ever came up with something superior
|> > to OS/2 (right, and pigs will fly), I'd try it, and maybe even use it. But it
|> > would take a lot to get me away from OS/2 to *any* other OS.
|> >
|> > I believe you are shortchanging yourself by dismissing OS/2 (and IBM in
|> > general).
I agree 100% here.
-Mike
Hans Andersen
>
> In message <4goa2q$9...@shellx.best.com> - mic...@cybtrans.com (Michael Hoffman
> ) writes:
> :>I am confident that NT will quickly become the dominant operating
> :>system for the next few years, whether or not it is technologically
> :>superior. I think it's a better career move to align myself with
>
>
>
> What consistency? APIs that change more frequently than my underwear?
You have to be careful about statements like that. Suppose you only change
underway once a year or so ?
Being so or not - what do you have against change or innovation ? Have
you been in the IBM-field for too long ?
> What popularity? About a million copies sold in 3 years? I think not.
> Price? Pay per server, pay per seat. They get you coming both ways.
Oh - I guess the popularity goes about the Windows program, that everybody
knows rather well after almost 6 years. Which incidentally has changed over
the years - much to your annoyance.
>
> :>Windows may suck in some ways technically. But technological
> :>superiority is not the driving factor for popularity. The important
> :>thing is the balance of price, popularity, and ease-of-use, as well as
> :>technically being good enough.
> :>
> :>Windows is technically good *enough*.
>
> Not for me it ain't. It failed to deliver the capabilities I required with
> version 3.0. It still fails to meet my standards. And all I want is a stable,
> reliable multitasking operating system.
Well then, why are you opposed to change then ? Windows NT v3.51 is working
rather well to my knowledge. If all you want is stability & multitasking, then
you shouln't complain about NT.
>
> :>When you consider the *balance* of factors, Microsoft Windows and NT
> :>has trounced the others and deserves to become predominant.
>
> ROTFLMAO
>
> :>I would appreciate any insight you can give me, especially regarding
> :>Unix vs. NT trends relating to a Web development career. I hope the
> :>responses, if any, contain *substantial* insights or criticism, not
> :>just empty flames that waste everyone's time.
>
> Re-evaluate the parameters and the information you have assimilated to this
> point. Some of the data is plain *wrong*.
I don't think so. They are opinions about futures, hence they cannot be wrong
now.
>
> UNIX is going to be around for a long, long time to come.
Maybe so, but not as the most popular platform.
>
> And as for hating OS/2 because IBM sells it; buy OS/2 1.3 from M$ if you feel
> that strongly. It's rock solid. But I have to empathize with you, It's likely
> that I despise M$ as much as you hate IBM (for reasons I'll pass on in email
> if you really need to know). But if M$ ever came up with something superior
> to OS/2 (right, and pigs will fly), I'd try it, and maybe even use it. But it
> would take a lot to get me away from OS/2 to *any* other OS.
>
> I believe you are shortchanging yourself by dismissing OS/2 (and IBM in
> general).
>
> Good luck whatever your decision.
--
Hans JS Andersen Royal School of Librarianship
Technical Director Birketinget 6, DK-2300 Copenhagen
mailto:h...@db.dk http://alpha.db.dk/esc/ha
Hans Andersen
>
> I read this and it gives me a good chuckle. Of all the graduate E.E.
> and Comp. Sci. friends that I have, all laugh when the idea of NT
> replacing UNIX comes up. The only people who would buy the
> self-prophesising dribble in most industry rags are those who can't
> think and evaluate for themselves.
You will be surprised to see, how many people fit that description.
You almost made the whole argument in one line.
Mats Andtbacka
>Sorry if I'm excessively cross-posting here.
please, don't apologize for your velveeta.
apologize because you posted content-free, argument-free opinionated
flamebait that belonged *nowhere* in the big seven, and hardly would
have fit even in alt.flame.
apologize because you had exactly nothing of substance to say, but
stated your humble opinion as were it The Truth(tm) nonetheless.
apologize because you did this in a way that could not possibly result
in anything but a massive, excessively crossposted flamefest.
apologize for not even having any half-decent reasons for holding your
opinions in the first place (your "hating X because Y makes it" - as
if that were an argument!)
then apologize for being an effing moron, cancel your article and go
lurk in news.groups for a year.
followups set appropriately.
--
"I'm more differed from than differing" -- Arthur Dent
Michael S. Barthelemy
> I propose that all future software sales will come from Microsoft, as
> the nontechnical 90% of the world suddenly starts using computers this
> year because of the Web. It's going to happen instantaneously. This
> is a prediction about the most major discontinuous change.
You're forgetting the 80/20 rule. Unless one has a monopoly protected by government regulation, you can typically only
achieve 80% market penetration with some number of other companies getting the other 20%. Of-course the numbers can
vary, but it's usually not enough to be worth arguing about. Microsoft will not be able to do any better than 80% of
the market.
Also, 90% of the world will not be suddenly using computers anytime in the near future. Trends show that the number of
people using computers is growing at a geometric rate. The world plods along at it's own speed, and neither you, I or
even Bill Gates can affect this.
Mike Barthelemy
m...@plexare.com
Michael Hermann
|> Yusseri Yusoff <ymy...@essex.ac.uk> wrote:
|>
|> >Michael Hoffman wrote:
|> >>
|> > <snip>
|> > <extremely long diatribe about NT, Unix, OS/2, etc..>
|> >>
|> >> -- Michael Hoffman, EE, technical writer, Web developer,
|> >> techno-trendmonger
|>
|> >Wow! What a great bait!
|>
|> >|Yusseri Md. Yusoff |
|> >|Dept. of Electronic Systems Engineering | email:ymy...@essex.ac.uk
|> >|University of Essex |
|>
|> and Unix a year from now? If Microsoft gets its head together and
|> merges Win95 and NT, is it guaranteed that NT will explode and
|> overshadow the Unix server market that is relatively large today?
No way, MS can be happy to have 10% by then.
If MS merges NT and Win95 nothing at all will change, Win95 already gets
served by NT. The only thing that will change is the sales rate of
Win95 and NT, but not of the two combined (at least not for the better).
-Mike
Michael Hermann
|> rma...@oz.net (Bob Madden) wrote:
|>
|> >>bad...@seanet.com (Michael R. Delahoz) wrote:
|>
|> >You need to provide some justification for this conclusion. Just your
|> >saying it, does not necessarily make it so.
|>
|> The numbers are valid, so far as they go. But I don't think that
|> calculating the total cost of server hardware and software is that
|> important. There are maintenance costs and compatibility issues. A
|> lot of companies that are used to buying all IBM equipment would like
|> to buy all Windows equipment, rather than having Unix servers and
|> Windows on the desktop. Currently they are refraining from buying NT
|> servers, but as soon as NT gets a little better, there could very well
|> be a stampede towards NT. One such crucial change for this to happen
|> is bringing the Win95 users closer to NT. I'd like to see Microsoft's
|> latest OS strategy (it's hard to keep track of). The Win3x/Win95/NT
|> split is the only thing holding MS back from really taking off.
What MS is holding back is that all their solutions aren't that great
|> >> It's not logical, it's twisted motives. Many
|> >>corporations *want* a monolithic Microsoft solution. This is the buzz
|> >>I'm picking up from reading various magazines and online news.
|>
|> >These same "various magazines" have been declaring OS/2's demise for
|> >over 3 years, that I know of, and it still hasn't happened. Could it
|> >be that they are wrong about a few other things as well? Don't just
|> >take everything you read as "Gospel".
|>
|> I've always thought the columnists were biased *towards* OS/2.
Now that really is funny...
Since when is the Bob attitude successful ?
Bundling got MS where they are today, everywhere else they failed.
|> Unix is good at one thing only: networking. Microsoft is good at
|> marketing, usability, pricing, and incremental evolution, and is
|> getting better with networking.
Incremental evolution at a snails pace. Marketing and pricing yes,
usability so-so, technology no.
|> >>There is a very large chance that Microsoft is about to become a
|> >>software monopoly far more than they are now, very soon. They've
|> >>passed a critical threshold of mindshare.
|>
|> >I must not have been paying attention again... Just when did this
|> >happen? And WHY didn't somebody TELL ME?
|>
|> It first happened when Windows 3.0 came out. It then happened during
|> the Win95 campaign, followed quickly by the Internet announcements and
|> demonstrations that showed the Microsoft was definitely committed to
|> the Internet. The current cover of _Internet World_ is "Microsoft
|> Declares War", showing general Bill commanding his tanks flying the
|> Windows flag. Windows 95 and Microsoft's Internet strategy are making
|> headlines everywhere, which have been telling you about the critical
|> threshold of mindshare that Microsoft surpassed between August '95 and
|> February '96.
Between 8/95 and 2/96 MS certainly didn't gain any mindshare, they
probably lost a lot. I don't know where you get your ideas.
|> >>I propose that all future software sales will come from Microsoft, as
|> >>the nontechnical 90% of the world suddenly starts using computers this
|> >>year because of the Web. It's going to happen instantaneously. This
|> >>is a prediction about the most major discontinuous change.
|>
|> ><ROFL> Those are some damned good drugs you're doing!
|>
|> You think it will be business as usual in the computer industry, with
|> growth rates the same as they have been? The Web changes everything.
Actually, yes I believe it will be business as usual. The web, while
a new and interesting factor, won't be that improtant. Most of the stuff
is and will be done without the net. The web is a new part of the
computing scene, but nothing drastic.
|> >>Microsoft will become a software monopoly for all the people who are
|> >>about to explode onto the Web. The Web will make the OS war
|> >>unimportant at the same time as Microsoft's OS captures the billions
|> >>of people, the vast majority of people who have never given a thought
|> >>about NeXT vs Amiga vs OS/2 vs Windows vs Mac vs Linux. The vast
|> >>majority of people are completely unaware of the OS war, and they are
|> >>about to start surfing the Web. If ever they have to pick an OS, they
|> >>will choose Microsoft, because it has the marketing mindshare and the
MS either already is the monopoly for those getting into the net, or
they won't become a monopoly. The net is too divers for a monopoly.
|> All the pieces are about to fall into place to create discontinuous
|> change. A month ago, people doubted that agreements could be reached
|> soon. But the telecomm bill suddenly passed, and Visa and Master Card
|> suddenly came to an agreement. The race is on, to get on the Web, and
|> no one has time for deadlocks. The Web is now recognized as the
|> digital Manifest Destiny, even though computer old-timers might be
|> inclined to dismiss the hype as unwarranted and fail to see the
|> explosion coming so soon. All the projections I've seen are
|> reasonable and conservative. I predict that universal Web access will
|> happen much sooner than people expect. The problems will be solved
|> quickly, now that everyone has simply identified the shared goal and
|> grasped the possibility.
|>
|>
|> @Home http://www.home.net/
|>
|> @Home is about to make 10 megabits per second available in Sunnyvale
|> for about $40 a month. That's 350 times faster than 28.8 kbaud.
|>
|> Secure online transactions are about to become available.
-Mike
Dmitry V. Irtegov
>Sorry if I'm excessively cross-posting here.
I glad you understand it :).
>I'm exploring the
>relationship between the OS war and the Web revolution. I should
>probably differentiate these issues into 2 or 3 threads, but they are
>interconnected in my mind. I hope this combination of concerns is
>interesting to you. If the thread explodes or wastes bandwidth, maybe
>someone out there can extinguish it.
[skip]
>I am facing some hard decisions now. Currently, the recommended way
>to write CGI scripts is to code in Perl on Unix. But I think there
>will be greater demand for NT-based CGI programming, over the next few
>years.
There is a Perl port for NT. Thence if you know Perl, you won't lose
in any case.
>I'm facing a similar hard choice between Java and Visual Basic Script.
>I am sure that I want to become a specialist in Microsoft solutions,
>because I like Microsoft products
Well, this reason is good enough. On other hand, I like many things, but
I do not trumpet about them through entire USENET.
>and they are going to become the
>most popular. Microsoft skills will have the greatest demand.
They definitely will have _some_ demand for some time, but
words "most popular" and "greates demand" are, at best, arguable.
Beware of job market saturation.
IMHO, people who know what they are doing will always be in better
position than people with limited skill set. :)
>I want to learn the most appropriate Microsoft-oriented tools and
>environments for Web development. I don't want to be ignorant of
>Unix, Perl, and Java, but I find I have to resist jumping into the
>Unix/Perl camp, which is currently dominant and tries to pull me in.
>Unix is yesterday's revolution and I'd rather minimize my involvement
>with it.
IMHO, words "appropriate" and "Microsoft-oriented" are contradictory
in most cases.
>NT is far from perfect, but the NT environment and mentality
>is far more up-to-date than Unix, which has its roots in the early
>70s. The Unix mentality is obsolete, never mind the technical
>details.
Practically all current concepts of OS design were established in
late 60s-early 70s. NT has it's roots in VMS, Unix, CP/M and Xerox Alto.
All three of them were designed in early 70s; main design concepts of VMS
are probably even older. Who knows release date of RSX-11?
[drivel skipped]
>Technical people might hate this trend, but based on reading the
>technology and business magazines and reports, NT and Microsoft are
>unstoppable. Whether or not this is desirable, it's happening, and
>I'm going along with it.
Should we create a topic "YKYHBRTRTLW" in alt.folclore.computers?
I.e "You Know You Have Been Reading Trade Rags Too Long When"
you start to beleive in ongoing Microsoft Era.
Or YKYHBRTRTLW you start crossposting that drivel to multiply
.advocacy groups.
>_________________
>Microsoft is the right balance of price, consistency, and popularity.
Consistency of what? Of delivering shit to uneducated consumers?
>I've always hated Apple because they were overpriced when I bought my
>first computer in 1988.
You hate things because they are overpriced?
Well, overpricing might be a reason for not buying them, but for _hating_?
I don't get it. In that case you should have continuous orgasm about Linux.
>I've always hated Unix because it's essentially an antiquated, chaotic
>character-mode environment with no serious GUI and poor compatibility
>among the zillion flavors.
Flamebait provocation? Posting this paragraph alone to several Unix-related
groups is a good way to significantly increase your mailbox :).
>I've always hated OS/2 because it's made by IBM.
Ugh huh. And there are many people who hate NT and even more who just
do not want to try it because it is made by Necrosoft. This is a reason
NT has less marketshare than it deserves.
IMHO, as more and more people try to use Win LY and/or 3.x for
doing serious work, number of these "irrational" Necrosoft haters
will grow exponentially.
>I've always liked Windows because it's affordable and dedicated to
>consistency of user interface.
Consistent one GPF per hour? Ya right.
>Windows may suck in some ways technically. But technological
>superiority is not the driving factor for popularity. The important
>thing is the balance of price, popularity, and ease-of-use, as well as
>technically being good enough.
>Windows is technically good *enough*.
Enough for what? For barely running a text editor on the Septium
with more RAM than my DOS partition?
Actually, while Windows 3.x/95 might be acceptable for standalone desktops
and/or small LANs (i.e. SOHO market), big heterogeneous corporate networks
and Internet put slightly other set of requirements on the O/S.
And, from my experience, both 3.x/LY are inadequate to these requirements.
Considering that NT is a bit heavy for desktop and Not There yet
for serious Internet server...
[skip]
>I would appreciate any insight you can give me, especially regarding
>Unix vs. NT trends relating to a Web development career. I hope the
>responses, if any, contain *substantial* insights or criticism, not
>just empty flames that waste everyone's time.
As your post reads, it looks that you already have set your mind.
>The OS wars are being affected in several ways by the Web revolution,
>including the Internet terminal concept and Java, which supposedly
>render the OS war irrelevant.
They might render _desktop_ OS wars irrelewant, and I hope they will.
But I'd love to see a $500 Internet appliance running SQL server :).
--
Cheers,
Fat Brother.
*
* When people make sweeping statements about "MS Crapware" or some
* such nonsense, it's almost directly related to trying to run those
* applications under Windows...
* Sang K. Choe, MS products advocate.
Donal K. Fellows
John Cathie <jca...@mindlink.bc.ca> wrote:
> In message <4goa2q$9...@shellx.best.com>
> mic...@cybtrans.com (Michael Hoffman) writes:
>> Microsoft is the right balance of price, consistency, and popularity.
>
I'm glad I don't work next to you then (I believe that they probably
leave the API alone for a week or so at a time... :^)
>> Windows may suck in some ways technically. But technological
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
You betcha!
>> superiority is not the driving factor for popularity. The important
>> thing is the balance of price, popularity, and ease-of-use, as well as
>> technically being good enough.
>>
>> Windows is technically good *enough*.
>
> Not for me it ain't. It failed to deliver the capabilities I
> required with version 3.0. It still fails to meet my standards. And
> all I want is a stable, reliable multitasking operating system.
Windows has a number of really amusing features, but being technically
good enough isn't one of them. It suffers from big-time clunkiness and
lack of speed (and I say this having had the benefit of seeing NT with
the new shell in operation - `running' gives the wrong impression -
the user of this particular system, who definitely knows what he's
doing, used to time the boot-up and reckoned that if it took over a
minute at one particular stage it had crashed. Sheesh! M$DOG is better
than that!)
>> When you consider the *balance* of factors, Microsoft Windows and NT
>> has trounced the others and deserves to become predominant.
>
> ROTFLMAO
I agree. ROTFLUMSH even.
NT has to go through quite a few more revisions before I'd assess it
as being even close to usable, and it's much vaunted security only
applies when it isn't networked (apparently). I think I'll stick with
Linux, which is at least (virtually) compatible (at the level I use it
at) with nearly every other system that I've used for work.
>> I would appreciate any insight you can give me, especially regarding
>> Unix vs. NT trends relating to a Web development career. I hope the
>> responses, if any, contain *substantial* insights or criticism, not
>> just empty flames that waste everyone's time.
For CGI scripts I use perl (if it isn't available for NT, it will be
:^) And Java looks to be the way that executable WWW content is
going. Currently, the most stable servers are all for Unices (though
this is helped by the fact that Unices are more stable than NT). I
particular, I have found that I am more likely to get a page first
time from a Unix-based site than an NT-based site, and companies _do_
care about stability and robustness. Make of that what you will.
> Re-evaluate the parameters and the information you have assimilated to this
> point. Some of the data is plain *wrong*.
>
> UNIX is going to be around for a long, long time to come.
Agreed. I would say that NT is only really a feasable option for a WWW
server if you have money to burn, and someone to run it. If this is
the case, you might as well have a Unix system instead and benefit
from the increased stability of the OS.
> And as for hating OS/2 because IBM sells it; buy OS/2 1.3 from M$ if
> you feel that strongly. It's rock solid. But I have to empathize
> with you, It's likely that I despise M$ as much as you hate IBM (for
> reasons I'll pass on in email if you really need to know). But if M$
> ever came up with something superior to OS/2 (right, and pigs will
> fly), I'd try it, and maybe even use it. But it would take a lot to
> get me away from OS/2 to *any* other OS.
Myself, I'd hesitate before using half an OS as a solution for
anything other than half-assed problems... :^)
But then, as a Linux advocate, I _would_ say that.
> I believe you are shortchanging yourself by dismissing OS/2 (and IBM in
> general).
That I would agree with. IBM has done a lot of good stuff, especially
on their mainframes.
Donal. [ Damn, I should have made this post shorter... :^) ]
--
Donal K. Fellows, (at work) | Donal K. Fellows, (at home)
Dept. of Computer Science, | 6, Randall Place, Heaton,
University of Manchester | Bradford, BD9 4AE
U.K. Tel: ++44-161-275-6137 | U.K. Tel: ++44-1274-401017
fell...@cs.man.ac.uk (preferred) | do...@ugglan.demon.co.uk (if you must)
--------------------------------------------------------------------------
<http://r8h.cs.man.ac.uk:8000/> for my home page
paul chemmanoor
(BOWEN JASON MICHAEL) writes:
> and Comp. Sci. friends that I have, all laugh when the idea of NT
> replacing UNIX comes up. The only people who would buy the
> self-prophesising dribble in most industry rags are those who can't
> think and evaluate for themselves.
> Jason
>
They laugh no more. And yet MVS is acknowledged to be better
than any of the OSes that now win popularity polls (and that
includes a broad spectrum from unixes to windows) in terms of
raw stability and reliability.
There is far more than technical superiority to consider when
talking about market shares of OSes.
2] And yes, I am a computer science graduate, and I have worked
with enough shades of unix (ultrix, hpux, NextStep, linux,
SunOS, Solaris, AIX ..) to appreciate what unix can do.
At the present time I run a mixture of Linux 1.3.61 kernel,
SunOS 4.1.3, Solaris 2.4, Win95, WFW, WinNT 3.51 Workstation and
WinNT 3.51 Server. In my experience the only truly bad OS in
that list is WFW [ in terms of what users complain about ].
I didnt realize I ran such a varied mixture of OSes until I
actually put that list down now. Anyway a small set of comments
about each.
1.Linux
(plus)
A computer science graduates dream :)
What fun it is to get AHA-2940 support in a special
kernel and then find out the 3C590 3 Com Etherlink III
isnt supported [ an excuse to get the latest kernel to
compile :) ]. On the whole impressive, runs in tight
space, you get free X, and DOOM :). A whole set of
Language utilities C, C++, Objective C, Lisp, Perl ...
A complete set of internet utilities.
Important net functionalities out of the box such
as sendmail, DNS, NFS ...
(minus)
In simple terms it is a nightmare for a non-cs person
to setup. Multitasking theoretically great but in practice
has problems with Netscape [ yeah I know this is probably
Netscape's fault.. but ... from a users perspective
it is a big problem ]
2.SunOS
(plus)
This is a revered OS of the Internet. Even today you will
probably find versions of internet utilities for SunOS 4.1.3
first before versions for any other OS. This is even today,
given enough resources, my first choice for a internet related
machine. NNTP, DNS, Sendmail ..
Stable. Slim [ Compared to Solaris ] Widely Supported.
(minus)
Needs a skilled Administrator. Much of the apparent ease with
which a unix admin gets his or her work done comes with a
good understanding of the fundamentals of unix. This understanding
is often mistakenly taken for granted by CS people. It is brought
home only when compelled to support the average business user.
e.g. Adding a www server given the binary in a inetd mode should
not take more than 15 minutes at the maximum and probably only 1 or 2
minutes at the minimum. But try explaining all the logic of what you
had to do to a new unix user from a non CS background amd you will see
what I mean .. This point becomes important when we consider NT.
And as for installation I think SunOS installation [ as well as nearly
all other unixes except linux is too much for a average business user
to handle ].
3. Solaris
Almost everything about SunOS applies to Solaris.
(plus)
Dynamic configuration of OS. No more building of kernels to tweak minor
parameter changes. NIS+. Better Multiprocessing. Slightly better Admin tools.
Future of SunOS. Probably best bet for long term Internetworking support.
(minus)
Bloated. Slow. Almost all the problems of administration inherited from
SunOS.
4. Win95.
(plus)
Better UI[ than win 3.1]. [ Yeah dont underestimate the importance of UIs.
Otherwise we
would still be rooting for block mode 3270s ]
Plug 'n Play. Yeah you have read how this is hype, but I was converted
when just after I plugged
in a cheap 4x Acer CDrom automagically a D: drive appeared which
could access said drive, with absolutely no help from me. Anyone
who has done addition of hardware will recognize the allure of such
a feature. I am sure there are problems with specific components but
on the whole a big step in the right direction. Win 3.1 compatibility.
I have not come across any 3.1 app that wont run under win95.
I am sure there are some, but the point is who cares as long as *you*
dont have to run those as part of your business ?
DirectX. Yeah you thought it was hype. I saw the demos.
Biggest asset: Makes win 3.1 people happy.
Yes I know they are starting from low expectations. :)
But the fact of the matter is when someone who has been whining about
how hard it is to be able to connect to Compuserve, Internet or whatever
under Win 3.1, is given Win95 they return with a happy grateful smile.
This is in my experience. Your mileage may vary.
(minus)
UI has a way to go. For the power user fvwm under Linux is probably
quicker and
more effective. Needs atleast 12 MB for reasonable performance.
In 8mb it thrashes
badly. Multitasking poor. It is more than what you get from Win 3.1.
But considerably
worse than any serious OS.
An extra note for this OS: The benefits of Win95 do not in my eyes
call for upgrading win 3.1 *unless* you are having problems that
you know go away with Win 95 : PPP, SLIP, TCP/IP ..
You have to measure the cost of upgrading[ price + downtime + effort ]
against the benefits.
On the other hand you probably would prefer Win95 over
Win 3.1 when you are buying a new machine, because price, downtime
and effort[ of installing] are all virtually zero in that case.
You see the customers are smarter than they get credit for :)
5. WFW 3.11[ or other variants such as Win 3.1 .... ]
(plus) um not much except the largest installed base and largest
selection of apps. :)
(minus) not much to leave out .. :)
6. Windows NT + Windows NT Server.
(plus)
Win 3.1 Compatability. Yes there are a few apps that wont run.
But they are few and far in between. We regularly run VB 3,
Oracle Win 3.1 clients, VC++ 1.5, MS Office Win 3.1 versions
without any real problems.
Stable. I have run NT from the early 3.1 betas.
The only crashes I have seen have been directly identified
to be hardware related: faulty IDE controller, bad Network card ...
We have been running NT servers for more than 2 years.
They are rock solid while running Oracle, SQL Server, Sybase etc.
Easiest user interface for System administration among the
serious OSes. Why do I call NT a serious OS ?
Because: It is Premptively MultiTasking, MultiUser [ some caveats apply],
Symmetricaly MultiProcessing, MultiThreading and C2 secure.
Increasingly adept at networking: Interoperable with SNA, AppleTalk,
DECNet, TCP/IP, IPX/SPX... networks.
Security model goes beyond vanilla unix rwx and into ACLs.
Domains and Trusted domain concepts simplify medium scale
system administration.
Journalled File System in NTFS.
Has base OS suport for Software RAID.
Extremely impressive integration with SQL Server.
[ Those tpc numbers talk dont they ? ]
Steadily accquiring internet toolset: nslookup, whois, DNS,
sendmail, WWW server, ftpd, telnetd, perl .., NNTP.
GCC binaries are available now. Almost the entire unix
utility set
has been ported quite a while ago. I run a free unixbin
set on my NT workstation.
Has an excellent built in performance monitor which can
be used to monitor all the vital signs of a machine
and yes you can do that remotely too.
Has exactly the same interface on multiple processor
platforms (Alpha, PowerPC, MIPS, Intel)
(minus)
Has no remotable GUI. [ big big minus ]
Has no directly exposable MultUser capabilities.
This is directly related to not having a remotable GUI
as the CLI equivalents of GUI commands are far in between
and a telnetd does not solve much.
Has immature internet utility set as of now.
Base NT does not have dynamic routing capability.
[ You can do static routing through some games with the
registry ]
Is a resource hog. It needs atleast 24 MB to run smooth.
[Although yes you can get pretty useful work out of 16 MB.
I ran a SQL Server in development mode on a 16 MB machine
without too much thrash.]
Has very few utilities with source available.
Doesnt scale beyond a small set of processors. [ This
is hearsay for me ]
Has a bad UI at this point.
5] And now the whole point of this discussion:
Why NT seems to be such a challenge to Unix.
The answer is simple: The set of things
that unix can do that
NT cannot is shrinking every day. The remoteable
GUI is already available from 3rd parties.
The UI has been updated to look like win95.
The Networking capabilities except for
internetworking already rivals unix. And Cairo
is slated to have both bay networks and CISCO
routing algorithms coded in.
As for Internetworking utilities: DNS, Sendmail,
ftpd, telnetd, gopherd, httpd are all available
in public domain and from 3rd parties.
A major boundary was crossed recently when MS
released the httpd server. I tested the server
against the other platforms. And I found IIS
( the MS solution ) to be faster and more
resistant to stress than Linux and SunOS.
Although it is true the other OSes were running
freeware the cost of IIS is zero too. And yes
IIS is running on a 32 MB pentium 60 while the
SunOS is on a 32 MB SparcStation 10. And the
linux is on a pentium 90 with 16 MB.
Put together the fact that NT now has the lowest
cost/performance ratio for productivity tools,
Databases [ yeah go take a look at the
tpC numbers ] and now the Web servers, and you can
see why the NT/Unix debate is heating up and certainly
is no
laughing matter.
People constantly underestimate the competetiveness
of Microsoft. It is worthwhile to recall the list
of companies MS has beaten before. If they percieve their
platform to be lagging because of some reason they will
add it if for nothing else than out of a desire to make
more money. [ I expect a remoteable GUI sometime soon. ]
In this respect I admire their willingness to swallow
their public utterances sometimes humiliatingly and to
address the problem. This in my eyes is better than
losing your business. Look to the example of Lotus which
refused to get on the Windows bandwagon until too late.
All this is not to say MS is a ideal company. They do
indulge in vaporware often. They do sometimes change
course in the middle of a product cycle. However with
NT they have
a product ahead of the hardware available in its target
audience. They have time on their side. With P6 and
cheap sub 25 dollar per MB memory NT Servers will be
formidable departmental servers. When they work out the
remotable GUI along with a decent telnetd and asociated
CLI versions of their programs unix will have *very*
few reasons to be chosen over NT at the departmental
level.
6] Okay now let me wind up :)
A note about university anti-NT fanaticism.
Conviction in an idea is no guarantee of its truth :)
Computer science grad students live in an sheltered
environment where they never have to worry about
the real world business compromises. [ And this is
good, first the fundamentals than the real world ]
Furthermore, CS departments are overwhelmingly unix
in flavour. So the "unix is best" argument is self propagating.
It is no accident that those of us who see some
good things in NT come from environments where multiple
OSes (not just unix flavours ) are the norm. From the
NT advocacy groups I see David LeBlanc, Sangria and Jim Frost
are all people with Unix backgrounds and I think all of them
run various flavours of unix [ and most run Linux for fun ]
in a day to day environment. Have you ever noticed its
the norm for NT advocates to be non-fanatic ? There is
a reason why.
For intelligable advocacy I would recommend reading
in some peace and quiet Maurice Bach's "Design of the
Unix Operating System" and then read Helen Custer's
"Inside Windows NT". If you can locate fundamental
weaknesses of Windows NT kernel you would be far more
justified in your hatred of NT. I am reasonably sure
you will come away surprised by how close NTs design
is to Unix [ in some ways better than vanilla flavours of
unix ]. Do not be fooled by the silly win 3.1 GUI.
After all changing it so radicaly to a Win 95 look has
already been demonstrated as trivial as far as OS
development cycles go. If remoteable GUI became such
a big Issue, how long do u think it would take for MS
to buy a company like Citrix that does WinFrame ?
Already some company has a product out for 99 dollars
[perhaps an evaluation copy ? any more info on this ? ]
which brings Posix.2 to NT. A complete unix system along
with things like symbolic links [ which std NT doesnt expose].
Posix.2 in turn makes NT attractive for the federal govt etc.
In short NT is gradually absorbing all that is good in Unix.
Which I think is a wise thing to do. Unix is a great system.
Which in turn brings up the question why buy a unix like NT
when you can buy the real McCoy: aka a real Unix like Solaris ?
And the answer is : It can now do everything Unix can and still run
the productivity tools, and still run the best cost/performance
ratio for databases, web servers, file servers, application
servers.
See the logic there ? :)
Today I would recommend different OSes for different tasks.
One thing I would not choose NT for, if given a choice is Internet
connectivity. Yes I know it can do most things but not as
gracefully as unix and not with as much felixibility.
Here Unix rules. So far ..
For databases I would choose a AlphaServer runnning 64 bit Oracle
on Unix,
if you had money to burn.
If you were on a tighter budget I would pick NT + SQLServer on a
dual processing P6 machine [ like HPs ].
7] Final note: I promise :)
The *negative* effects of rapid acceptance of NT are already visible
to me: Non CS grads who dont have the first idea about OSes but know
only to point and click because of NTs ability to hide so much, rather
like the electronics designer who claims a transistor is where the
two colored lines cross on a CAD program.
I actually am hoping NT doesnt become too successful: the long term
economic fatcors are too much in its favour. And I would love to see
competition among OSes. In my ideal world I would have Linux balance
out NT for the power user on the desktop. And NT challenged by unix
at the departmental level. The enterprise glue is probably going to be
NT in the future [ Unicenter-CA, SMS , Exchange Server, NetView etc
cant hurt ] for companies with upto say 1000 to 3000 employees. The
almost obscene glee emitted by penzoil, chevron and Intergraph are
precursors ..
But
beyond that unix will probably rule for a long while. Of course beyond 5 years
it is probably to be expected NT will no longer have much resmblance to
its current state. And neither will unix. For all we know we could be
running ATM at 622 Mps to our desktops making do with Java applets and
an internet terminal. :) :) :)
Stay Cool. Its not often I get the time to post.
So if you flame me, dont expect a guaranteed response.
Dont let that stop you tho.
Email could be better.
- Paul Chemmanoor.
Kelvin Parker
>
> Sorry if I'm excessively cross-posting here. I'm exploring the
<BODY CUT>
>
> -- Michael Hoffman, EE, technical writer, Web developer,
> techno-trendmonger
Michael,
I really appreciate this excellent post and analysis. I am not a
programmer or developer, but am responsible for our company's computer
network, and its web sites and future development. I have considered all
the issues you discuss, and have come to the same conclusion.
I love WebObjects/NeXTSTEP/OPENSTEP etc, but too expensive, not widely
enough used. I want a powerful compnay network, but forget command line
UNIX, we're going for WinNT. So on and so forth.
I hope to catch responses to your post (from those who technically know
better than I), for a business executive who has to make similar choices,
it will be a real learning experience.
Thanks,
Kelvin Parker
Neo-Tech Worldwide
Loren Petrich
Michael Hoffman <mic...@cybtrans.com> wrote:
>Sorry if I'm excessively cross-posting here. I'm exploring the
>I am becoming a professional Web developer. I have to decide what
>tools to learn and what skills to develop. ...
I'd recommend learning something that's just emerging, so you'll
get a jump on other people. And one thing that is doing so is electronic
commerce. The idea would be that one would do one's catalog-viewing,
order-composing, and paying online, and thus avoid a heck of a lot of
intermediaries. For that, you'll need to know about encryption (can't
have fake order forms :-), and some AI at both the client (Java) and
server (CGI scripts) ends. One problem here is what is the appropriate
division of labor between client and server -- should the server do all
the work, or should it upload a Java applet and let the client do all the
work -- or anything in between.
... However, looking at the
>*direction* that things are going, looking at the latest *trends*, I
>think NT has a much better future than Unix.
I don't think that the UNIX guys will take this lying down :-)
And this argument ignores all the WWW servers running the MacOS.
Yes, that fruit company's OS. I suggest that you check it out, because it
is absolutely uncrackable (there was a competition to do just that, and
it got no entries), and that will be vital for e-commerce.
>I think that technical arguments can miss the point. Superiority of a
>technology does not correspond with dominance of that technology.
Ever heard of a niche market? You could do well in such a market,
because you will simply have less competition.
[NT vs. UNIX...]
>I am facing some hard decisions now. Currently, the recommended way
>to write CGI scripts is to code in Perl on Unix. But I think there
>will be greater demand for NT-based CGI programming, over the next few
>years.
With the MacOS, one can write CGI scripts in AppleScript, which I
find to be a very elegant programming language. Currently, it is rather
slow, but if one uses the time span you suggest, then Apple will have
released Copland (System 8), and AppleScript will be much faster. Copland
will also have preemptive multitasking and protected memory, though
user-interactive processes will share one memory space and will have a
faked-out cooperative multitasking between them. However, file and
network stuff, what are essential to WWW servers, will have both of these
PM's.
Copland will have the advantage over NT of less resource
requirements -- unless, of course, M$ hires some *really* good
programmers :-) Unless Apple commits some serious design blunder, which
is unlikely, it will be at least as bulletproof as NT, and be as
uncrackable as the current MacOS (System 7.5). And it will also have a
superior user interface and compatibility with several years of 32-bit
MacOS apps (the MacOS has been 32-bit since Day One).
>I'm facing a similar hard choice between Java and Visual Basic Script.
I'd go with Java, since Sun is more committed to openness than
M$. Also, Java plug-ins are now more available than Visual Basic ones.
Even M$ has given in and decided to go along with Java.
>I am sure that I want to become a specialist in Microsoft solutions,
>because I like Microsoft products and they are going to become the
>most popular. Microsoft skills will have the greatest demand.
If you *like* M$ products, I'm not going to deprive you of your
fun, but I don't, with exceptions such as Excel. Do you know what my
favorite utility for managing files on PeeCee disks is? The MacOS Finder.
>Technical people might hate this trend, but based on reading the
>technology and business magazines and reports, NT and Microsoft are
>unstoppable. Whether or not this is desirable, it's happening, and
>I'm going along with it.
I wonder if the M$ guys slip the reporters some under-the-table
payments :-)
More seriously, M$ has remarkable gifts of self-hype; it has the
ability to make itself seem like it invented practically *everything*.
Apple is almost *totally* lacking in anything comparable, despite
occasional flashes of brilliance such as "1984". Thus, it was reported by
someone here that someone noticed a Macintosh and said that there was a
computer already running Windoze95.
>Microsoft is the right balance of price, consistency, and popularity.
I disagree, for numerous reasons.
>I've always hated Apple because they were overpriced when I bought my
>first computer in 1988.
Apple's repented of that disaster, happy to say. You'd be
surprised at how low some Apple prices now are. Furthermore, one can now
choose some Mac clones, such as those made by Power Computing. In the
months to come, the PowerPC Platform (PPCP; formerly the Common Hardware
Reference Platform [CHRP]) boxes will start coming out, and there will be
cloners galore competing in this market. You'll be able to run Copland,
any of several flavors of UNIX (AIX, Solaris, Linux, ...), and yes,
WindozeNT. And multi-boot will be built in to all but el-cheapo boxes, so
you can try out different OSes. And some PPCP-box makers will almost be a
*LOT* better at self-hype than Apple (it's hard to do worse :-).
>I've always hated Unix because it's essentially an antiquated, chaotic
>character-mode environment with no serious GUI and poor compatibility
>among the zillion flavors.
I agree with that assessment -- I've found UNIX *very*
user-hostile -- worse than other CLI OSes I've used, like VMS and even
VM/CMS (I still miss VMS after all these years).
>I've always hated OS/2 because it's made by IBM.
I don't feel that inclined to spite IBM.
>I've always liked Windows because it's affordable and dedicated to
>consistency of user interface.
I find DOS and Windoze to be piles of something whose mention
would seriously agitate Senator Exon. The Windoze File "Manager" is so
clumsy and cruddy (can't rename files by editing their names in place, to
name just *ONE* example) that I often use straight DOS or that
aforementioned PeeCee-disk file utility. MDI content-free root windows?
Yecch. And my only comment about DOS is what can make 7 out of 8 megabytes
of a computer's memory totally inaccessible without special software?
... Macintosh may in some ways have better ease-of-use than NT.
I did see a copy of WindozeNT running -- it had the same cruddy
Windoze3.x interface. And by the time that WindozeNT 4.0 comes out, Apple
will be a *lot* further along with Copland.
--
Loren Petrich Happiness is a fast Macintosh
pet...@netcom.com And a fast train
My home page: ftp://ftp.netcom.com/pub/pe/petrich/home.html
Or ftp to ftp.netcom.com and go to /pub/pe/petrich
Liang-Shing Ng
>I am becoming a professional Web developer. I have to decide what
>tools to learn and what skills to develop. I'm pretty much at the
>start of my career, and I want to learn systems that will be relevant
>in the future.
In the future, the Web tools will be very easy to use, on Windows or
Unix, that people like you will be out of job, or there are just too
many of you, just like the job title 'programmers'. And the real artists
rule. Just like the advertising industry.
Web developer will be graded on the particular skill or field, like
financial Web writer, or political Web writer, or news Web writer. Not
on the basis of Unix or Windows.
Just like programmers are graded on the basis of field, like dbms,
scientific, os, hardware driver etc. Not the kind of language they can
use.
Why bother the debate about OS?
LSN
Ravi Krishna Swamy
paul chemmanoor <ica...@access.digex.net> wrote:
> 1.Linux
> (minus)
> In simple terms it is a nightmare for a non-cs person
> to setup. Multitasking theoretically great but in practice
> has problems with Netscape [ yeah I know this is probably
> Netscape's fault.. but ... from a users perspective
> it is a big problem ]
??? I never had any problem with Linux, Netscape, and multitasking.
I used them all for 9 months and it always seemed quite smooth to
me. Netscape transferred large files just as quick whether it was
in the foreground or background.
Ravi
--
Ravi K. Swamy http://www4.ncsu.edu/~rkswamy/www/
rks...@eos.ncsu.edu ro...@genom.com
David LeBlanc
>In article <3131C0...@access.digex.net>,
>paul chemmanoor <ica...@access.digex.net> wrote:
>> 1.Linux
>> (minus)
>> In simple terms it is a nightmare for a non-cs person
>> to setup. Multitasking theoretically great but in practice
>> has problems with Netscape [ yeah I know this is probably
>> Netscape's fault.. but ... from a users perspective
>> it is a big problem ]
>??? I never had any problem with Linux, Netscape, and multitasking.
Even on my 20MB machine, Netscape is a huge, bloated pig, and is quite
capable of bogging everything down, esp. if a Java applet loads. It
chews up NT almost as badly. Maybe the 1.2 version didn'd do this - I
don't know.
David LeBlanc
dleb...@mindspring.com
Happily running NT and Linux
Arun Gupta
> [snip]
> 1.Linux
>
> [snipped the plusses]
> (minus)
> In simple terms it is a nightmare for a non-cs person
> to setup. Multitasking theoretically great but in practice
Shouldn't Linux for the PowerMac make it very easy for a non-cs
person to setup linux (e.g., no weird hardware configs. to worry
about ? ) IMO, Linux + PowerMac can bring Linux to the masses.
-arun gupta
Bob Madden
>>>>Netscape has already in great detail elaborated on the cost differential
>>>>between Netscape's server solution and M$' own bullshit "solution."
>>>Yes, I read that. Very interesting. But somehow unconvincing, or
>>>beside the real point.
>>You need to provide some justification for this conclusion. Just your
>>saying it, does not necessarily make it so.
>The numbers are valid, so far as they go. But I don't think that
>calculating the total cost of server hardware and software is that
>important. There are maintenance costs and compatibility issues. A
>lot of companies that are used to buying all IBM equipment would like
>to buy all Windows equipment, rather than having Unix servers and
>Windows on the desktop. Currently they are refraining from buying NT
>servers, but as soon as NT gets a little better, there could very well
>be a stampede towards NT.
The same case could easily be made for an ALL OS/2 solution. In fact,
the case could be made even better. Why have even two platforms such
as NT, and 95, when various OS/2 implementations can do it ALL?
>One such crucial change for this to happen
>is bringing the Win95 users closer to NT. I'd like to see Microsoft's
>latest OS strategy (it's hard to keep track of). The Win3x/Win95/NT
>split is the only thing holding MS back from really taking off.
There has never been any mystery to Microsoft's overall strategy.
Anyone who's been in the computer industry for any length of time
should be intimately familiar with it.
What's holding them back is their inferiority. About the only thing
that Microsoft has succeeded at recently is preloads, and even they
aren't what was anticipated. BillG just recently announced that his
great, and benevolent company was going to donate millions of dollars
worth of Win95 to some worthy charity. Whatcha wanna bet they're
unsold upgrade editions. He's just getting some PR mileage out of
product that would otherwise be occupying a landfill somewhere.
>>>>To completely implement the Netscape version an IS shop has to invest about
>>>>$1,700.00 more or less, to do with M$' "free" solution which BTW only runs on
>>>>NT Server, the projected costs are in the vicinity of $6,000.00
>>>That's simple logic. I get the impression that large corporations are
>>>yearning to move to NT, and MS's giveaway of the server software is
>>>the excuse they will use to choose NT servers rather than Unix servers
>>>running Netscape.
>>Maybe Microsoft finally woke-up, and started charging what the
>>software was actually worth.
>I thought IIS was pretty much brand new.
I've learned over the years that when someone holds out a hand
offering me a FREEBIE, to always ask what's in that hand they're
hiding behind their back.
>>> It's not logical, it's twisted motives. Many
>>>corporations *want* a monolithic Microsoft solution. This is the buzz
>>>I'm picking up from reading various magazines and online news.
>>These same "various magazines" have been declaring OS/2's demise for
>>over 3 years, that I know of, and it still hasn't happened. Could it
>>be that they are wrong about a few other things as well? Don't just
>>take everything you read as "Gospel".
>I've always thought the columnists were biased *towards* OS/2.
How can you possibly read their proclaiming "OS/2 is Dead" at every
opportunity as biased "towards" OS/2?
>I sure
>don't take the magazines as gospel. When Win3.0 came out, I instantly
>knew it would take off like mad, but the first articles underestimated
>the rapidity of its acceptance.
It was still a piece of JUNK.
>The worst problem in the desktop OS
>industry at the time was poor interoperability. Windows provided a
>badly needed common environment to bring together all the PC software
>and hardware. I was more intensely aware of this need than the
>magazines.
Hurray for your insight. However, my recollection is of a Windows
that was pre-loaded everywhere, but only rarely used. It was the poor
unfortunate developers who incorrectly perceived Windows as some kind
of World-Beater. The vast majority of it's users that I knew were
loading some kind of DOS-based menuing system, and loading Windows
only when it was needed for a specific program.
>I have often spotted a trend before the columnists, or violently
>disagreed with them. There are still a few lingering articles by the
>Internet contrarians -- "conservative analysts" who think the Internet
>is a fad because it's (CURRENTLY) insecure and filled with 32 year old
>white computer geeks. And articles that prove that the Internet
>won't happen, because marketing analysis shows that only a certain
>type of person is (CURRENTLY) on the Internet.
>My predictions about the likely explosion of NT are based on what I
>read in the tabloids, but I read critically.
If you say so...
>>>I'm arguing from market perception, rather than from plain economics
>>>or technical points. I've seen several references to NTs
>>>"familiarity" and "ease of use" over Unix. This translates to lower
>>>maintenance costs.
Someone who is maintaining a "Server" is going to need a certain
amount of technical expertise to be a successful administrator, and no
amount of familiar "GUI" will help with this. Your conclusions as
regards "familiarity", and "ease of use" might hold water when applied
to client machines, but NOT to servers.
>>That is generally what Microsoft does quite well, since it is so
>>difficult to argue in favor of their products on technical grounds
>>alone.
>>> Windows is positioned as everyman's
>>>OS; it is in no way elite. It is humble and "good enough". It is
>>>very Bob.
>><GAG!><CHOKE!>
>*Because* of Microsoft's cutthroat strategy and marketing skill, they
>are very well positioned to promote NT even if it is in some ways
>technically inferior to Unix. Marketing is more important than
>technology, to promote a product. The Bob attitude is disgusting, and
>successful.
What has been successful? NT hasn't been. Bob hasn't been. Win95 is
only successful because of pre-loads. Take away Win95 pre-loads, and
where does that leave Microsoft? All their much touted marketing has
in reality failed to convince anyone, other than OEM's.
>Unix is good at one thing only: networking. Microsoft is good at
>marketing, usability, pricing, and incremental evolution, and is
>getting better with networking.
Maybe with small to medium sized LANs, but from most reports I've seen
WinNT continues to fail rather misserably on larger projects.
>>>There is a very large chance that Microsoft is about to become a
>>>software monopoly far more than they are now, very soon. They've
>>>passed a critical threshold of mindshare.
>>I must not have been paying attention again... Just when did this
>>happen? And WHY didn't somebody TELL ME?
>It first happened when Windows 3.0 came out. It then happened during
>the Win95 campaign, followed quickly by the Internet announcements and
>demonstrations that showed the Microsoft was definitely committed to
>the Internet. The current cover of _Internet World_ is "Microsoft
>Declares War", showing general Bill commanding his tanks flying the
>Windows flag. Windows 95 and Microsoft's Internet strategy are making
>headlines everywhere, which have been telling you about the critical
>threshold of mindshare that Microsoft surpassed between August '95 and
>February '96.
I know all about the "Most Favored Nation" type of treatment Microsoft
gets from the industry press. However, I am still not impressed, and
I am not convinced that any sort of threshold has been crossed.
Well... Maybe my BS threshold. God knows... I sometimes think that if
I see even one more Microsoft Marketing Brochure pretending to be some
type of honest article, I'll PUKE.
>>>I propose that all future software sales will come from Microsoft, as
>>>the nontechnical 90% of the world suddenly starts using computers this
>>>year because of the Web. It's going to happen instantaneously. This
>>>is a prediction about the most major discontinuous change.
>><ROFL> Those are some damned good drugs you're doing!
>You think it will be business as usual in the computer industry, with
>growth rates the same as they have been? The Web changes everything.
I think you're basic missunderstanding lies with your perception of
the "Web" as the underlying cause of recent changes in the industry,
when it is NOT. What is the "driving force" is the HUGE technological
leaps being made wrt "Hardware", without which, the "Web" would not be
feasible. What is really interesting to watch nowadays is the race to
be the provider of this increased bandwidth. At this point, the Cable
companies seem to have a healthy lead with their "Cable-Modem"
technology...
I dunno... We'll just have to see.
>>>Microsoft will become a software monopoly for all the people who are
>>>about to explode onto the Web. The Web will make the OS war
>>>unimportant at the same time as Microsoft's OS captures the billions
>>>of people, the vast majority of people who have never given a thought
>>>about NeXT vs Amiga vs OS/2 vs Windows vs Mac vs Linux. The vast
>>>majority of people are completely unaware of the OS war, and they are
>>>about to start surfing the Web. If ever they have to pick an OS, they
>>>will choose Microsoft, because it has the marketing mindshare and the
>>>If most people end up using cheap Internet terminals, they might find
>>>themselves using Sun hardware (Java chips) or accessing Web pages that
>>>happen to reside on Unix servers. Or maybe mainframes or NT servers.
>>>I sure haven't read much about OS/2 Web servers yet.
I think this is where the one major difference lies... The level of
technical sophistication of someone looking for "Server" solutions is
a bit greater than your average user, and they are also better
informed.
>>>This year, there will be a huge explosion as the majority of people
>>>all over the world, rather than a very small minority, rush onto the
>>>Web. The market for computing power -- whether desktop computers,
>>>servers, or terminals -- is going to grow more this year than it has
>>>in all past years, combined.
>>You keep proceeding from what are at best... "faulty assumptions".
>>Before the "rush" you are talking about, can really happen, the
>>low-cost, high-bandwidth connections will need to become reality.
>All the pieces are about to fall into place to create discontinuous
>change.
The computer industry is synonymous with "discontinuous change".
>A month ago, people doubted that agreements could be reached
>soon. But the telecomm bill suddenly passed, and Visa and Master Card
>suddenly came to an agreement. The race is on, to get on the Web, and
>no one has time for deadlocks. The Web is now recognized as the
>digital Manifest Destiny, even though computer old-timers might be
>inclined to dismiss the hype as unwarranted and fail to see the
>explosion coming so soon. All the projections I've seen are
>reasonable and conservative. I predict that universal Web access will
>happen much sooner than people expect. The problems will be solved
>quickly, now that everyone has simply identified the shared goal and
>grasped the possibility.
What you fail to realize is that the marketplace has often told
businesses what they can do with their "manifest destiny"...
>@Home http://www.home.net/
>@Home is about to make 10 megabits per second available in Sunnyvale
>for about $40 a month. That's 350 times faster than 28.8 kbaud.
I have investigated @Home's Website, and have found it very un-clear
as to the specifics of how this new service will be delivered to the
end-user. It seems they are going to GREAT lengths to control the
type of access I will get as a user, and I made it a point to E-mail
them in a very clear manner... that what I want is a High-Speed, Low
Cost connection to the Internet. As far as controlling the method I
must use to receive their signal, or the content thereof, I am
strongly opposed. It's fine if they want to lease me their
proprietary cable-modem, another kind of cable-box, but I do not want
them mucking about with anything else.
>Secure online transactions are about to become available.
Secure, or not... I am still likely to do most of my purchasing from
local stores.
>>>What hardware and OSs will predominate a year from now, after this
>>>remarkable event that changes the whole scope of computing? Watch out
>>>-- we're all about to be trampled by discontinuous change.
>>So... Now you've pontificated for several paragraphs, and all you can
>>come up with is that things are going to change... a LOT?
>I've predicted that NT servers will grow much faster than Unix
>servers, but that the OS war will become less important.
Of course NT servers will show a better "growth rate". Based on sales
to date, just about anything would be an improvement.
>New operating environments will have a chance, because they only have to
>make themselves compatible with the Web, Java, and HTML or SGML. The
>idea of a "platform" as an isolated island will be replaced by the
>idea of flavors of environment, all interoperable. And I have asked,
>what predictions do you have, for the impact of the Web on the OS war?
>This investigation is driven by my need to decide which Web
>development skills and environments to learn.
As always, this depends on your own abilities. Assuming the Web takes
off like it should, I believe there will be a strong demand for Web
authors with artistic talent. To a very large extent "artistic
talent" just can't be taught. Programmers will still be in demand to
write underlying sub-routines that perform a myriad of functions. A
good example that we are all aware of is a "Search" engine, but there
are many more examples of things that will still need to be "coded".
One other area that is, and will be experiencing growth (believe it,
or not) is Assembly language programming. Due to dwindling demand,
they have become pretty much a dying breed, and there will always be a
demand for people who really understand how everything works.
It is not that the need for them will increase greatly... It wont, but
the supply has been dwindling, and at some point the supply of
available Assembly language programmers will fall below what is
needed, and they will once again be in demand.
Try to steer yourself towards learning about underlying technologies
such as speech-recognition. Although nobody has a real clear idea
just what the future will hold in terms of what the marketplace will
accept, these underlying technologies will be absolutely essential.
IOW... You may not know just who's design will be accepted for
building a structure, but if you're in the business of manufacturing
the girders required to make it all happen, it won't matter who gets
the contract, they'll have to come to you to get the building
materials.
Have a GREAT Day
Bob M.
Chris Worth
>>Now this is just sad. It's obvious that you put a lot of time and effort
>>into composing this post (137 lines!), but you'll never be able to
>>successfully troll a newsgroup until you learn how to properly forge a
>>newsgroup header.
>
>>-Erik
>
>
>and Internet terminals on the OS war?
>
>Do you think it is likely that NT will very rapidly gain acceptance
>over Unix for corporate intranets? Will the Unix-to-NT Web server
>ratio suddenly flip-flop?
>
That really depends, how many companies have UNIX boxes already, chances are
the won't simply punt UNIX and got to NT. why would they? NT might pick up
some sales for the first timers, particularly the FUD-susceptible. Afterall WarpServer
is to be released soon, and it handles bigger loads than NT. as to a UNIX-NT flip
flop. NOT a chance. if nothing else because there are thousands upon thousands
of UNIX sites( most all the EDU, and many COM) locations are already UNIX, and if
it works, why change?
chris
Thomas G. McWilliams
: -- Michael Hoffman, EE, technical writer, Web developer,
: techno-trendmonger
Web developer? What a goofy passtime. The Web is the the CB craze
for the mid 1990s. It is a fad like pet rocks, platform shoes and
CB radio. 10-4 good buddy. Got a smokey eastbound.
Most folks have no need for the Web at all. Folks who normally spend
their time watching "The Simpsons" re-runs from inside their
double-wide have momentarily fired up the Packard-Bell during a
commercial break. Randomly clicking and looking at pretty pictures has
about as much staying power as Rubik's cube.
Coming soon to QVC: "You too can be a Web Developer, no experience
needed". It airs in between the "Gold by the inch" and "Psychic
Friends Network" infomercials.
Wayne J. Hyde
dleb...@mindspring.com (David LeBlanc) writes:
> rks...@unity.ncsu.edu (Ravi Krishna Swamy) wrote:
>> In article <3131C0...@access.digex.net>, paul chemmanoor
>> <ica...@access.digex.net> wrote:
>>> 1.Linux (minus) In simple terms it is a nightmare for a non-cs
>>> person to setup. Multitasking theoretically great but in practice
>>> has problems with Netscape [ yeah I know this is probably
>>> Netscape's fault.. but ... from a users perspective it is a big
>>> problem ]
>> ??? I never had any problem with Linux, Netscape, and
>> multitasking.
> Even on my 20MB machine, Netscape is a huge, bloated pig, and is
> quite capable of bogging everything down, esp. if a Java applet
> loads. It chews up NT almost as badly. Maybe the 1.2 version
> didn'd do this - I don't know.
Netscape can bring both of our 40MB and 24MB NT machines to their
knees. Netscape will start allocating gobs of memory if you don't
kill it every so often and restart. I even disabled Java and it still
goes RAM crazy. The AIX version of Netscape is a fat pig too.
--
Wayne Hyde | System Administrator | http://www.cis.ufl.edu/~wjh
218 Newins-Ziegler Hall | Fla Cooperative Fish & | Off the keyboard, through
Gainesville, FL 32611 | Wildlife Research Unit | the router, over the
(904) 392-1861 | w...@cis.ufl.edu | bridge, nothing but net!
Geoffrey Alexander
>|> >>I propose that all future software sales will come from Microsoft, as
>|> >>the nontechnical 90% of the world suddenly starts using computers this
>|> >>year because of the Web. It's going to happen instantaneously. This
>|> >>is a prediction about the most major discontinuous change.
>|>
>|> ><ROFL> Those are some damned good drugs you're doing!
>|>
>|> You think it will be business as usual in the computer industry, with
>|> growth rates the same as they have been? The Web changes everything.
>Actually, yes I believe it will be business as usual. The web, while
>a new and interesting factor, won't be that improtant. Most of the stuff
>is and will be done without the net. The web is a new part of the
>computing scene, but nothing drastic.
By the time any number of the great unwashed make it to the Web, most of
what they'll find without digging will be commercially oriented -- and aside
from the few mass transaction sites where shoppers will congregate (as though
at a shopping mall -- the metaphor is apt) few sites will attract real interest
-- few people groove on television commercials that don't move -- and the happy
millions will, I think, defer from configuring QT just to watch a Lexis commercial.
Word will get out on the real attractions of the Web, and it's real debits in the
way of what's there -- and the market for Web products will deflate. We know what
the web is good for -- but none of those things will translate to the kind of
cultural impact some are talking about. How many people turn the sound up on their
TV when commercials come on? How crowded is the reference desk at your local library?
"The Internet is an elite organization. Most of the people in the world have
never even made a phone call." -- Noam Chomsky
--
geoffrey alexander : http://www.netins.net/showcase/sahaja
If you can talk brilliantly enough about a problem, it can
create the consoling illusion that it has been mastered...
stanley kubrick
Orc
Chris Worth <cawo...@roentgen.bcc.louisville.edu> wrote:
>In <4gomdr$9...@shellx.best.com>, mic...@cybtrans.com (Michael Hoffman) writes:
>>>Now this is just sad. It's obvious that you put a lot of time and effort
>>>into composing this post (137 lines!), but you'll never be able to
>>>successfully troll a newsgroup until you learn how to properly forge a
>>>newsgroup header.
>>
>>>-Erik
>>
>>
>>So what do you think about the impact of the cross-platform Web, Java,
>>and Internet terminals on the OS war?
>>
>>Do you think it is likely that NT will very rapidly gain acceptance
>>over Unix for corporate intranets? Will the Unix-to-NT Web server
>>ratio suddenly flip-flop?
>>
>
>That really depends, how many companies have UNIX boxes already, chances are
>the won't simply punt UNIX and got to NT. why would they?
Look at the Apple 2 or the collection of Z-80 boxes that used to
be sold. Not too many of them around anymore, are they? It shows
what happens when a disorganized collection of vendors gets hit by
a computer-company invented Standard.
Fortunately Unix runs (or walks, in the case of Solaris) on the
same hardware that NT runs on. This makes it more difficult to argue
that NT is better because it adheres to invented Standards, while Unix
Does Not.
____
david parsons \bi/ But you should still worry.
\/
Chad Irby
(Some groups trimmed to follow the Rule of Seven...) *
t...@netcom.com (Thomas G. McWilliams) wrote:
> Michael Hoffman (mic...@cybtrans.com) wrote:
> : -- Michael Hoffman, EE, technical writer, Web developer,
> : techno-trendmonger
>
> Web developer? What a goofy passtime. The Web is the the CB craze
> for the mid 1990s. It is a fad like pet rocks, platform shoes and
> CB radio. 10-4 good buddy. Got a smokey eastbound.
Better than that- being a "Web Developer" right now is right up there with
being a professinal typesetter in 1984. There's a lot of folks making
really ridiculous amounts of money making up plain-vanilla Web pages for
clients with too much money and not enough brains.
Meanwhile, there are a handful of Web authoring tools on the way that are
starting to make basic HTML-crafting as simple as PageMaker made basic
desktop publishing. PageMill is just the start...
*(The Rule of Seven is: don't crosspost to seven or more newsgroups.)
--
ci...@magicnet.net
I gave up on the fantasy and am looking
for a really good reality...
Ravi Krishna Swamy
David LeBlanc <dleb...@mindspring.com> wrote:
>rks...@unity.ncsu.edu (Ravi Krishna Swamy) wrote:
>
>>In article <3131C0...@access.digex.net>,
>>paul chemmanoor <ica...@access.digex.net> wrote:
>
>
>>> 1.Linux
>>> In simple terms it is a nightmare for a non-cs person
>>> to setup. Multitasking theoretically great but in practice
>>> has problems with Netscape [ yeah I know this is probably
>>> Netscape's fault.. but ... from a users perspective
>>> it is a big problem ]
>
>
>Even on my 20MB machine, Netscape is a huge, bloated pig, and is quite
>capable of bogging everything down, esp. if a Java applet loads. It
>chews up NT almost as badly. Maybe the 1.2 version didn'd do this - I
>don't know.
I kept it open for weeks at a time on my 486dx33 w/ 16 mb ram.
It was version 1.1 and I'd download 5 mb files with it while
editing and compiling all the time. I could never notice any
drop in its transfer rate except when using another network app.
I got the very first 2.0 beta for Linux but the "frames" feature
annoyed me so much that I'm still using XMosiac 2.7 and Netscape 1.12
It would be great if all that "functionality" were not in one huge
~4 mb app. I've already got upteen mailreaders to choose from. I
don't need another in my web browser.
Loren Petrich
Michael Hoffman <mic...@pdr-is.com> wrote:
>Pricing strategy is crucial. Microsoft has been very careful to never
>ask for too much money all at once. The NeXT and its pricey
>applications required a large lump sum. This demand for a financial
>sacrifice all at once prohibited OS/2, Mac, Unix, and NeXT from taking
>off.
UNIX has been around for years, and there are free UNIXes
available. Ask the Linux groupies some time :-)
And I really don't think that M$ deserves to be rewarded for
phony accounting.
I just hope that its competition can stop being so damn *inept*
-- given that CHRP/PPCP will be an open hardware standard, I would not be
surprised if at least some of the licensees of it will be certified
hypemeisters.
... The MS solution is not cheap when you total the costs, but the
>ability to incrementally upgrade PCs using building blocks from a vast
>arena of suppliers has enabled people to make baby steps toward
>ever-better versions of Windows. ...
So that's why we must suck up to the Boy Blunder of Redmond? :-)
Jay Urbanski
:
:Wow! You really believe what you said don't you? How adding a poor GUI
:to NT can make it more attractive as a server is beyond me. You honestly
:think unix boxes are only used for webserving and being something to run
:on a PC. There's a lot more that unix is being used for than just webstuff.
:Banking transactions, instrumentation and test, telecommunications,
:satellite systems, imaging, special effects, traffic control, warfare
:systems, weather mapping to name a few.
100% correct. Web serving is one of the few areas I think Unix is in
danger of being crowded aside a bit by NT, because it's something NT does
well and it's a no-brainer to set up. However web serving is a pretty
trivial task, and for the huge installed base of Unix users out there doing
complex work, they aren't going to be switching to NT any time soon.
:NT has a lot of bases to cover
:to even come close to knocking on unix's door. Plus it needs to scale
:better and be more open.
Yup.
:I guess after we (hopefully) open your eyes to the scale and folly of
:what you propose with the timeframe you've stated, you will then try
:to say NT will replace mainframes by next fall.
Mainframes are no more in danger of being replaced by NT than they are in
danger of being replaced by Unix servers. :)
Jay Urbanski
MCSE
Certified Solaris Administrator
Paranet
Sangria
> And this argument ignores all the WWW servers running the MacOS.
Until they figure out a way to increase it's load capabilities, it's
not going to be all that useful outside of a relatively small site or
one that's willing to invest in quite a bit of hardware to compensate.
>Yes, that fruit company's OS. I suggest that you check it out, because it
>is absolutely uncrackable (there was a competition to do just that, and
>it got no entries), and that will be vital for e-commerce.
It's absolutely uncrackable since httpd do nothing more than listen
on port 80 and pump bits through the connection.
Take any Unix, NT, hell even DOS based system with a single port connection
opened on port 80 and see if anyone can "crack" it. Even with CGI support,
if I get to put the CGI on the machine, I can garantee you, there is no
way you can "crack" it.
Let's see what happens to that MacOS based system when it's running a
telnetd, sendmail, rlogind, etc...
The *vast* majority of the CERT warnings issued were revolving around
sendmail, not a web server.
>>I am facing some hard decisions now. Currently, the recommended way
>>to write CGI scripts is to code in Perl on Unix. But I think there
>>will be greater demand for NT-based CGI programming, over the next few
>>years.
>
> With the MacOS, one can write CGI scripts in AppleScript, which I
>find to be a very elegant programming language.
As far as CGI is concerned, there really is no "default" language.
You can pretty much use whatever the system will suport as long as
you follow the interface guidelines.
> Copland will have the advantage over NT of less resource
>requirements
Do you know this for a fact?
NT 4.0 was tested to run "decent" in 8MB. And pre-liminary tests
of my own show that it's video subsystem is considerably faster
than it was in 3.51. While beta 1 still needs some more work, it's
getting there...
Copland's initial release will have two strikes against it:
It's a 1.0 release.
And it's primarily going to run on a RISC platform.
I seriously doubt Copland will have much of an advantage, if any
over NT in terms of resource usage.
>...Unless Apple commits some serious design blunder, which
>is unlikely, it will be at least as bulletproof as NT,
I wouldn't count on Apple not to have problems. Remember, for all
intent and purposes, this is a 1.0 release of a new OS. I expect
Apple's first release to have problems.
As for as bulletproof as NT, sorry. NT doesn't suffer from having
the interface die on it when an application decides to take breather.
And unless Apple did something different with MacOS's file system,
it won't be close to NT's fault tolerance. And it most certainly
will not approach the feature sets of NT.
>and be as uncrackable as the current MacOS (System 7.5).
That isn't saying much.
MacOS, running the full spectrum of network services would find itself
on the losing end of security in a real hurry... MacOS has no concept
of process ownership. This by definition would imply there is no
"security" model for MacOS.
>>I'm facing a similar hard choice between Java and Visual Basic Script.
>
> I'd go with Java, since Sun is more committed to openness than
>M$. Also, Java plug-ins are now more available than Visual Basic ones.
>Even M$ has given in and decided to go along with Java.
Correct.
However, learning both would be better.
Also, I wouldn't count on the current crop of Java-plugins...suckers
are buggy as hell. They exhibit some phenonemal memory leaks under
NT...
> More seriously, M$ has remarkable gifts of self-hype; it has the
>ability to make itself seem like it invented practically *everything*.
Not quite.
MS's isn't the best in this field.
MS's biggest strength is their tenacity.
Often their first release of a product bites. This has been the case
with pretty much every product they've released. However, what MS does
better than anyone else is keep pushing the product, keep improving the
product, until it becomes arguably one of the better products available.
>>... Macintosh may in some ways have better ease-of-use than NT.
>
> I did see a copy of WindozeNT running -- it had the same cruddy
>Windoze3.x interface. And by the time that WindozeNT 4.0 comes out, Apple
>will be a *lot* further along with Copland.
How close to release do you think Copland will be by June/July of this year?
Beta 1 is already in the hands of MSDN members, and various "NT Champs"
subscribers. Beta 2 will be in March (just a few days away) and the release
date is around June/July of this year.
Last I checked Copland isn't expected to ship until end of this year
(which is pretty much what I figured). And recall, this is NT's fourth
version while it's Copland's first. Expect to see Copland "teething"
pains with this release.
If anything, I expect NT's rate of growth to far surpass that of MacOS,
Unix, and even OS/2 in 1996.
-- Sang.
*************************************************************
* Sang K. Choe san...@inlink.com *
* http://www.inlink.com/~sangria/index.html *
* Or http://sangria.inlink.com/index.html *
* finger: sa...@sangria.inlink.com *
*************************************************************
JeroenHoppenbrouwers
> 100% correct. Web serving is one of the few areas I think Unix is in
> danger of being crowded aside a bit by NT, because it's something NT does
> well and it's a no-brainer to set up. However web serving is a pretty
> trivial task, and for the huge installed base of Unix users out there doing
> complex work, they aren't going to be switching to NT any time soon.
What happens here all over again is the misconception that *all* serving
possible is either file/print serving or database/Web serving. Just like
with the PC adoption: before the PC, 100% of computing was quite heavy
because of the price of computers. Then the PC came and word processing
became feasible. Within a few years, 90% of the installed computers were
running only word processing--and everybody thought that the PC had kicked
the big computers off the Earth.
Nowadays, there is a growing base of simple servers that serve only Web
or files. These simple servers can even run on DOS if somebody took the time
to write a proper program (that, of course, took the whole machine, but that
was the purpose of the server anyway). The rapid growth of the simple machines
that do Web leads to the misconception that *all* servers can/should run NT
because NT serves the Web as good as UNIX.
If the only thing you want to do is word processing, get Win 3.1. Many people
do.
If the only thing you want to do is Web serving, you can get NT and hope you
understand it in time. You can also get that dedicated DOS program, which
would be even simpler to set up.
But NT is not going to replace *all* servers. Simply because it cannot.
But nobody is interested in the servers in the cellars that keep the Earth
running. Visibility is all people want, and Windows, being a graphical
front-end program, gets all of it.
--
Jeroen Hoppenbrouwers, research engineer at | Stop connecting computers;
Infolab, Tilburg University, The Netherlands | start connecting people!
http://infolabwww.kub.nl:2080/infolab/people/hoppie
Robert Brockway
: I think that technical arguments can miss the point. Superiority of a
: technology does not correspond with dominance of that technology.
True. The doninance of ms-dos is proof of that.
: I am confident that NT will quickly become the dominant operating
: system for the next few years, whether or not it is technologically
: superior. I think it's a better career move to align myself with
: NT-based Web development, than Unix-based Web development.
Why not use both?
: I am facing some hard decisions now. Currently, the recommended way
: to write CGI scripts is to code in Perl on Unix. But I think there
: will be greater demand for NT-based CGI programming, over the next few
: years.
Actually Unix is more popular than ever before in the busines world.
: I want to learn the most appropriate Microsoft-oriented tools and
: environments for Web development. I don't want to be ignorant of
: Unix, Perl, and Java, but I find I have to resist jumping into the
: Unix/Perl camp, which is currently dominant and tries to pull me in.
: Unix is yesterday's revolution and I'd rather minimize my involvement
: with it. NT is far from perfect, but the NT environment and mentality
: is far more up-to-date than Unix, which has its roots in the early
: 70s. The Unix mentality is obsolete, never mind the technical
: details.
Hardly. look at Unix of the 70s, now look at Unix of today. There have
been many improvments.
: While Unix will live on, it will be killed in the sense of being
: overwhelmed by NT, as everyone in the world becomes involved in
: computers. The 90% of people who have not yet touched a computer will
: not buy into Unix; they will buy into NT.
Unlikely. They will use win95 :-)
NT is such resource a big of an operating system that i would not consider
it for my desktop. I don't have enough money to get a good machine. Most
families would not (or can not) spend the required cash to get a good NT
machine. They will therefore not use it (or use it for a while on
insufficient hardware and change).
: Internet terminals will shield the masses from dirtying their hands
: with either NT or Unix. These terminals will be served by Unix
: servers at first, but soon, the entire business world will switch to
: NT. Technical issues aside, Unix has no future; Microsoft is rapidly
: becoming the OS monopoly. OS/2 is completely out of the picture.
The OS/2 people may disagree with you here :-)
: Technical people might hate this trend, but based on reading the
: technology and business magazines and reports, NT and Microsoft are
: unstoppable. Whether or not this is desirable, it's happening, and
: I'm going along with it.
This is a circular argument. Everyone else is using Cod liver Oil, so i
will too.
: Microsoft is the right balance of price, consistency, and popularity.
You must be kidding....I have suspect this is a troll...now i'm sure :-)
: I've always hated Apple because they were overpriced when I bought my
: first computer in 1988.
True.
: I've always hated Unix because it's essentially an antiquated, chaotic
: character-mode environment with no serious GUI and poor compatibility
: among the zillion flavors.
Are you serious? Microsoft has never had a network transparent windowing,
Unix has had one for years (i can't remember just when X was developed, but
it was early 80s i think). Have you heard of POSIX? I guess not. It is
the Unix standard.
: I've always hated OS/2 because it's made by IBM.
Subjective. If you want us to take you seriously you must provide reasons.
: I've always liked Windows because it's affordable and dedicated to
: consistency of user interface.
How much is the NT you were telling us about before? US$400 isn't it? or
more perhaps. Hardly what a family wants to pay for an OS. Linux is free :-)
: Windows is technically good *enough*.
I disagree. One opinion is as valuable as another :-)
: In choosing and advocating an operating environment, find the best
: *balance* of:
Cheers,
-Robert
--Robert Brockway, email: ec53...@student.uq.edu.au
WWW: http://student.uq.edu.au/~ec531667
"Do not sell IP routers to hostile governments."
Harley Hahn, 1993.
Al Kozakiewicz
says...
>Practically all current concepts of OS design were established in
>late 60s-early 70s. NT has it's roots in VMS, Unix, CP/M and Xerox Alto.
>All three of them were designed in early 70s; main design concepts of VMS
>are probably even older. Who knows release date of RSX-11?
I first used RSX-11 in 1974. The first PDP-11 (PDP-11/20) was released
around 1970, but I cannot remember which was the first non-paper tape
operating system available for it (RSTS. RSX or RT).
VMS did not borrow much from RSX except at the most superficial level.
The first versions of VMS supported the PDP/RSX compatibility mode built
into the earliest VAX machines, but that was not an especially well
integrated part of VMS. Outside of the fact that both systems were
event-driven with a round-robin scheduler, and that VMS had some API calls
in common with RSX (and probably in common with every event-driven OS in
existence ;^)), they really didn't share much architecturally. After all,
one machine was a 32-bit virtual memory CPU, the other a 16 bit memory
mapped processor.
That being said, VMS was *the* heir to most of the RSX-11 customer base.
VMS dates to about 1977.
Regards,
Al
--
________________________________________________________________________
Al Kozakiewicz, President phone: 518.452.9062
Hourglass Systems, Inc.
Microsoft Solution Provider
________________________________________________________________________
Norman D. Papernick
will kill Unix in a .. by William Un...@physics.ub
> Uh, but most of that 90% will not be running servers. They will be using
> ISP's to provide them with the path onto the internet. I have never been
> able to tell what machine it was that served me my web pages. Have you?
> So what if 90% use Win95 or even Dos? (and by the way that 90% sure
> won't be buying NT either).
Just to poke at this one: the .HTM has always been a give-a-way. And if
there are lots of spaces in the file names, then there is a good chance
it is mac, but not certian.
<note: the To: line has been pruned.> --norm
---
Norman Papernick | Information is browser independent. Fluff is not.
no...@andrew.cmu.edu | HTTP is the best and worst thing to happen to the
CMU Computer Science | internet. <url:http://www.cs.cmu.edu/~norm>
Michael Hoffman
Thanks for all constructive insights -- I'm a quarter of the way
through reading the latest batch carefully. This is cross-posted to
almost all the os.advocacy newsgroups and the www.advocacy group.
It would take all day to respond point-by-point. I could say a lot
more than the following points, but I'll leave it at that. I hope I'm
addressing issues that interest a lot of people, Web-related issues
that haven't been debated in these newsgroups yet.
Someone points out that you can't charge such high rates for Windows
as you can for the technologically more refined niches such as NeXT,
but that misses my main proposition, which I now restate as:
_______________
Thesis:
NT Web servers are well-positioned to greatly increase in sheer
quantity over Unix Web servers, if Microsoft improves certain aspects
of NT. But the Web is by far the most powerful development to ever
happen in computing, because it provides high-level universal
connectivity and interoperability that is driving a feedback effect --
a universal gold rush to get wired within the next year. The Web
could rapidly render the entire OS war irrelevant and wipe out the
current concepts of "operating system" and "platform".
_______________
NEW BYTE COVER STORY ON INTERNET TERMINALS
There is a *great* cover article about Internet terminals in the
latest Byte, March 1996, which arrived yesterday.
The Disintegration of the "Platform" concept, enabling old
technologies to be abandoned:
A "platform" is a monolithic island -- a system consisting of an
exclusive OS, user interface or operating environment, and software.
The Web could finally dissolve this concept by providing a single
point of compatibility. As long as your CPU, or OS, or GUI, or
application works with universal standards such as Java, the Web, or
SGML, you can mix and match whichever OS and hardware you want. With
this interoperability, we are no longer shackled by
backward-compatibility! This means that outdated OSs can be surpassed
by radically new OSs, and archaic chip families such as the Intel x86
family can be abandoned at last for *much* faster and cheaper chips.
This could enable the obsolescence of Unix or other OSs that don't
measure up to the hottest OSs in the research labs. Sure, today Unix
might be the most powerful, but it pales compared to OSs in the
research labs. This fits with _The Unix-Haters Handbook_. Unix falls
far short of what it could be. Wouldn't it be great to abandon it and
write something far more up-to-date, from scratch? Standards,
interoperability, and the Web will enable this to happen.
For servers, one OS may dominate, while one clients, another OS might
dominate. Or, due to interoperability of OSs that are connected
through the common point of the Web, many OSs may proliferate,
including OSs far superior to Unix -- OSs that are currently just
fascinating projects in the computer research labs. The same applies
to CPUs.
Raw Unix vs. Unix GUIs
*Good* Unix GUIs such as NextStep and possibly SGI are crucial to Unix
servers capturing the exploding Web server market.
The raw, command-line Unix 'operating environment' can't really be
used by casual users such as travellers in an airport terminal trying
to use the Web to make travel arrangements, do research, log in to
work and the workgroup, make some Web "phone calls", and pass the
time. But GUIS that completely hide the mess that is raw Unix, such
as NextStep, would work. But I doubt Unix is the best choice for the
OS for such Internet terminals.
Unix could improve and capture an exploding market for Web servers.
NT could vastly improve as well.
With certain improvements, NT could "multiply much faster" than Unix
-- I mean both relative and absolute quantities. However, the best
outcome would be for both NT and Unix to be surpassed by a completely
new operating system.
WHY THE WEB WILL GET EVERYONE INVOLVED IN COMPUTERS WITHIN A YEAR
The Web is the driving factor behind a computer explosion happening
this year, because the Web is about *convergence* -- the digital
convergence. The idea may be old, but finally a specific technology
enables the great digital convergence to happen. Sure, a lot of
people will merely hang out in the shopping malls. The fact that
everyone is wired will creat a giant feedback loop shooting the demand
for computer technology through the roof. When everyone has access,
it makes all the difference in the world. Few engineers grasp this
*feedback* effect -- or they are too busy *acting* on that confident
understanding to write rebuttals to the many technological
contrarians. Many engineers are incredibly skeptical of the
importance of the Web. But all businesses have extremely compelling
reasons to re-orient themselves with the Web right in the middle of
the entire business world. When 100 people have a telephone, it's
insignificant. When *everyone* has a telephone, it's greatly
significant. It's the fax effect -- when everyone has a fax, buying a
fax becomes mandatory.
It is now *mandatory* that every business re-orient itself to take
fullest advantage of the Web. The Web is not moderately important.
It is *absolutely* important.
The ubiquity of the Web, with everyone using it seriously every day,
will have great impact on the OS war, even enabling leapfrogging of
completely new operating systems that only need to be compatible with
the Web. Unix or NT could become obsolete overnight. We can all
switch to our favorite GUI or OS-de-jour, due to interoperability that
is guided around the one requirement: Web compatibility.
So what should I learn to be a Web developer? (My driving question.)
I want to work with content more than with systems integration.
I've adjusted my previous list a little. I'll concentrate more
towards the top of this list:
Java
HTML authoring
VB
Perl for CGI
Windows Web servers and clients
Database front-ends
Unix Web servers and clients
Note that Java, HTML, Perl, and CGI skills are largely cross-platform,
so I don't need to exclusively commit to either Unix or NT. But I
sure don't want to spend a lot of time becoming a Unix guru. I hope
Unix becomes relatively unimportant and is retired. It really rubs me
the wrong way, through and through, (though I could tolerate the NeXT
GUI).
Jon A. Maxwell
] As for as bulletproof as NT, sorry. NT doesn't suffer from having
] the interface die on it when an application decides to take breather.
I was using NT today with the "NewShell" and an error dialog came
up (just a little permission denied thing, nothing of vital
importance). So I clicked on another window --oops, not allowed
sorry BEEP.
I clicked on Start and ran something else --oops, drive D not
accessible (gave me a little box so I could wait).
I opened up find file I think it was, and the screen had "holes"
through the find window. Great. NT really kicks Mac's hinny in
the interface area, uhuh!
When I got to the machine, it told me to hit "control-alt-delete"
to log in! Now /that/ is an excellent interface alright!!
--
thur Mail Address: LordA...@vt.edu or jmax...@vt.edu
n r
a JAMax "Last week I saw a woman flayed, and you will hardly
h o w believe, how much it altered her person for the
tan lle worse." --Jonathan Swift
TOMMY K. HWANG
> The commerical implementation of Linux, Caldera, costs $99. Unixware costs
> about $200. Many two user license Unix costs about $600 to $700, like SCO
> and AIX. Unlimited licenses can be obtained for about $1000 like BSDI. You
> can't get unlimited licenses for NT Server at about $1000. Frankly, Linux
> for free is for a virtually unlimited of users too.
ummm... How did you get Unixware for $200.00? I am not Unixware person..
but price seemed a little low, especially when my SCO Openserver costs over 1500
for 15 licenses (and no developer) and Solaris is over $3000 for unlimited user.
I never liked BSDi or any varients of BSD.
> So where did NeXT come in? NeXT isn't even dead yet. Nor is it's fate
> of NeXT the fate of all Unices. NeXT did not even thrive as well as all the
> other Unices. NeXT is not even on the same category as Windows so why make a
> comparison?
From "Rumors", NeXTStep is in 'hold'. No developments after 4.0, but
support still available if people still buys it. They seems to be moving
towards NT I think.???
David LeBlanc
>Sangria wrote:
>] As for as bulletproof as NT, sorry. NT doesn't suffer from having
>] the interface die on it when an application decides to take breather.
>I was using NT today with the "NewShell" and an error dialog came
>up (just a little permission denied thing, nothing of vital
>importance). So I clicked on another window --oops, not allowed
>sorry BEEP.
Did you notice that the new shell for 3.51 is considered alpha code?
It isn't even close to ready for prime time. IN fact, nearly everyone
has abandoned using it, since the bugs got really mean and nasty.
NT 4.0 beta 1, OTOH, has been running flawlessly so far on my machine.
(knock on wood - having said that, I'm likely doomed).
>When I got to the machine, it told me to hit "control-alt-delete"
>to log in! Now /that/ is an excellent interface alright!!
It is, actually - alt-ctrl-del gives a hardware level interrupt, which
means you can't easily write a trojan log-in screen to steal passwords
with. First time you've used NT?
Shimpei Yamashita
>
>rks...@unity.ncsu.edu (Ravi Krishna Swamy) wrote:
>
>>In article <3131C0...@access.digex.net>,
>>paul chemmanoor <ica...@access.digex.net> wrote:
>
>
>>> 1.Linux
>>> (minus)
>>> In simple terms it is a nightmare for a non-cs person
>>> to setup. Multitasking theoretically great but in practice
>>> has problems with Netscape [ yeah I know this is probably
>>> Netscape's fault.. but ... from a users perspective
>>> it is a big problem ]
>
>>??? I never had any problem with Linux, Netscape, and multitasking.
>
>Even on my 20MB machine, Netscape is a huge, bloated pig, and is quite
>capable of bogging everything down, esp. if a Java applet loads. It
>chews up NT almost as badly. Maybe the 1.2 version didn'd do this - I
>don't know.
20MB? Heh. Netscape w/Java can even slow down the 100MHz HP PA-RISC
workstation I get to play with at work. The darned thing has half a
gig of RAM as well, and Java apparently chews right through
them. Unbelievable.
--
Shimpei Yamashita <http://www-leland.stanford.edu/~shimpei/index.html>
Shimpei Yamashita
>
>paul chemmanoor <ica...@access.digex.net> wrote:
>> [snip]
>> 1.Linux
>>
>> [snipped the plusses]
>> In simple terms it is a nightmare for a non-cs person
>> to setup. Multitasking theoretically great but in practice
>
>person to setup linux (e.g., no weird hardware configs. to worry
>about ? ) IMO, Linux + PowerMac can bring Linux to the masses.
This means, at best, that Linux+PowerMac will be about as easy for a
sysadmin as UNIX on workstations made by Sun, HP or DEC, all of which
make Unices for their own workstations. OTOH, Solaris, HPUX and OSF/1
are hardly OS'es for the masses. Making UNIX easy entails writing
good GUI front ends for admin functions. This is almost completely
orthogonal to the hardware configuration--once you get past the
hardware, MacLinux will be just as much of a pain to use as x86 Linux.
Jon A. Maxwell
] >] the interface die on it when an application decides to take breather.
] >I was using NT today with the "NewShell" and an error dialog came
] >up (just a little permission denied thing, nothing of vital
] >importance). So I clicked on another window --oops, not allowed
] >sorry BEEP.
]
] Did you notice that the new shell for 3.51 is considered alpha
] code? It isn't even close to ready for prime time. IN fact,
] nearly everyone has abandoned using it, since the bugs got
] really mean and nasty. NT 4.0 beta 1, OTOH, has been running
] flawlessly so far on my machine. (knock on wood - having said
] that, I'm likely doomed).
So how do I go back to using the old interface? I have
guest-level privileges, whatever that means.
] >When I got to the machine, it told me to hit "control-alt-delete"
] >to log in! Now /that/ is an excellent interface alright!!
]
] It is, actually - alt-ctrl-del gives a hardware level
] interrupt, which means you can't easily write a trojan log-in
] screen to steal passwords with. First time you've used NT?
Just break the control key --most people probably hit that one
first or second so they won't even notice if it doesn't work.
Plus, the program can call ImpersonateLogInAsUser() and log in so
the victim never even knows. NT: security through false assumptions.
Later, they'll be doing something else when they discover it is
broken and won't make the connection. And if they do, it's too
late anyhow as their password has been changed (using a call to
ImpersonateUserChangePassword) and only the perpetrator knows it!
--
thur Mail Address: LordA...@vt.edu or jmax...@vt.edu
n r
a JAMax "Though it be long, the work is complete and finished
h o w in my mind. I take out of the bag of my memory what
tan lle has previously been collected into it." --Mozart
Blake Stone
: Sangria wrote:
: ] As for as bulletproof as NT, sorry. NT doesn't suffer from having
: ] the interface die on it when an application decides to take breather.
: I was using NT today with the "NewShell" and an error dialog came
: up ...
Wow! You experienced problems running an alpha release designed
as a technology demonstration? What a shock.
: When I got to the machine, it told me to hit "control-alt-delete"
: to log in! Now /that/ is an excellent interface alright!!
Only if you are actually interested in security. The point being
that no application under NT can trap the ctrl-alt-delete
keysequence to immitate the log-in dialog. Now if you just want
to play with pretend security you could use a Mac, sure ...
Trying to claim that the macintosh OS is a more stable or secure
platform than NT is a losing battle. Pick on NT 3.51's user
interface if you like, or the relative lack of 32-bit software to
date that takes advantage of NT, but at least pick something
where you have a point.
--
-------------------------------------------------------------------------
Blake W. Stone bst...@arcane.com
Object Addict - Arcane Systems Ltd. 'Twas brillig, and the slithy toves
Publishers of ThreadKit Did gyre and gimble in the wabe...
Mark Komarinski
: In article <4gt6bg$3...@firebrick.mindspring.com>
: dleb...@mindspring.com (David LeBlanc) writes:
: > rks...@unity.ncsu.edu (Ravi Krishna Swamy) wrote:
: >> In article <3131C0...@access.digex.net>, paul chemmanoor
: >> <ica...@access.digex.net> wrote:
: >>> 1.Linux (minus) In simple terms it is a nightmare for a non-cs
: >>> person to setup. Multitasking theoretically great but in practice
: >>> has problems with Netscape [ yeah I know this is probably
: >>> Netscape's fault.. but ... from a users perspective it is a big
: >>> problem ]
: >> ??? I never had any problem with Linux, Netscape, and
: >> multitasking.
: > Even on my 20MB machine, Netscape is a huge, bloated pig, and is
: > quite capable of bogging everything down, esp. if a Java applet
: > loads. It chews up NT almost as badly. Maybe the 1.2 version
: > didn'd do this - I don't know.
: Netscape can bring both of our 40MB and 24MB NT machines to their
: knees. Netscape will start allocating gobs of memory if you don't
: kill it every so often and restart. I even disabled Java and it still
: goes RAM crazy. The AIX version of Netscape is a fat pig too.
That's what you get when you have one application that tries to be a:
HTML browser
Java compiler/.executor
Mail program
News reader
etc
All in one binary. Even MS isn't so stupid as to put *everything* in one
executable.
--
- Mark Komarinski - koma...@craft.camp.clarkson.edu
Smile. It makes people wonder what you're up to.
Jon A. Maxwell
] : ] the interface die on it when an application decides to take breather.
] : I was using NT today with the "NewShell" and an error dialog came
] : up ...
]
] Wow! You experienced problems running an alpha release
] designed as a technology demonstration? What a shock.
Actually it was NT 4.0 (appearantly I wasn't supposed to know
that because the guy claimed it was 3.51 with new shell, but when
I rebooted it it clearly said NT version 4.0 build 1057 or
something like that).
It certainly explains why it was so slow, anyhow... I imagine it
has a lot of debugging code in there. (Still, it is a 32 meg
pentium.)
] : When I got to the machine, it told me to hit "control-alt-delete"
] : to log in! Now /that/ is an excellent interface alright!!
]
] Only if you are actually interested in security. The point
] being that no application under NT can trap the ctrl-alt-delete
] keysequence to immitate the log-in dialog. Now if you just
] want to play with pretend security you could use a Mac, sure
You are the one talking pretend security if you think console is
secure! If I wanted to get your password from the console I
would boot into DOS from a floppy and run an NT look alike. Or
plug some hardware between the keyboard (to have ctrl-alt-del
send different signal or not at all), or break the keys, etc.
How does NT prevent any of that?
] Trying to claim that the macintosh OS is a more stable or secure
] platform than NT is a losing battle. Pick on NT 3.51's user
] interface if you like,
Which is what I was doing. Thank you for supporting my position.
--
thur Mail Address: LordA...@vt.edu or jmax...@vt.edu
n r
a JAMax "All that we see or seem
h o w Is but a dream within a dream."
tan lle --Edgar Allan Poe
Christopher Robato
writes:
:>
:>bow...@ucsub.Colorado.EDU (BOWEN JASON MICHAEL) wrote:
:>
:>>I read this and it gives me a good chuckle. Of all the graduate E.E.
:>>and Comp. Sci. friends that I have, all laugh when the idea of NT
:>>replacing UNIX comes up. The only people who would buy the
:>>self-prophesising dribble in most industry rags are those who can't
:>>think and evaluate for themselves.
:>>Jason
:>
:>Whether or not they can think and evaluate for themselves, a lot of
:>people are eager to buy NT, *if* certain changes are made. A lot of
:>people buy Unix servers reluctantly and wish that NT were a little
:>more mature.
:>
:>Graduate EE and CS students can be blind to factors that are as
:>important as the technical pros and cons. If technical factors were
:>so important, NeXT would have taken off. Marketing, mindshare,
:>pricing, and industry attitudes are in some ways more influential than
:>technical factors. When the NeXT came out, who would have thought
:>that developers would be forced to abandon it in favor of Windows,
:>which was in an ugly, early version at the time? The NeXT watchers
:>laughed at Windows, but nevertheless were forced to accept it.
:>
:>Pricing strategy is crucial. Microsoft has been very careful to never
:>ask for too much money all at once. The NeXT and its pricey
:>applications required a large lump sum. This demand for a financial
All multiuser/site/client-server style apps cost a bundle. Do you know how
much NT Backoffice costs? It's about $3,500. That's typical of the going
rate of multiuser/site license applications. How much do you
think 100 users for NT would cost?
If you want price, compare it with other Unices.
How about Linux---this is free. You only pay for the distribution media.
The commerical implementation of Linux, Caldera, costs $99. Unixware costs
about $200. Many two user license Unix costs about $600 to $700, like SCO
and AIX. Unlimited licenses can be obtained for about $1000 like BSDI. You
can't get unlimited licenses for NT Server at about $1000. Frankly, Linux
for free is for a virtually unlimited of users too.
On a single user basis, I never found Unix apps more expensive than Windows
apps. Let's take a seat in Lotus Notes or Framemaker.
>sacrifice all at once prohibited OS/2, Mac, Unix, and NeXT from taking
:>off. The MS solution is not cheap when you total the costs, but the
:>ability to incrementally upgrade PCs using building blocks from a vast
:>arena of suppliers has enabled people to make baby steps toward
:>ever-better versions of Windows. The next step is for the Win95 crowd
:>to inch toward NT Workstation. Then, look out.
:>
:>
:>NT is currently a joke compared technically to Unix.
:>
:>Windows was a joke compared technically to NeXT.
:>Windows continually improved and is thriving, while NeXT is dead.
:>
:>
:>The grad geeks can and should laugh at the idea of NT replacing Unix.
:>This highly possible outcome is as ironic as Windows racing past NeXT.
:>
So where did NeXT come in? NeXT isn't even dead yet. Nor is it's fate
of NeXT the fate of all Unices. NeXT did not even thrive as well as all the
other Unices. NeXT is not even on the same category as Windows so why make a
comparison?
The grad geeks can well afford to laugh because the predicted did not
happen. NT did not replace Unix and not anytime soon.
Rgds,
Chris
>>>>** Sailor Moon Joins Team OS/2 **<<<<
FUD covers the city, turning millions into lemmings. Serena and
friends raises their shiny Warp CD ROMs. "Warp 32 bit Power, Tranform!"
Serena turns into Sailor Moon, and they into Sailor Mars, Sailor Mercury,
Sailor Jupiter and Sailor Venus. The Sailor Team OS/2 girls
crashes into the Red Moon palace. Hordes of lemmings rose to fight them.
With moonlight beaming behind their silhouttes, Sailor Moon threatens
the evil Queen Beryl Gates and the diabolical Windowsverse forces,
"In the name of I-B-Moon, I shall right FUD and that means you!"
[[[ cro...@kuentos.guam.net ]]]
Jon A. Maxwell
Another person wrote:
[netscape is huge, godzillion things all in one binary]
] MS put almost all of that in one exe (though they did leave off
] mail), but their browser is still _much_ smaller than Netscape.
] And here I was thinking MS was the king of bloated apps (I like
] some of them, but they are big honkers)...
Oh really? Gee, Internet explorer runs on, what, two platforms?
(Mac and Windows) while Netscape runs on Unix+x, Mac, Windows,
and many different types of unix, too! Also, netscape probably
had to use some bloated windows interface library since Microsoft
saved the /real/ one for their own programs (to make them appear
smaller and faster than the competition, of course).
] If Netscape keeps up in this direction, they won't stay on top
] of the heap for long - having the MS version be faster and
] smaller ought to be a warning sign.... I think MS's exe is 3
] times smaller...
Has that stopped MS? Then why should it stop Netscape?
] Now if we could only get MS to port IE to Linux 8-/
] fat chance, eh?
Right, they would have to make it big and slow then, so that
people could all see how much better the windows version was...
--
thur Mail Address: LordA...@vt.edu or jmax...@vt.edu
n r
a JAMax "Laws are like cobwebs, which may catch small flies, but
h o w let wasps and hornets break through."
tan lle --Jonathan Swift
Jon A. Maxwell
] >So how do I go back to using the old interface? I have
] >guest-level privileges, whatever that means.
]
] you have to have admin priveledge to run the script that undoes
] the stinking thing. I don't know what they are using it for,
] but I don't consider NT with the new shell suitable for
] anything resembling a production environment.
You consider NT with Win3.1 interface usable? Even the NT
version uses project groups and all that, right? If so (I can't
find out since I'm stuck with NewShell) I can't imagine how you
could call that usable...
] Best thing to do is see if someone has a copy of DevNet level
] II, and then they should have the NT 4.0 beta - then you can
] run the new GUI and not suffer horribly.
They insalled the NT so they could print from Windows. I'm not
sure, but I don't think they really want to be running a slow
beta...
Appearantly NT, for me anyhow, gets much more usable once you
have admin access!
--
thur Mail Address: LordA...@vt.edu or jmax...@vt.edu
n r
a JAMax "Mediocrity knows nothing higher than itself; but talent
h o w instantly recognizes genius." --Complete Sherlock Holmes
tan lle Valley of Fear, by Sir Arthur Conan Doyle
Michael Hoffman
><note: the To: line has been pruned.> --norm
Is there any way to bring this thread that I started to a halt, or
alter the distribution? It would be powerful if Usenet enabled the
initiator of a thread to control the distribution of the thread over
time. I doubt it's possible to keep the conversation on track and
retain the complex topic of my question.
The real topic of this thread is the impact of the Web on the OS war,
particularly the potential threat of NT Web servers to Unix, and how
this revised OS contest affects the choice of skills to learn for Web
development.
Aside from *hard numbers*, I've made up my mind well enough. Due to
my Subject line, and the lack of discussion on the Web's
cross-platform nature, this thread is at risk of narrowing down to an
NT vs. Unix debate, that should not be posted to the entire list of
newsgroups I included.
Jon A. Maxwell
]
] It lies. The version number also breaks lots of stuff that thinks it
] is then running on Win95. Build 1057 was 3.51 release. Build 1234 is
] the NT 4.0 beta 1.
Talk about kludge... I thought NT was supposed to be a real
operating system, and Win32 was supposed to be a real API --programs
can't even tell what different API calls to make[*] without having
to just "know" which magic incorrect version number means what!!
* -- since win95 and NT don't implement Win32 in the same way; in
NT some calls work/return errors and in win95 some others are
that way (might as well be two APIs, really).
Besides, Dave, weren't you the one arguing how much better NT's
graphic system was than X windows? (Maybe it was Sang?) When I
move a window so that it is /covering/ another window more and
more, the other window refreshes over and over even though no new
area is being exposed. This is with continuous drag on NT 3.51?
4.0? Talk about suck! I wonder, how do I turn the option on to
make the OS remember what was below a window? (So that I don't
have to see all that annoying refresh) I have guest access
priveledges (whatever that means).
The only good thing I have to say about NT so far is that it
really loaded some things fast (and it didn't crash...yet). Like
the paintbrush app came up almost instantly! Then again, when I
opened up the control panel it took seconds for it to draw all
the icons (it felt like I was using an Amiga back in the .info
days). When I closed it and opened it again it was very quick.
So I guess that NT cache and 32 megs of ram helps out a lot! I
wouldn't say the filesystem is all that fast though, because it
did take quite a while to load the control panel the first time
(This is on a pentium, with scsi thunderspeed harddrive or
something like that... I can look up the exact name, but I get
the impression the actual speed is pretty fast).
When they get a compiler on there then I will compile some speed
programs. The computers also have FreeBSD and NextStep (but I
don't think the NS ones have guest account) so I can make some
good comparisons -- and I'll report the bad-for-nt results and
let people like Dave and Sang report the pro-NT ones.
BTW, does anybody know where I can get GCC for NT? I would like
to use gcc on both NT and FreeBSD!
--
thur Mail Address: LordA...@vt.edu or jmax...@vt.edu
n r
a JAMax "There are three kinds of lies:
h o w lies, damned lies and statistics."
tan lle Attributed to: Mark Twain
Joris Zwart
Michael Hoffman <mic...@cybtrans.com> wrote:
>So what do you think about the impact of the cross-platform Web, Java,
>and Internet terminals on the OS war?
I think Java is a security hole you can drive a bus through, and I think
Internet terminals are a nice idea but that nobody needs them. The Internet
is just today's version of the 1980s viewdata (or whatever you call it)
hype. It has been shown often enough that Joe Average has no interest in
things like that.
>Do you think it is likely that NT will very rapidly gain acceptance
>over Unix for corporate intranets? Will the Unix-to-NT Web server
>ratio suddenly flip-flop?
No and no. It's ridiculous to even think that, with the free Unices around
as a reliable low-cost solution for WWW. I won't even talk about Unix on
corporate networks, where Unix does a lot NT hasn't yet even begun to learn.
Anybody thinking that NT will 'kill' Unix is living in a fantasyworld.
Before we argue further, you first need to explain why the domain in your
NNTP posting host is different from the domain in your e-mail address, which
often means that the post is faked. Asking for advice on advocacy groups when
you have such a controversial opinion, and spend so much space showing it
off, often means a troll. You might well be a phony, so you'll have to
explain that first.
Joris
------------If debugging is the art of removing bugs, then programming must
Joris S. Zwart be the art of putting them in - Edsger W. Dijkstra
jo...@stack.urc.tue.nl http://www.stack.urc.tue.nl/~jozwa/
(PGP public key available by finger)----------------#include <disclaimer.h>
David LeBlanc
>Wayne J. Hyde (w...@rock.cis.ufl.edu) wrote:
>: dleb...@mindspring.com (David LeBlanc) writes:
>: > Even on my 20MB machine, Netscape is a huge, bloated pig, and is
>: > quite capable of bogging everything down, esp. if a Java applet
>: > loads. It chews up NT almost as badly. Maybe the 1.2 version
>: > didn'd do this - I don't know.
>: Netscape can bring both of our 40MB and 24MB NT machines to their
>: knees. Netscape will start allocating gobs of memory if you don't
>: kill it every so often and restart. I even disabled Java and it still
>: goes RAM crazy. The AIX version of Netscape is a fat pig too.
>That's what you get when you have one application that tries to be a:
>HTML browser
>Java compiler/.executor
>Mail program
>News reader
>etc
>All in one binary. Even MS isn't so stupid as to put *everything* in one
>executable.
MS put almost all of that in one exe (though they did leave off mail),
but their browser is still _much_ smaller than Netscape. And here I
was thinking MS was the king of bloated apps (I like some of them, but
they are big honkers)...
I really don't know what Netscape _does_ with all of that RAM. Oddly
enough, it doesn't eat much just doing news or mail - it is when you
start browsing the web that it goes nuts.
If Netscape keeps up in this direction, they won't stay on top of the
heap for long - having the MS version be faster and smaller ought to
be a warning sign.... I think MS's exe is 3 times smaller...
Now if we could only get MS to port IE to Linux 8-/
fat chance, eh?
David LeBlanc
David LeBlanc
>David LeBlanc wrote:
>] Did you notice that the new shell for 3.51 is considered alpha
>] code? It isn't even close to ready for prime time. IN fact,
>] nearly everyone has abandoned using it, since the bugs got
>] really mean and nasty.
>So how do I go back to using the old interface? I have
>guest-level privileges, whatever that means.
Unfortunately, the new shell replaces lots of system files, so you
have to have admin priveledge to run the script that undoes the
stinking thing. I don't know what they are using it for, but I don't
consider NT with the new shell suitable for anything resembling a
production environment.
Best thing to do is see if someone has a copy of DevNet level II, and
then they should have the NT 4.0 beta - then you can run the new GUI
and not suffer horribly.
Don't judge the OS overall by that - it was meant for testing apps
that use the new desktop stuff. Plus how well it works tends to be a
function of patch level, where as patch level increases, bugs increase
- the thing was released at the same time as SP1 - SP2 + newshell
breaks all sorts of stuff - like File Manager, route, etc, etc.
David LeBlanc
>Blake Stone wrote:
>] Jon A. Maxwell (jmax...@csugrad.cs.vt.edu) wrote:
>] Wow! You experienced problems running an alpha release
>] designed as a technology demonstration? What a shock.
>Actually it was NT 4.0 (appearantly I wasn't supposed to know
>that because the guy claimed it was 3.51 with new shell, but when
>I rebooted it it clearly said NT version 4.0 build 1057 or
>something like that).
It lies. The version number also breaks lots of stuff that thinks it
is then running on Win95. Build 1057 was 3.51 release. Build 1234 is
the NT 4.0 beta 1.
David LeBlanc
Pohl Longsine
> 100% correct. Web serving is one of the few areas I think Unix is in
> danger of being crowded aside a bit by NT, because it's something NT does
> well and it's a no-brainer to set up.
My experience is that NT is inappropriate for an internetworked world. One
must be in front of the machine to set it up for web-services. Give me a
telnet session to a UNIX box, and I can do it from anywhere. Even if you
install telnetd on NT, it doesn't help you control GUI-only applications.
> C:\FOOBAR\INSTALL.EXE
(press enter; dial your client in California: "Hi. Could you click 'ok'
for me? Thanks.")
This has not been a difficult thing to explain to clients, who now consider
themselves to be educated about the pitfalls of using NT for web services.
Al Kozakiewicz
says...
>Actually it was NT 4.0 (appearantly I wasn't supposed to know
>that because the guy claimed it was 3.51 with new shell, but when
>I rebooted it it clearly said NT version 4.0 build 1057 or
>something like that).
>
>It certainly explains why it was so slow, anyhow... I imagine it
>has a lot of debugging code in there. (Still, it is a 32 meg
>pentium.)
Installing the newshell preview on a 3.51 system will show the version
number as NT 4.0 on boot up. It isn't. The build number does not change,
however. 105x (can't remember the last digit) indicates 3.51.
K. D. Spoelstra
>Sangria wrote:
>] As for as bulletproof as NT, sorry. NT doesn't suffer from having
>] the interface die on it when an application decides to take breather.
>I was using NT today with the "NewShell" and an error dialog came
Okay, you were using NT with alpha software that gets into the
kernel , fill in a bug report and send it to microsoft, don't bother
us about it...... This is a post for nt.misc ,nt.my-new-shell-killed-nt or
ofcourse nt.pre-release.
If you've got a problem with alpha software , try to find people who
can help you, not using the problem for bashing other people.
BTW: MS states that you mustn't use the new-shell, except for developers
and even then it isn't recommended, hence the term technology PREVIEW
Friendly Greetings,
Kees
Sorry for crossposting....
K. D. Spoelstra
>Actually it was NT 4.0 (appearantly I wasn't supposed to know
>that because the guy claimed it was 3.51 with new shell, but when
>I rebooted it it clearly said NT version 4.0 build 1057 or
>something like that).
It should be 1234, which is the first beta, or even older, still not
release software..
>It certainly explains why it was so slow, anyhow... I imagine it
>has a lot of debugging code in there. (Still, it is a 32 meg
>pentium.)
There were some rumors that it was a checked build, this thing is
that big and is that slow
>You are the one talking pretend security if you think console is
>secure! If I wanted to get your password from the console I
>would boot into DOS from a floppy and run an NT look alike. Or
>plug some hardware between the keyboard (to have ctrl-alt-del
>send different signal or not at all), or break the keys, etc.
How could you boot of a floppy in my machine, if you can't open
my case, can't get in the bios......
>How does NT prevent any of that?
By setting boot access to floppy of...
The other things are only possible if you have long enough access
to the physical system itself, so you don't hack NT, you hack
the physical system.....
>] Pick on NT 3.51's user interface if you like,
>Which is what I was doing. Thank you for supporting my position.
I believe the newshell was your target, but I jumped somewhat later in this
thread.....
Friendly Greetings,
Kees
David LeBlanc
>David LeBlanc wrote:
> Another person wrote:
>[netscape is huge, godzillion things all in one binary]
>] MS put almost all of that in one exe (though they did leave off
>] mail), but their browser is still _much_ smaller than Netscape.
>] And here I was thinking MS was the king of bloated apps (I like
>] some of them, but they are big honkers)...
>Oh really? Gee, Internet explorer runs on, what, two platforms?
>(Mac and Windows) while Netscape runs on Unix+x, Mac, Windows,
>and many different types of unix, too!
I think I made this point - also, Windows is really 2 platforms now.
There are 16 and 32-bit versions. So Netscape can hog the RAM on
UNIX, plus all the same platforms IE runs on. A true cross-platform
RAM hog. Oh joy.
>Also, netscape probably
>had to use some bloated windows interface library since Microsoft
>saved the /real/ one for their own programs (to make them appear
>smaller and faster than the competition, of course).
Get a grip. Your MS-bashing is interfering with your logic. Netscape
has access to the same tools MS uses. No appearance about it - it
_is_ smaller and faster. I have both on the same machine - as I said,
it is a sad state of affairs when Netscape has something which makes
the MS offering look small and fast. Besides, if MS has the magic
small and fast library, then why the hell didn't they give it to the
Word and Excel group?
>] If Netscape keeps up in this direction, they won't stay on top
>] of the heap for long - having the MS version be faster and
>] smaller ought to be a warning sign.... I think MS's exe is 3
>] times smaller...
>Has that stopped MS? Then why should it stop Netscape?
The memory leaks. IE is smaller, faster, and won't wipe out my
machine. I don't know about you, but I don't appreciate having to
telnet into my Linux box to unfreeze the UI. If it was a stand-alone,
I'd be forced to hit the reset. I also don't appreciate having
Netscape eat all of my RAM and swap on my NT box. When something
burns me badly, I tend to remember it.
>] Now if we could only get MS to port IE to Linux 8-/
>] fat chance, eh?
>Right, they would have to make it big and slow then, so that
>people could all see how much better the windows version was...
Yeah, right. Bashing for the sake of bashing, eh?
David LeBlanc
>David LeBlanc wrote:
>] I don't know what they are using it for,
>] but I don't consider NT with the new shell suitable for
>] anything resembling a production environment.
>You consider NT with Win3.1 interface usable?
I call it stable. I hate the UI, but I'll take stability and speed
over nasty bugs and slow response. i used the new shell for a while
and went back, so I speak from experience. I don't consider alpha
code appropriate in a production environment.
>They insalled the NT so they could print from Windows. I'm not
>sure, but I don't think they really want to be running a slow
>beta...
The beta is faster and more stable than 3.51 with the new shell. Look
at it this way - would you prefer to run alpha code or beta code?
David LeBlanc
>David LeBlanc wrote:
>] jmax...@csugrad.cs.vt.edu (Jon A. Maxwell) wrote:
>] >Actually it was NT 4.0 (appearantly I wasn't supposed to know
>] It lies. The version number also breaks lots of stuff that thinks it
>] is then running on Win95. Build 1057 was 3.51 release. Build 1234 is
>] the NT 4.0 beta 1.
>Talk about kludge... I thought NT was supposed to be a real
>operating system, and Win32 was supposed to be a real API --programs
>can't even tell what different API calls to make[*] without having
>to just "know" which magic incorrect version number means what!!
Yes, the new shell for 3.51 is a kludge. I've been telling you that -
it is alpha code, and is a mess. However, don't blame the API for
sloppy programming. The system call returns both version number and
OS type - if I remember correctly, the high bit determines if it is NT
or Win95. Some people get sloppy and don't check the high bit - just
the version. Those people will have to issue fixes as the current
version of Win95 will return 4.0, and so will the next version of NT.
>* -- since win95 and NT don't implement Win32 in the same way; in
>NT some calls work/return errors and in win95 some others are
>that way (might as well be two APIs, really).
Not really. There are differences, but they aren't that bad.
>Besides, Dave, weren't you the one arguing how much better NT's
>graphic system was than X windows? (Maybe it was Sang?) When I
>move a window so that it is /covering/ another window more and
>more, the other window refreshes over and over even though no new
>area is being exposed. This is with continuous drag on NT 3.51?
>4.0? Talk about suck! I wonder, how do I turn the option on to
>make the OS remember what was below a window? (So that I don't
>have to see all that annoying refresh) I have guest access
>priveledges (whatever that means).
I wasn't arguing that - my only gripe with X is that you have to run a
font server or the whole damn UI freezes while something loads.
If you have guest access, then it means the admin considers you a
total newbie, and you have the lowest access level possible. He
didn't even see fit to make you user level. You can't configure
anything by yourself, and if he has it set up right, you probably
couldn't break it if you tried.
>The only good thing I have to say about NT so far is that it
>really loaded some things fast (and it didn't crash...yet).
Most of what you report can be attributed to the new shell.
>BTW, does anybody know where I can get GCC for NT? I would like
>to use gcc on both NT and FreeBSD!
Look in ftp.cygnus.com - they have an on-going project on this.
David LeBlanc
>In <313282a2....@news.fastlane.net> Jay Urbanski wrote:
>> 100% correct. Web serving is one of the few areas I think Unix is in
>> danger of being crowded aside a bit by NT, because it's something NT does
>> well and it's a no-brainer to set up.
>My experience is that NT is inappropriate for an internetworked world. One
>must be in front of the machine to set it up for web-services. Give me a
>telnet session to a UNIX box, and I can do it from anywhere. Even if you
>install telnetd on NT, it doesn't help you control GUI-only applications.
This is one reason you can typically configure most newer HTTP
processes from a browser. Also, it depends on the httpd - it is what
is termed a service. Services can be started, stopped, paused and
installed remotely. The values the GUI would set are held in the
registry, which is editable both remotely, and from a CLI. I can even
write a Perl script to edit the registry.
It helps if you know what the system is capable of - when you do, lack
of a telnet seems like less of a problem all the time.
Jon A. Maxwell
] > Another person wrote:
] >[netscape is huge, godzillion things all in one binary]
] >Also, netscape probably had to use some bloated windows
] >interface library since Microsoft saved the /real/ one for
] >their own programs (to make them appear smaller and faster
] >than the competition, of course).
]
] Get a grip. Your MS-bashing is interfering with your logic.
] Netscape has access to the same tools MS uses.
Mostly I was joking about that one. "Mostly" because who's to
know if MS has held back some API calls, like I understand they
did with Windows 3.1?
] >Has that stopped MS? Then why should it stop Netscape?
]
] The memory leaks. IE is smaller, faster, and won't wipe out my
] machine. I don't know about you, but I don't appreciate having to
] telnet into my Linux box to unfreeze the UI.
Windows 3.1 had all sorts of memory problems, was slow, UI
freezed a lot... did any of that stop it? I ask again, why
should it stop Netscape then?
] Yeah, right. Bashing for the sake of bashing, eh?
That's right. Although when it comes to NT-bashing, it's because
I don't think the world really needs another VMS. =)
--
thur Mail Address: LordA...@vt.edu or jmax...@vt.edu
n r
H.J. Lu
> >character-mode environment with no serious GUI and poor compatibility
> >among the zillion flavors.
>
> I agree with that assessment -- I've found UNIX *very*
> user-hostile -- worse than other CLI OSes I've used, like VMS and even
> VM/CMS (I still miss VMS after all these years).
I found quite opposite. I can do almost whatever I want under Unix. But
it is a pain under Windows95 when your hardware is slightly different
than what Microsoft thought.
H.J.
Jon A. Maxwell
(Or something)
David LeBlanc wrote:
] jmax...@csugrad.cs.vt.edu (Jon A. Maxwell) wrote:
] >Talk about kludge... I thought NT was supposed to be a real
] >operating system, and Win32 was supposed to be a real API
] >--programs can't even tell what different API calls to make[*]
] >without having to just "know" which magic incorrect version
] >number means what!!
]
] The system call returns both version number and OS type - if I
] remember correctly, the high bit determines if it is NT or
] Win95. Some people get sloppy and don't check the high bit -
] just the version.
This is what I am talking about, exactly! The high bit
determines if it is NT or Win95 -- and when can'tCopland comes
out it will use a different API I take it, since you can't get 3
possibilities out of 1 binary bit!
This is exactly the kind of sloppy programming (and now API
design) that would crash AmigaBASIC unless you had a 68000.
They used "unused" bits in pointers to store some data. Needless
to say with the new processors that /used/ those bits came out,
AmigaBASIC would crash. Pathetic!
] >* -- since win95 and NT don't implement Win32 in the same way; in
] >NT some calls work/return errors and in win95 some others are
] >that way (might as well be two APIs, really).
]
] Not really. There are differences, but they aren't that bad.
Then why does it matter if programs think they are running on
another operating version? The whole point of having a common
API is that programs /don't/ have to be concerned with this kind
of incompatibility!
] If you have guest access, then it means the admin considers you
] a total newbie, and you have the lowest access level possible.
They took away my guest account after only a day! Can you
beleive it? But I will try to get a personal account.
] >BTW, does anybody know where I can get GCC for NT? I would
] >like to use gcc on both NT and FreeBSD!
]
] Look in ftp.cygnus.com - they have an on-going project on this.
Thanks, Dave. If I can get my account back this will really help!
--
thur Mail Address: LordA...@vt.edu or jmax...@vt.edu
n r
a JAMax "You can fool all the people some of the time, and some
h o w of the people all the time, but you can not fool all
tan lle the people all of the time." --Abraham Lincoln
David LeBlanc
>David LeBlanc wrote:
>> The beta is faster and more stable than 3.51 with the new shell. Look
>> at it this way - would you prefer to run alpha code or beta code?
>I've heard some people say the beta isn't all that fast; probably due to being
>a checked build (debug build). If you say it is faster than 3.51, speed should
>only get better in the release.
I left off the YMMV. I have heard very different reports about video
speed, which accounts for a lot of the subjective differences in
speed. In terms of program load speed, it is about the same, and it
also seems I have more free RAM. My video is _much_ faster - very
close to as fast as Win95 is on this machine. Someone whose video was
slower might think the OS was slower.
I do think 4.0 will end up smaller and faster overall, but unlikely to
be small enough to justify an 8MB recommended minimum. The beta has
been very solid, and nearly bug-free so far. I was worried a first
beta wouldn't so well, but so far I'm a happy camper.
Gerald W. Edgar
: bad...@seanet.com (Michael R. Delahoz) wrote:
: >To completely implement the Netscape version an IS shop has to invest about
: >$1,700.00 more or less, to do with M$' "free" solution which BTW only runs on
: >NT Server, the projected costs are in the vicinity of $6,000.00
: >BB
: That's simple logic. I get the impression that large corporations are
: yearning to move to NT, and MS's giveaway of the server software is
: the excuse they will use to choose NT servers rather than Unix servers
: running Netscape. It's not logical, it's twisted motives. Many
: corporations *want* a monolithic Microsoft solution. This is the buzz
: I'm picking up from reading various magazines and online news.
If you consider the cost of developing a web site $1,700 is a very small
part of the overall cost. But that 1,700 is an insurance policy for the
capaibity to move to a larger UNIX box. As a professional web developer I
think the money is worth it. If you are not even considering moving to a
larger platform then it does not matter. For small departmental servers
the NT solution is (relatively) cheap. But for a corporate wide site that
may get hit lots, with lots of dynamic pages (ie database access) there
may be a requirement to have a large box for the load. Th bottle neck
would probably be the network connection - how big can it be for a PC
(even running NT) It's larger for a UNIX box.
Gerald Edgar
Professional Internet Developer.
Larry Margolis
>Well, Michael Hoffman has done it again.
>
>It seems he has learned enough since his initial post to decend from his
>self-appointed dogmatic throne; but he's not sufficiently omniscient
>(whatever he may think of himself) to pull together every bit of
>existing and future computing/networking technology/science into a
>coherent and comprehensible model. Now he's finally starting to come
>close to the point where many of us started from.
>
>However, this last post was so full of fallacies, inconsistencies, and
>gross over simplifications that its import can best be summarized quite
>succinctly: "I, Michael Hoffman, do not know what the present or future
>of computation is or will be, but I think my opinions are important
>enough that everyone should hear them."
I'm with you so far...
>To which the only reasonable response is:
>
>CAN WE PLEASE JUST IGNORE MICHAEL HOFFMAN'S MISGUIDED PROGNOSTICATIONS
>AND RETURN TO AN ATTEMPT AT CALM, SENSIBLE, REASONABLE, RATIONAL, AND
>RESPONSIBLE DISCOURSE ON COMPUTERS, COMPUTING, AND TECHNOLOGY?
Calm? Sensible? Reasonable? Rational? Responsible? In the
*.advocacy groups? Are you out of your ever-lovin' *mind*? :-)
--
Larry Margolis, MARGOLI@YKTVMV (Bitnet), mar...@watson.IBM.com (Internet)
Kevin
Gupta) wrote:
> Shouldn't Linux for the PowerMac make it very easy for a non-cs
> person to setup linux (e.g., no weird hardware configs. to worry
> about ? ) IMO, Linux + PowerMac can bring Linux to the masses.
Well, yes and no. The hardware side will be fairly solid, but you
still have the vast array of various directories of system files and
drivers and such on the software side. Linux on PowerMac hardware is still
Linux (Unix by any other name...).
Being a Mac fan, I'm a little torn. I'm glad it's happening, because
it will make the platform that much stronger. I'm just curious as to why
they're taking the time since the Common Platform is only a short time
away.
-Kevin
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
spi...@tickets.stanford.edu
** Spiral Scratch Software: http://tickets.stanford.edu/kevin/ **
** Stanford Ticket Office on the Internet: http://tickets.stanford.edu/ **
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Jon A. Maxwell
] Jon A. Maxwell, jmax...@csugrad.cs.vt.edu says...
[NewShell changes NT 3.51 version to 4.0 so that programs still
work]
]
] >Talk about kludge... I thought NT was supposed to be a real
] >operating system, and Win32 was supposed to be a real API
] >--programs can't even tell what different API calls to make[*]
] >without having to just "know" which magic incorrect version
] >number means what!!
]
^--misspelled rabbit
ulTRIX is for kids...
] The problems experienced with NT 3.51 + Alpha shell is due
] primarily to how the *installation* programs deal with version
] numbers.
]
] For example, Netscape's brain-damaged setup program will
] actually think NT 3.51 + new shell is a Windows 95 machine.
]
] The result in an invalid entry in the Registry. Hence to "fix"
] Netscpe, one goes in and deletes that entry. Presto! No more
] "Can't create a socket connection" error.
Can't create a socket connection -- hardly sounds like a version
problem. After all, the API call to create a socket is the API
call to create a socket right? Or do all programs that use
sockets not run under Win95/NT without having to know which API
to use?
Besides, this is merely a cheap excuse because if Microsoft had
not artificially inflated the NT version number to begin with
this wouldn't have happened! Plus, if Microsoft had not skimped
on their bits (storing the "OS" bit in the version number) this
also would not have happened!
] In otherwords, it's not the OS, it's not the applications, it's
] the silly install programs. They do a GetVersionEx() and don't
] bother looking beyond dwMajorVersion (which would mistakenly
] report 4...)
huh? if the API is the same, why should the install programs
have to care?
--
thur Mail Address: LordA...@vt.edu or jmax...@vt.edu
n r
a JAMax "Though it be long, the work is complete and finished
h o w in my mind. I take out of the bag of my memory what
tan lle has previously been collected into it." --Mozart
Sangria
>I was using NT today with the "NewShell" and an error dialog came
>up (just a little permission denied thing, nothing of vital
>importance). So I clicked on another window --oops, not allowed
>sorry BEEP.
Welll, gosh. Maybe that's why they call that an "Alpha" version?
You want to start arguing whose Alpha release is more buggy? :-)
>When I got to the machine, it told me to hit "control-alt-delete"
>to log in! Now /that/ is an excellent interface alright!!
Yep that part is.
Ever considered how easy it would be to write a trojan horse to
look and behave exactly like a logon screen?
CTRL+ALT+DEL is to get around that problem. No application can
"trap" those three key combination.
-- Sang.
*************************************************************
* Sang K. Choe san...@inlink.com *
* http://www.inlink.com/~sangria/index.html *
* Or http://sangria.inlink.com/index.html *
* finger: sa...@sangria.inlink.com *
*************************************************************
Robert M. Martel
: Yes, but I've found managing a Sun to be easier than managing a
: Windows/DOS box.
No doubt in my mind about that! (I manage multiple unix platforms *and*
two PC Networks -guess which takes up 75% of my time?)
--
******************************************************************************
Bob Martel - System Administrator | I met someone who looks a lot like you
Levin College of Urban Affairs | She does the things you do
Cleveland State University | But she is an IBM
(216) 687-2214 |
b...@meeker.csuohio.edu | -Jeff Lynne
b...@cua6.csuohio.edu |
******************************************************************************
Sangria
>I've heard some people say the beta isn't all that fast; probably due to being
>a checked build (debug build). If you say it is faster than 3.51, speed should
>only get better in the release.
Two words: Video Driver.
What's included with NT 4.0 beta 1 sucks.
It's like putting in a $400 video card and getting a generic
800x600 display with no acceleration whatsoever. YUCK!
Once Matrox releases a "real" driver, I would expect the speed to
be somewhat better.
This is also the reason why some people claim "it's faster" while
other's claim "it's slower". I'm betting that these two groups
don't have the same video card nor are their video drivers of the
same quality...
Sangria
>po...@screaming.org (Pohl Longsine) wrote:
>...
>>My experience is that NT is inappropriate for an internetworked world. One
>>must be in front of the machine to set it up for web-services. Give me a
>>telnet session to a UNIX box, and I can do it from anywhere. Even if you
>>install telnetd on NT, it doesn't help you control GUI-only applications.
>
>This is one reason you can typically configure most newer HTTP
>processes from a browser.
Or in the case of IIS, you would simply fire up ISM and tell it go
and find the IIS running on the remote machine. Once that's done,
you just configure the setup any way you want.
>...The values the GUI would set are held in the
>registry, which is editable both remotely, and from a CLI. I can even
>write a Perl script to edit the registry.
Ok, you can do that too...although, that's a bit more clumsy, but
if you don't have a copy of ISM on your machine at the time this would
be the way to do. Everything that can be done from ISM can be done
by directly messing around with the Registry. You may need to stop
and restart the service if you alter one of the "initilization"
information, but that's nothing more than a quick NET * command.
(ISM would do that automatically for you...)
Michael Hoffman
>In <313400...@wolfenet.com>, Mark Hamstra <mham...@wolfenet.com> writes:
>>Well, Michael Hoffman has done it again.
>>
>>It seems he has learned enough since his initial post to decend from his
>>self-appointed dogmatic throne; but he's not sufficiently omniscient
>>(whatever he may think of himself) to pull together every bit of
>>existing and future computing/networking technology/science into a
>>coherent and comprehensible model. Now he's finally starting to come
>>close to the point where many of us started from.
>>
>>However, this last post was so full of fallacies, inconsistencies, and
>>gross over simplifications that its import can best be summarized quite
>>succinctly: "I, Michael Hoffman, do not know what the present or future
>>of computation is or will be, but I think my opinions are important
>>enough that everyone should hear them."
>I'm with you so far...
Could you be more specific? Your general assessment of my previous
long posting doesn't go into any detail, so it cannot persuade. I
don't know what your position is on my various predictions.
What is your prediction about the impact of the Web on the OS war?
Will the underdog OSs such as Amiga and GEOS have a chance to increase
their user base? Will the cross-platform nature of the Web enable new
OSs to come out of the labs and race past both NT and Unix?
>>To which the only reasonable response is:
>>
>>CAN WE PLEASE JUST IGNORE MICHAEL HOFFMAN'S MISGUIDED PROGNOSTICATIONS
>>AND RETURN TO AN ATTEMPT AT CALM, SENSIBLE, REASONABLE, RATIONAL, AND
>>RESPONSIBLE DISCOURSE ON COMPUTERS, COMPUTING, AND TECHNOLOGY?
This conversation has never departed from calm and reasonable
discussion except for about 3 postings out of 117.
Rob J Meijer
>What's your prediction for the most likely future scenario regarding
>the Web's impact on the various OS markets? Servers, desktop OSs, and
>Internet terminals might have very different market shares. For
>example, it's highly possible that most servers will be Unix, while
>most desktop computers will run some form of Windows. Your scenario
>could affect my decision of how much to learn about the various web
>tools and environments.
I think the internet is making a big contribution in showing people that
unix OS systems are way better performers than systems running a MS OS.
Although i think hardware vendors will keep on selling their hardware
with MS osses on them for a while, giving MS some more time to live.
Since Win 95 has been out, i have seen more of my internet active friends
replace their win3.xx os by Linux, than by Win-95.
And ass for the server side, a P90 based www server running Linux
outperforms the same machine running NT.
I think MS could be swallowed by UNIX on the internet within the next 5
years. Lets face it, MS just hasn't made a real internet ready OS yet,
and the internet makes Unix verry available, and support for the os even
more available.
Jay Urbanski
:I left off the YMMV. I have heard very different reports about video
:speed, which accounts for a lot of the subjective differences in
:speed. In terms of program load speed, it is about the same, and it
:also seems I have more free RAM. My video is _much_ faster - very
:close to as fast as Win95 is on this machine. Someone whose video was
:slower might think the OS was slower.
:
:I do think 4.0 will end up smaller and faster overall, but unlikely to
:be small enough to justify an 8MB recommended minimum. The beta has
:been very solid, and nearly bug-free so far. I was worried a first
:beta wouldn't so well, but so far I'm a happy camper.
I've run into two problems with it, one a bug - the other poor design. The
bug is it often fails trying to shutdown if I have Netscape open - the
video just goes crazy and I have to cold boot. The other pisses me off
because it's just pure stupidity. The task manager is back, no doubt from
all the howls from it being removed with newshell - great. Except the only
reason I ever want to use the damn thing is to kill explorer if it hangs
(and it does), but it relies on explorer to launch it! Hang explorer, then
press ctrl-alt-delete, select task manager, and watch it bring the hung
explorer to the foreground - no task manager. If you don't have a command
line open it's reset button at this point. Well, just wanted to vent.
Jay Urbanski
MCSE
Certified Solaris Administrator
Paranet
K. D. Spoelstra
>>When I got to the machine, it told me to hit "control-alt-delete"
>>to log in! Now /that/ is an excellent interface alright!!
>Yep that part is.
>Ever considered how easy it would be to write a trojan horse to
>look and behave exactly like a logon screen?
>CTRL+ALT+DEL is to get around that problem. No application can
>"trap" those three key combination.
No application can trap it in NT, except: the keyboarddriver
you can write a keyboarddriver which trapped it, but then
you're hacking on hardware again, not on the software.
And you have to get the sysadmin to install it ;-)
Friendly Greetings,
Kees
Kristian Koehntopp
>When I got to the machine, it told me to hit "control-alt-delete"
>to log in! Now /that/ is an excellent interface alright!!
This does actually make sense. The infamous three finger
salute does generate a certain signal (a non maskable
interrupt) that is going directly to the OS and circumvents
any application currently running. Thus, the user can be sure
to talk directly to the kernel and not to some application
spoofing the login panel.
This feature is called a "secure attention key (SAK)" to open a
"trusted communication path" and can be found in some Unices as
well. IBM AIX for example has the same feature if you configure
it correctly. It uses the characters Control-A (ASCII 01) for
this purpose.
Kristian
--
Es steht dort im Prozessorfeld | Da kommt ein Rechnerarchitekt
Seit kurzem ein Professorzelt. | Und fragt ihn, ob der Chip denn schmeckt.
Davor sitzt ein Professor | Da laechelt der Professor:
Und brät einen Prozessor. | Nur Software schmeckt noch besser.