Microsoft toadies strike again!
Jeff Bonwick
>
> MS innovates just fine.
>
> * first non-UNIX mainstream OS to support SMP -- Windows NT
First non-Chrysler, non-GM car company to make air bags standard: Ford Motors!
Hey, let's not forget the many other brilliant innovations from Microsoft:
* first Windows-compatible OS without an 8.3 filename limit
* first non-Macintosh, non-workstation, non-Xerox window system
* first non-vi, non-emacs, non-WordStar, non-WordPerfect, non-Mac visual editor
* first non-Lotus spreadsheet
* first non-Internet, non-AOL, non-CompuServe, non-Prodigy on-line service
* first non-Mosaic, non-Java, non-Netscape Web browser
BTW, I just can't wait to get Encarta, Microsoft's new on-line encyclopedia.
I want to use it to learn more about the history of computing, such as how
Bill Gates invented networking, multitasking, virtual memory, threads, etc.
Jeff Bonwick
Solaris Performance
pie...@jersey.net
>> m...@aplace.com (Smoke Crack and Worship Satan) writes:
>>
>> MS innovates just fine.
>>
>> * first non-UNIX mainstream OS to support SMP -- Windows NT
>
>First non-Chrysler, non-GM car company to make air bags standard: Ford Motors!
>
>Hey, let's not forget the many other brilliant innovations from Microsoft:
>
Hell, even most of *these* aren't Microsoft innovations:
>* first Windows-compatible OS without an 8.3 filename limit
That would be OS/2.
>* first non-Macintosh, non-workstation, non-Xerox window system
That would be Apple's Lisa.
>* first non-vi, non-emacs, non-WordStar, non-WordPerfect, non-Mac visual editor
That would be Valdocs, probably others too.
>* first non-Lotus spreadsheet
That would be VisiCalc.
>* first non-Internet, non-AOL, non-CompuServe, non-Prodigy on-line service
That would be GEnie and EWorld, among others.
>* first non-Mosaic, non-Java, non-Netscape Web browser
That would be WebExplorer, at least (there may be others I don't know about).
>
>BTW, I just can't wait to get Encarta, Microsoft's new on-line encyclopedia.
>I want to use it to learn more about the history of computing, such as how
>Bill Gates invented networking, multitasking, virtual memory, threads, etc.
>
>Jeff Bonwick
>Solaris Performance
Yeah, isn't Microsoft the greatest company on earth?
Shawn Dessaigne
/-------------------------------------------------\
|pie...@jersey.net |"It's difficult to work |
|pgf...@prodigy.com |a group when you're |
|------------------------|omnipotent." -Q |
|OS/2, Amiga, Apple IIgs,|------------------------|
|Epson QX-10, Epson QX-16| WHS Class of '97 |
\-------------------------------------------------/
J. Poag
: lily-white. But they do deserve credit for almost single handedly bringing the
: PC to where it is today.
You should be fucking tied down and whipped bloody.
Because of Gates and Microsoft, the computer industry has had to endure
over a decade of completely poor, SHAMEFUL, pathetic excuses for software
and program launchers dressed up as real OS'es... Wake up out of the trance,
boy--Microsoft's PR folks would like you to believe that Gates is a step
below Jesus Christ, leading us poor innocent lost sheep into the promised-land
of computing.
Grab a clue and ask yourself WHO DRAGGED US INTO THIS MESS TO BEGIN WITH.
: The only way standards get set in the PC industry is by
: sheer force of market weight. It's a bit like the way Netscape proprietary HTML
: extensions are now "standard." Running roughshod over the industry to promote
: your own agenda may not be pleasant, but that's how things get done in the
: absence of a controlling entity like Apple over Macintoshes. And MS has
: generally done a very good job of it-- certainly better than IBM, Borland,
: Lotus, Novell, et al. MS is about the only software company that has managed to
: avoid shooting itself in the foot over the years.
--
+-+----------------------------------------------------------------p-+-q
|O| Signature v1.15b |o|O|
+-+----------------------------------------------------------------b-+-d
| |
| Im a whiz-kid of the Superinformation Highway! I wash ALL my clothes |
| with BOLD Laundry Detergent, I spend ALL my time in coffee-shops AND |
| I drive a Subaru Impreza--Because it's just like punk-rock!" |
| |
+---------------------------------------------------------------------pq
+=====================================================================bd
Bowie J. Poag (b...@interaccess.com) 50 Mhz A2500HD/030/882,AmiTCP/IP 4.2
(I drive an '85 Dodge Omni, hate coffee, and listen to Devo. Bite me.)
Jeff Bonwick
> operating system.
There are a lot more machines running Unix than Windows NT --
and despite repeated predictions of its imminent demise,
Unix continues to grow and prosper. Irritating, isn't it?
> Would you like to walk your grandmother through installing Linux
> on her PC? And installing X Windows? And purchasing shrink wrapped
> applications for her UNIX based PC?
About as much as I'd like walking my grandmother through adding a disk
to her PC. "This is the J4100 jumper. According to the manual we should
place this little do-hickey across it, because the EISA configuration
utility told us that this IRQ is already taken by ..." Or explaining
to her that because of a bug in some other program, the machine locked
up and she's just lost the document she was working on.
Let's face it, computers have a *long* way to go. The differences
in ease of use between Unix and Windows are negligible compared to
the difference between where we are and where we need to be.
It's like the Incas and the Mayas arguing over whose technology
is better positioned to put a man on the moon.
> MS is about the only software company that has managed to
> avoid shooting itself in the foot over the years.
So far that's been true, but at the moment Billy is holding a loaded
pistol that appears to be pointing in a generally downward direction.
Microsoft is trying to figure out how to "own" the Internet, which
reveals a deep misunderstanding of what the 'net really is.
It's going to be an interesting year.
Jeff Bonwick
Solaris Performance
T. R. Kinney
> > m...@aplace.com (Smoke Crack and Worship Satan) writes:
> >
> * first non-Lotus spreadsheet
> Jeff Bonwick
> Solaris Performance
You mean non VisiCalc spreadsheet. Lotus is just another
non-innovative copy cat. Like Apple, IBM, an everybody
else who didn't breast feed at the PARC.
T. R. Kinney
> : It's very popular to rag Microsoft these days. And I agree that Microsoft is not
> : lily-white. But they do deserve credit for almost single handedly bringing the
> : PC to where it is today.
>
> You should be fucking tied down and whipped bloody.
>
> Because of Gates and Microsoft, the computer industry has had to endure
> over a decade of completely poor, SHAMEFUL, pathetic excuses for software
> and program launchers dressed up as real OS'es... Wake up out of the trance,
> boy--Microsoft's PR folks would like you to believe that Gates is a step
> below Jesus Christ, leading us poor innocent lost sheep into the promised-land
> of computing.
>
> Grab a clue and ask yourself WHO DRAGGED US INTO THIS MESS TO BEGIN WITH.
Get a life, and then realize that everyone else has one, and it's different than yours.
And most are probably more than satisfied to use what makes "them" successful. Though I would
like to try whatever OS encourages such "I'm Mr.Bad Ass" posts..
Jeff Bonwick
> a world of difference between Win95/System 7.5.x and UNIX. It's the
> difference between CLI and GUI.
Eh? Unix has had window systems (GUIs) since before Microsoft existed.
And because the window system is itself just another application in the
Unix world, you can run any window system or any app you want. You can
even run MS Windows/x86 binaries on Solaris/SPARC systems via Wabi.
So with MS I get exactly one window system and a complete joke of a CLI.
With Unix I get whatever window system I want and a really powerful CLI.
Expect to see even more cross-environment support from the Unix world.
Expect to see open, public, free, hardware-neutral technologies like
Java destroy the Microsoft model (formerly known as the IBM model)
of locking people into one vendor. Expect the rules to change.
Jeff Bonwick
Solaris Performance
James D Pollock
Um, actually, Lotus was only the biggest fish in the IBM PC world. VisiCalc
predates Lotus, and was, in fact, the first spreadsheet software.
>BTW, I just can't wait to get Encarta, Microsoft's new on-line encyclopedia.
>I want to use it to learn more about the history of computing, such as how
>Bill Gates invented networking, multitasking, virtual memory, threads, etc.
It's very interesting reading. Gates gets an article (with a video in it no less).
Paul Allen (you may have heard of him) gets nothing.
Also oddly absent (at least from the 95 edition) is Intel. Apparently, inventing
the memory chips and the microprocessor used in most of the machines the software is
run on just isn't very important.
Ian G Batten
Jeff Bonwick <bon...@jurassic.eng.sun.com> wrote:
> > m...@aplace.com (Smoke Crack and Worship Satan) writes:
> >
> >
> > * first non-UNIX mainstream OS to support SMP -- Windows NT
>
> First non-Chrysler, non-GM car company to make air bags standard: Ford Motors!
Not merely is claiming things as new because you're the first
implementor aside from the Unix crowd ludicrous, on this occasion it's
arguably wrong anyway. Depending upon how S you think the MP in Multics
was, of course.
ian
Charles
: Cute, but you miss the point. UNIX will never be a mainstream operating system.
: Would you like to walk your grandmother through installing Linux on her PC? And
: installing X Windows? And purchasing shrink wrapped applications for her UNIX
: based PC?
I wouldn't want to walk my grandmother through installing Windows .95 either.
You missed his point too. He wasn't promoting Unix has a mainstream OS,
he's just sick of MS, and other "PC" companies promoting what they do
as "NEW" technologies when the Unix and even mainframe world have been
it doing for 20 years.
--
Harnad describes the status of Usenet aptly:
a communication medium with revolutionary intellectual
potential being used mostly as a global graffiti board.
[Harnad 1991]
Dana Myers
Steve Kay <st...@starbug1.demon.co.uk> wrote:
>In article <48f551$7...@engnews2.Eng.Sun.COM>
> bon...@jurassic.eng.sun.com "Jeff Bonwick" writes:
>
>> > m...@aplace.com (Smoke Crack and Worship Satan) writes:
>> >
>> > MS innovates just fine.
>> >
>> > * first non-UNIX mainstream OS to support SMP -- Windows NT
>>
>>
>> BTW, I just can't wait to get Encarta, Microsoft's new on-line encyclopedia.
>> I want to use it to learn more about the history of computing, such as how
>> Bill Gates invented networking, multitasking, virtual memory, threads, etc.
>>
>
>to declare it as a superior open-systems OS of course :-)
It is novel to look back and remember that Microsoft actively sold
Xenix (originally based on V7, later System III) in the early 1980s.
I even recall an article by Gates in Byte circa mid 1980 plugging
Xenix.
Of course, some people would say that V7 was the last great version
of Unix, but that was before people expected things like networking,
virtual memory, threads, etc. in the OS.
--
* Dana H. Myers KK6JQ, DoD#: j | Views expressed here are *
* (310) 348-6043 | mine and do not necessarily *
* Dana....@West.Sun.Com | reflect those of my employer *
Steve Chapin
}}
}} Microsoft is almost single-handedly for the state of PC Computing
Make that "single-handedly responsible." Stupid fingers.
s...@mcs.kent.edu Steve Chapin Today's Usage Lesson:
I advise that you take his advice.
The worth of an idea is independent of its progenitor.
Cathy Mancus
In article <48gnn1$s...@nntp.crl.com>, m...@aplace.com (Smoke Crack and Worship Satan) writes:
> In article <48gdgn$d...@engnews2.Eng.Sun.COM>, bon...@cathy.eng.sun.com (Jeff
> Bonwick) wrote:
> And there will be more machines running NT than UNIX by 1997 IMHO.
I am resisting the urge to make a very large bet with you... I
will instead just settle for laughing uproariously.
/-------------------------------------------------------------------\
| Catherine Mancus <ca...@zorac.cary.nc.us> |
| PP-SEL, N5WVR "God is a sponge." |
\-------------------------------------------------------------------/
Dave Cinege
>> > > * first non-UNIX mainstream OS to support SMP -- Windows NT
>> >
>Motors!
>
>> Not merely is claiming things as new because you're the first
>> implementor aside from the Unix crowd ludicrous, on this occasion it's
>> arguably wrong anyway. Depending upon how S you think the MP in Multics
>> was, of course.
>
>Honeywell, Control Data Corporation, NCR, Burroughs.
>
>First OS to support SMP other than those OS's that preceded it - Windows NT.
What about OS/2 SMP? That's been around about as long as NT if not longer...
----------------------------------------------------------------------------
Help stop organized crime......
KILL A COP!
___ _|
// /\ /\
Ted! Dan! Stop that!
Lawson English
: bon...@jurassic.eng.sun.com (Jeff Bonwick) writes:
: > > m...@aplace.com (Smoke Crack and Worship Satan) writes:
: > >
: > * first non-Lotus spreadsheet
: > Jeff Bonwick
: > Solaris Performance
: You mean non VisiCalc spreadsheet. Lotus is just another
: non-innovative copy cat. Like Apple, IBM, an everybody
: else who didn't breast feed at the PARC.
Jef Raskin, inventer of the Mac, came up with the idea and guiding
principles before Jobs ever went to PARC.
In fact, Raskin visited PARC when it first opened.
In fact, Raskin described WYSIWYG printing in his 1967 Master's Thesis:
The QUick Draw Graphics Engine.
PARC was cool, but there were people worried about human interface design
before PARC was founded.
--
-------------------------------------------------------------------------------
Lawson English __ __ ____ ___ ___ ____
eng...@primenet.com /__)/__) / / / / /_ /\ / /_ /
/ / \ / / / / /__ / \/ /___ /
-------------------------------------------------------------------------------
Smoke Crack and Worship Satan
> There are a lot more machines running Unix than Windows NT --
> and despite repeated predictions of its imminent demise,
> Unix continues to grow and prosper. Irritating, isn't it?
>
Not irritating at all. UNIX has its place. But there are many more machines running System
7.5.x and Windows than UNIX. And there will be more machines running NT than UNIX by 1997
IMHO.
>
> Let's face it, computers have a *long* way to go. The differences
> in ease of use between Unix and Windows are negligible compared to
> the difference between where we are and where we need to be.
> It's like the Incas and the Mayas arguing over whose technology
> is better positioned to put a man on the moon.
>
This sounds suspiciously like DOS advocates arguing how lame the Mac OS is. Yes, all
computers are a pain in the butt, that's true. But there's a world of difference between
Win95/System 7.5.x and UNIX. It's the difference between CLI and GUI. Hardcore UNIX users
love the CLI, as they should-- that's what UNIX is all about. Not that it can't be
efficient, it's just inappropriate for almost everyone except hardcore computer geeks.
> So far that's been true, but at the moment Billy is holding a loaded
> pistol that appears to be pointing in a generally downward direction.
> Microsoft is trying to figure out how to "own" the Internet, which
> reveals a deep misunderstanding of what the 'net really is.
> It's going to be an interesting year.
Yes, it will be interesting. I don't think MS is interested in "owning the net" per se-- I
think that's people's image of Microsoft. As if anyone could own the net! It's not possible.
There will always be HTML 2.0, Gopher, newsgroups, etc. regardless of what Netscape or
anyone else specified above and beyond that. The least common denominator stuff will be
around for a long time. Furthermore, lost in all the din is that the network is *not* the
computer. The computer and the network are equally important IMHO.
Steve Chapin
}}
}} It's very popular to rag Microsoft these days. And I agree that
}} Microsoft is not lily-white. But they do deserve credit for almost
}} single handedly bringing the PC to where it is today.
I wouldn't put it quite that way. It would be accurate to say that
Microsoft is almost single-handedly for the state of PC Computing
today. However, when you realize that:
(a) MS-DOS is a non-reentrant program monitor, i.e. 1960's technology
appearing in the 1980's,
(b) Windows NT is VMS all over again, i.e. 1970's technology
appearing in the 1990's, and
(c) Windows 95 is knock-off of the '87 Mac OS, i.e. 1980's
technology in the 1990's,
you'll notice a trend. To wit, Microsoft is consistently delivering
out-of-date technology to users. Their OS products lack innovation,
but they rely on the ignorance of their users and their monopolistic
marketing position (thanks to IBM) to keep the gravy train going.
You are correct that Microsoft is responsible for bringing, or rather
holding back, the PC where it is today. The sad part comes when you
think about where the PC could be if Microsoft applied some of their
resources towards advancing the state-of-the-art. Hopefully some of
the better ideas from their U.S. and Australian research labs will
start to affect the commercial side of the company.
Ling Wang
>Would you like to walk your grandmother through installing Linux on her PC? And
>installing X Windows? And purchasing shrink wrapped applications for her UNIX
>based PC?
I sure wouldn't like to show my grandma how to install any MS
windows application. Having to run to every desktop with 20
floppies and install each one. And even with CDROM installs,
we would have to sit pretty much thru the whole install cycle
to answer questions here and there. NFS and automount is not
a idea even MS could percieve. I see PC SA's running to each
and every of hundreds of PC's and doing physical mounts of file
systems every time a new file system has to be mounted. And when
you talk to MS/PC people about mounts that is not constrained to
a drive letter, they do not grab the concept. MS brainwashed
all the PC users to the concept of A:-Z: that anything else is a
foriegn or non-existent language to them.
Our Sun systems with jumpstart, all we have to do is put a
brand new unistalled machine on the user's desktop and hook it
up the network and turn it on. Everything else is done
automatically, from installing the OS, options, special system
mods, softwares and reboots itself at the end so the user could
login. This is for valid for local and remote sites. The only
thing my grandma has to do is turn on the power and walk away.
Installing applications on the server does not take much time.
and with NIS, NFS, automount, VUE/CDE's action based way of
starting applications. I will do it on the server, and the only
thing every entitled user has to do is rebuild their actions
database. There is absolute no other requirement on the client
side. Everything is done on server, even when it is on a new server
or new file system or new mount.
My group of 3 SA's support 200-300 Sun boxes at 11 sites in 8 cities.
In the regional sites, we only have some PC techs who would change
tapes or power cycle hosed machines for us. I cannot imagine
having to support it if they were all PC's.
--
### ### +------------------------
# ###### #### | Ling Wang
## ### ########## | ICST Unix Admin
# ###### ## ### | Voice: (212)236-2183
# # ## ## #### | Fax: (212)236-3565
# ## #### | Email: lwang@.ml.com
### ### ## +------------------------
## ## ### #### ### | Merrill Lynch
## ## ## ### | Enterprise Services
### #### ## ### | World Financial Center
## ###### ########## ## | South Tower,14th Floor
# ## ## # # ## ### ## | New York, NY 10080-6114
# ## #### # ## # ## +------------------------
##### #### ####
Smoke Crack and Worship Satan
Poag) wrote:
>
> : It's very popular to rag Microsoft these days. And I agree that Microsoft is
> not
> : lily-white. But they do deserve credit for almost single handedly bringing
> the
> : PC to where it is today.
>
>
> Because of Gates and Microsoft, the computer industry has had to endure
> over a decade of completely poor, SHAMEFUL, pathetic excuses for software
> and program launchers dressed up as real OS'es... Wake up out of the trance,
> boy--Microsoft's PR folks would like you to believe that Gates is a step
> below Jesus Christ, leading us poor innocent lost sheep into the promised-land
> of computing.
>
> Grab a clue and ask yourself WHO DRAGGED US INTO THIS MESS TO BEGIN WITH.
>
Oh, stop. You're reinforcing all my stereotypes of hysterical internet users. And let me be
the first to tell you that if it wasn't Microsoft, it would have been another company doing
the same kinds of stuff. You know... whoever has the gold (market share) makes the rules.
Net people just get all gooshy and doe-eyed when they talk about Netscape. And what's
Netscape doing? Running around setting proprietary HTML standards and languages through
sheer brute force-- just like Microsoft! Amazing how that works, isn't it? Messianistic
complex happens to the best of us.
And look at the track records of every other major software publisher in the PC world:
abysmal! Microsoft isn't brilliant-- the competition just screws up all the time. What the
hell happened to WordPerfect, 1-2-3, or OS/2? Novell, Lotus, and IBM, that's what happened!
Smoke Crack and Worship Satan
Bonwick) wrote:
> Eh? Unix has had window systems (GUIs) since before Microsoft existed.
> And because the window system is itself just another application in the
> Unix world, you can run any window system or any app you want. You can
> even run MS Windows/x86 binaries on Solaris/SPARC systems via Wabi.
> So with MS I get exactly one window system and a complete joke of a CLI.
> With Unix I get whatever window system I want and a really powerful CLI.
Of course UNIX has grafted-on windows systems. But what self respecting UNIX jock *uses* a
windowing system? People who use UNIX learn to take advantage of the CLI-- that's how you
get the gritty stuff done. It's rather like strapping Windows, desqview, and Norton
Commander on top of DOS IMHO.
> Expect to see even more cross-environment support from the Unix world.
> Expect to see open, public, free, hardware-neutral technologies like
> Java destroy the Microsoft model (formerly known as the IBM model)
> of locking people into one vendor. Expect the rules to change.
UNIX will always have a life as a server and expensive workstation OS. The question I would
pose is, "what can the flavors of UNIX do that NT can't"? In those situations, people should
use UNIX. Otherwise, NT is a hell of a lot easier.
Alan L. Stange
> Cute, but you miss the point. UNIX will never be a mainstream operating system.
> Would you like to walk your grandmother through installing Linux on her PC? And
> installing X Windows? And purchasing shrink wrapped applications for her UNIX
> based PC?
>
> UNIX is great for those who need it, but for everyone else it sucks. Period.
>
> It's very popular to rag Microsoft these days. And I agree that Microsoft is not
> lily-white. But they do deserve credit for almost single handedly bringing the
^^^^^^
interesting mis-spelling of the word _blame_....
> PC to where it is today. The only way standards get set in the PC industry is by
> sheer force of market weight. It's a bit like the way Netscape proprietary HTML
> extensions are now "standard." Running roughshod over the industry to promote
> your own agenda may not be pleasant, but that's how things get done in the
> absence of a controlling entity like Apple over Macintoshes. And MS has
> generally done a very good job of it-- certainly better than IBM, Borland,
> Lotus, Novell, et al. MS is about the only software company that has managed to
> avoid shooting itself in the foot over the years.
--
Alan L. Stange Brookhaven National Lab sta...@ccd.bnl.gov
http://dcsadm.bnl.gov/~stange/
Smoke Crack and Worship Satan
(Charles) wrote:
>
> You missed his point too. He wasn't promoting Unix has a mainstream OS,
> he's just sick of MS, and other "PC" companies promoting what they do
> as "NEW" technologies when the Unix and even mainframe world have been
> it doing for 20 years.
>
Right, but the very fact that many of these technologies are *available* on the
PC is impressive. Ever tried to do SMP on a Commodore 64? You can buy dual
processor Pentium-133s *today* from Micron, for God's sake!
Basically, being sick of MS is really boring. Unless you've got something
factual and concrete to add, whine to someone else about how MS is going to
destroy the world. It's not original. It's not even true. Ever read that Aesop's
fable about the fox and the grapes? If you're pissed off at MS, it's more
effective to spend your time developing killer products to compete with MS. Or
you can just whine, which is what most people seem to be doing nowadays.
Sinan Karasu
Smoke Crack and Worship Satan <m...@aplace.com> wrote:
>In article <48icl3$p...@brtph500.bnr.ca>, man...@bnr.ca (Cathy Mancus) wrote:
>>
>> : And there will be more machines running NT than UNIX by 1997 IMHO.
>>
>> I am resisting the urge to make a very large bet with you... I
>> will instead just settle for laughing uproariously.
>
>what percentage of those users will opt for Windows NT. As for UNIX-- the largest
>selling RISC machine is the PowerMac. Let's make it 1998. :)
What is happening right now under our noses:
1) Unix companies are pushing Win3 users to switch to win95 and NT
by making browsers/languages that run under win95 and NT but not under
win3.x (I should not have to spell this out. Perceptively
challenged need not understand.)
2) Thus making the Windows programs portable to Unix.
3) Creating libraries to make all Windows 95/NT programs
run natively under Unix (once compiled)
4) Making free/cheap products (such as Netscape) to include
things like Word/Excel/... thus eating into Microsoft
market share.
5) MakeServers/Browsers provide things like Encarta etc....
to eat into that market share by Microsoft.
Mirosoft knows that their bread and butter comes by selling hard
products. They have not been able to break into the Net provider
business. It also is a red-herring for Microsoft. While they
while away their time on Blackbird,the attack from other directions
will eat them alive. 10% revenue loss is tantamount to losing
many quarters profits.
Sinan
PS: Don't waste your flames by private e-mail. They'll
be filtered.If you wish to flame, please post-it....
--
Redistribution by Microsoft Network is prohibited.
Jeff Bonwick
> Of course UNIX has grafted-on windows systems.
You're confusing 'grafted-on' with 'well-modularized'. SunView was
built into the SunOS 3.x kernel many years ago -- it was an *advance*
to move the window system out of the kernel. If you want examples
of grafted-on, how about the way Windows converts long filenames into
*visible* underlying 8.3 filenames? Or its use of the "f:" and "g:"
"network disks" as a network programming model?
> But what self respecting UNIX jock *uses* a windowing system?
Just about all of us. When's the last time you saw Unix -- on a
VT-100 connected to a VAX? ;-)
To put it in perspective: I'm writing this message from the comfort of
home on an SS-2 running OpenWindows under Solaris 2.5. I'm connected
to the Sun Wide-Area Network via a secure ISDN line and Firewall-1.
I happened to fire up this news reader during a remote login to the
SS-20 in my Menlo Park office. The news reader is getting its input
from a server in Milpitas, also protected by Firewall-1, and sending
its output to my SS-20's X server, which is xhost'ing the display from
there, to the Building 17 router, to the Menlo Park backbone, to the
Internet firewall, over ISDN, and into my workstation at home, where
the local X server decodes the incoming XWIN protocol into a pretty
display that you'd swear must be a local news reader. Point-and-click,
drag-and-drop, everything works as you'd expect it to. It's all done
with Unix and it's all totally transparent. Try *that* with Windows.
Jeff Bonwick
Solaris Performance
Dan Stromberg
>Bonwick) wrote:
>>
>> You're confusing 'grafted-on' with 'well-modularized'. SunView was
>> built into the SunOS 3.x kernel many years ago -- it was an *advance*
>> to move the window system out of the kernel. If you want examples
>> of grafted-on, how about the way Windows converts long filenames into
>> *visible* underlying 8.3 filenames? Or its use of the "f:" and "g:"
>> "network disks" as a network programming model?
>>
>
>letters, you address a file as "\\server\scsidrive\myfolder\myfile.doc" which I think you'll
>agree is a vast improvement. This works on Win95 and of course NT. I use it all the time; I
>avoid mapping drive letters unless I'm installing 16-bit software that expects it.
It's good they've finally got those silly drive letters on the way
out.
Then again, I've seen people do this with an automounter before.
Of course, it's kind of a short-sighted way to -use- an automounter,
at least unless you also get a CNAME for your server (but if you
aren't careful, people will ignore the CNAME). Fortunately, an
automounter will allow you to use other naming schemes - without such
potential problems. And not surprisingly, automounting is
substantially more universal, than what I see quoted above.
Trivia: Who was it that said "Those who do not understand Unix are
condemned to reinvent it, poorly."? It's surprising how often one
sees this re-affirmed.
Smoke Crack and Worship Satan
Smoke Crack and Worship Satan
Bonwick) wrote:
>
> You're confusing 'grafted-on' with 'well-modularized'. SunView was
> built into the SunOS 3.x kernel many years ago -- it was an *advance*
> to move the window system out of the kernel. If you want examples
> of grafted-on, how about the way Windows converts long filenames into
> *visible* underlying 8.3 filenames? Or its use of the "f:" and "g:"
> "network disks" as a network programming model?
>
Well, Windows 95 logoed apps support Universal Naming Convention. Rather than mapping drive
letters, you address a file as "\\server\scsidrive\myfolder\myfile.doc" which I think you'll
agree is a vast improvement. This works on Win95 and of course NT. I use it all the time; I
avoid mapping drive letters unless I'm installing 16-bit software that expects it.
>
> To put it in perspective: I'm writing this message from the comfort of
> home on an SS-2 running OpenWindows under Solaris 2.5. I'm connected
> to the Sun Wide-Area Network via a secure ISDN line and Firewall-1.
> I happened to fire up this news reader during a remote login to the
> SS-20 in my Menlo Park office. The news reader is getting its input
> from a server in Milpitas, also protected by Firewall-1, and sending
> its output to my SS-20's X server, which is xhost'ing the display from
> there, to the Building 17 router, to the Menlo Park backbone, to the
> Internet firewall, over ISDN, and into my workstation at home, where
> the local X server decodes the incoming XWIN protocol into a pretty
> display that you'd swear must be a local news reader. Point-and-click,
> drag-and-drop, everything works as you'd expect it to. It's all done
> with Unix and it's all totally transparent. Try *that* with Windows.
>
Remote logins with screen display can be done over ISDN, network, & modem with PcAnywhere32
for NT and 95. NT and 95 support TCP/IP and firewalls transparently.
Smoke Crack and Worship Satan
Chapin) wrote:
>
> I wouldn't put it quite that way. It would be accurate to say that
> Microsoft is almost single-handedly for the state of PC Computing
> today. However, when you realize that:
>
> (a) MS-DOS is a non-reentrant program monitor, i.e. 1960's technology
> appearing in the 1980's,
> (b) Windows NT is VMS all over again, i.e. 1970's technology
> appearing in the 1990's, and
> (c) Windows 95 is knock-off of the '87 Mac OS, i.e. 1980's
> technology in the 1990's,
>
> you'll notice a trend. To wit, Microsoft is consistently delivering
> out-of-date technology to users. Their OS products lack innovation,
> but they rely on the ignorance of their users and their monopolistic
> marketing position (thanks to IBM) to keep the gravy train going.
>
Utterly ridiculous.
First of all, the consumer market naturally lags the high-end workstation
market. So bringing VMS and even a program monitor to the PC is hardly
"recycling old technology"-- it's permitting home users to take advantage of OS
features formerly reserved for workstations.
Second, The primary factor holding back "innovation" is end users-- you and I.
We demand that all of our apps, all based on old techology, run on the latest
operating system.
If that wasn't a requirement, we'd all have BeBoxes and NextStep! So your
argument is unrealistic.
It all boils down to the old chicken and egg problem: what do developers write
software for? The OS that permits them to make the most sales. What OS do people
buy? The one with the most software.
Robert Krawitz
In article <48gv6k$h...@engnews2.Eng.Sun.COM>, bon...@cathy.eng.sun.com (Jeff
Bonwick) wrote:
> Eh? Unix has had window systems (GUIs) since before Microsoft existed.
> And because the window system is itself just another application in the
> Unix world, you can run any window system or any app you want. You can
> even run MS Windows/x86 binaries on Solaris/SPARC systems via Wabi.
> So with MS I get exactly one window system and a complete joke of a CLI.
> With Unix I get whatever window system I want and a really powerful CLI.
Of course UNIX has grafted-on windows systems. But what self
respecting UNIX jock *uses* a windowing system? People who use UNIX
learn to take advantage of the CLI-- that's how you get the gritty
stuff done. It's rather like strapping Windows, desqview, and
Norton Commander on top of DOS IMHO.
I consider myself a self respecting Unix jock. I've hacked the
kernel, etc. And I do use X. Granted, I basically use it to give me
more command line interfaces on a single screen. Unfortunately, AIX
(with SMIT) and Solaris (with admintool) make it unnecessarily
unpleasant to use the CLI to accomplish tasks that on a normal Unix
system are a piece of cake -- edit the system files directly.
UNIX will always have a life as a server and expensive workstation
OS. The question I would pose is, "what can the flavors of UNIX do
that NT can't"? In those situations, people should use
UNIX. Otherwise, NT is a hell of a lot easier.
Unix isn't also burdened (performance-wise) with all that flashy
glitter. Also, if you know what you're doing, it's probably faster to
do things in CLI mode than it is filling out dialog boxes and so
forth.
--
Robert Krawitz <r...@tiac.net> http://www.tiac.net/users/rlk/
Member of the League for Programming Freedom -- mail l...@uunet.uu.net
Tall Clubs International -- tci-r...@aptinc.com or 1-800-521-2512
J.J.Bailey
-Anyone really want to bet that everything you can do on a Unix box
-today and tomorrow won't be able to be done on an NT server in 2 or 3
-years? I wouldn't take that bet. And don't tell me about security on a
-Unix box vs. NT and the superiority of Unix. I would agree. Unix is a
-superior operating system to NT. Right now. And even in 2 or 3 years
-it probably will be. But so what? Unix is not what sells. Microsoft is
-what sells. As sad as that may be.
This scenario is incomplete. You forgot to add the part explaining how
UNIX development stands still for "2 or 3 years" waiting for Micro$oft to
catch up.
--
J.J.Bailey
Consultant
j...@bcc.com
Mr. Who
schm...@adsnet.net (Chris Shmidtc) wrote:
>m...@aplace.com (Smoke Crack and Worship Satan) wrote:
>
1.
>In 2 or 3 years, the next big release of Windows For Everyone will
>essentially be NT Workstation with a Win95 front end, available now
>for all the NT users anyway. You cannot tell me that there will be
>more Unix users in 1997 than there will be Windows NT users. By 1997
>or '98 all Win95 users will be NT users.
2.
>It's just a short jump from there for the bean counters and suits to
>say "Let's get rid of Novell and our Unix boxes. Everything those
>operating systems do can be under NT anyway and it's cheaper to buy,
>maintain, and operate and it will all be standardized."
>
3.
>Anyone really want to bet that everything you can do on a Unix box
>today and tomorrow won't be able to be done on an NT server in 2 or 3
>years? I wouldn't take that bet. And don't tell me about security on a
>Unix box vs. NT and the superiority of Unix. I would agree. Unix is a
>superior operating system to NT. Right now. And even in 2 or 3 years
>it probably will be. But so what? Unix is not what sells. Microsoft is
>what sells. As sad as that may be.
>Chris
I have a few questions/responses on Chris' post and have
labelled Chris' main points 1, 2 and 3 for reference.
1. How are you defining a "Unix user" vs. an "NT user" ??? Are
you counting the number of copies of each sold ??? Most copies
of Unix are sold on multi-user platforms that support 5-1000
users. So if you're counting 1 copy for a machine that services
multiple "Unix users" you aren't counting them the same way. If
you add up all the users on Unix systems vs. NT in this fashion I
would bet that they come up pretty even. Also, you're assuming
that Microsoft will be on time with Cairo (or is it Memphis) --
in spite of the fact that MS is chronically late with their releases.
Win95 was released > 1 year behind schedule. You're target of 1997
or 1998 is a bit optimistic IMO.
2. Most "bean counters" and "suits" defer the judgement to their
technical people. You are talking about a COMPLETE refitting
of a company's IS infrastructure -- something that won't be done
in a corporate environment lightly. Remember the mainframe was
supposed to be dead a few years ago ! It is possible that NT will
replace Novell given Novell's current state of affairs.
3. Are the big names of commercial databases (Oracle, Informix, Sybase
etc.) available for NT ? If not are they under development ? I
don't know ? Does NT scale the way Unix does -- IMHO it does not.
Intel (oops until) an NT system can provide the SAME or BETTER
performance on a large commercial database in a heavy transaction
processing environment NT will not replace Unix systems. I don't
think that will be 1997/8.
--
+---------------------------------------+
| Mr. Who is: |
| jo...@sigg.com (Business) |
| jkot...@interaccess.com (Personal) |
| |
| Geek Code: under construction & too |
| busy surfing to calculate it !!! |
+---------------------------------------+
Chris Shmidtc
This is an extremely good point. Windows 95, even according to
Microsoft, is sort of a stop-gap, interim operating system, designed
to allow the remaining DOS users (especially) and previous Windows
users to migrate to the beginnings of a 32-bit OS, where DOS is seen
less and less.
In 2 or 3 years, the next big release of Windows For Everyone will
essentially be NT Workstation with a Win95 front end, available now
for all the NT users anyway. You cannot tell me that there will be
more Unix users in 1997 than there will be Windows NT users. By 1997
or '98 all Win95 users will be NT users.
It's just a short jump from there for the bean counters and suits to
say "Let's get rid of Novell and our Unix boxes. Everything those
operating systems do can be under NT anyway and it's cheaper to buy,
maintain, and operate and it will all be standardized."
Anyone really want to bet that everything you can do on a Unix box
today and tomorrow won't be able to be done on an NT server in 2 or 3
years? I wouldn't take that bet. And don't tell me about security on a
Unix box vs. NT and the superiority of Unix. I would agree. Unix is a
superior operating system to NT. Right now. And even in 2 or 3 years
it probably will be. But so what? Unix is not what sells. Microsoft is
what sells. As sad as that may be.
I use both operating systems (actually, throw in MacOS, also). I just
don't know but I'm not willing to bet against the tide right now.
Chris
schm...@adsnet.net
509-962-4638
Jan Ekholm INF
:
: In article <48gnn1$s...@nntp.crl.com>, m...@aplace.com (Smoke Crack and Worship Satan) writes:
: > In article <48gdgn$d...@engnews2.Eng.Sun.COM>, bon...@cathy.eng.sun.com (Jeff
: > Bonwick) wrote:
: > And there will be more machines running NT than UNIX by 1997 IMHO.
:
: I am resisting the urge to make a very large bet with you... I
: will instead just settle for laughing uproariously.
Yeah, Linux will probably spread faster than NT ever could imagine of...
--
Jan 'Chakie' Ekholm \ CS at Abo Akademi University, Turku, Finland
jek...@mail.abo.fi +-----------------------------------------------
cha...@mousehouse.abo.fi / 'What u have to learn to do, u learn by doing'
http://www.abo.fi/~jekholm\ Linux Inside since fall '94.
Jeff Bonwick
> Remote logins with screen display can be done over ISDN, network,
> & modem with PcAnywhere32 for NT and 95.
And that means that NT "supports" remote login? Hmmm... the table
beneath my workstation is "supporting" networking as we speak.
Therefore, I claim that the table is "internet-ready".
You're right that many of the deficiencies of Windows have been addressed
by third-party products, and in many cases that's a reasonable solution.
But in the networked world it doesn't work very well because you want to
be able to assume that any machine you're talking to has some minimal set
of capabilities. If those capabilities aren't part of the base operating
environment, you just can't do that. It's no big deal if you're running
a half-dozen machines in a dentist-office LAN. It becomes a very big deal
when you have to administer a large site.
This reminds me of the claim that "A Windows PC can do almost anything
a Unix workstation can." All you have to do is buy a top-of-the-line
P6 processor, lots of memory, a high-end networking card, a high-end
disk controller, some fast disks, a whizzy graphics card, a high-res
24-bit color monitor, a bunch of expensive networking software and
drivers, some LAN and WAN administration utilities, backup software,
performance monitoring tools, ftp, telnet, the Windows versions of
the Unix shell and popular utilities from /usr/bin, and voila!
Half a workstation for twice the price!
> NT and 95 support TCP/IP and firewalls transparently.
I'm talking about *being* the firewall. How secure would you feel
knowing that the only thing between your corporate network and the
hackers of the world was Win95 or NT?
Jeff Bonwick
Solaris Performance
Jaime Lares Rebelo Pinto
: You mean non VisiCalc spreadsheet. Lotus is just another
: non-innovative copy cat. Like Apple, IBM, an everybody
: else who didn't breast feed at the PARC.
Apple came up with lots of innovations along this years, denying
that is absurd. I believe that Byte published a long list of Apple
innovations recently, but I didn't read it.
People from Apple went to PARC because a debate was going on inside
Apple about investing or not investing in the GUI. It seems that Jobs
didn't consider it a good idea, but Runskin (bad spelling, sorry)
considered it a good idea and had already done some research on the
subject. It wasn't possible for the people in favor of the GUI to
get something working fast, but Runskin new that research on GUIs
was being done at PARC, so they went there.
Jobs completely changed is mind and Apple invested in the GUI.
They were a bit influenced by what they saw, but Apple paid to XEROX
for that. However, they had their own ideas, and came up with key
ideas for MacOS. I did read several articles by people that saw
both products, and all people agreed that the products were considerably
different, and general opinions I read were in favor of the MacOS
GUI.
This is not very surprising, because people form Apple went to
PARC several years before the project at PARC was finished, and
there were already ideas about the subject inside Apple.
Apple didn't ignore the technology until the key ideas were
established and the worth of the technology in the market
was impossible to deny. They were pioneers, like the people at
PARC, but they developed products that proved the value of the
technology. Only after that some other companies invested in it...
Cheers,
Jaime.
Chris Shmidtc
>This scenario is incomplete. You forgot to add the part explaining how
>UNIX development stands still for "2 or 3 years" waiting for Micro$oft to
>catch up.
Really. Well name some goals that Unix "wants" to accomplish in the
next 2 or three years. And yet, Microsoft's goal is to make the 1997
version of NT be nearly as complete and functional as the 1997 version
of Unix. And that's really beside the point.
NT is what is going to sell. Win95 users are not going to be jumping
boat to go to some flavor of Unix. They'll want something that looks
and acts like something they're familiar with. NT97. It really doesn't
matter if NT97 has caught up with Unix97. In 1997 the perception will
be that it has caught up. Microsoft markets. That's what they do best.
Actually, what they do best is market someone else's technology in a
new wrapper. In they next two or three years, Microsoft will spend
billions developing and marketing NT. No one will come close to that
for any other operating system with the exception of the Macintosh.
It's hard to fight billions of development dollars and millions of
development hours.
And I'll be willing to bet that any version of NT in 1997 will be
pretty close in functionality, security and performance to any version
of Unix that year. Billions of dollars can buy a lot of performance.
It can also buy a lot of perception.
Chris
schm...@adsnet.net
509-962-4638
David Bellamy
>
> > > > * first non-UNIX mainstream OS to support SMP -- Windows NT
> > >
> > > First non-Chrysler, non-GM car company to make air bags standard: Ford
> Motors!
>
> In article <DI72M...@ftel.co.uk>, I.G.B...@ftel.co.uk (Ian G Batten) wrote:
> > Not merely is claiming things as new because you're the first
> > implementor aside from the Unix crowd ludicrous, on this occasion it's
> > arguably wrong anyway. Depending upon how S you think the MP in Multics
> > was, of course.
>
> Don't forget IBM MVS, IBM VM, Sperry Univac, Digital VMS, Apollo Domain,
> Honeywell, Control Data Corporation, NCR, Burroughs.
>
Not to mention some of the innovations from the otherside of the
Atlantic, ie ICL's VME (mid 70's) and I'm sure that some of the Georges
on the 1900 series (late 60s early 70s) had both VM and SMP - although
no doubt called something else then. And the Atlas (late 50s ?) made by
one of ICLs forerunners (ICT ??) had paged memory.
Anhow, I think this thread is moving more to comp.history.
> First OS to support SMP other than those OS's that preceded it - Windows NT.
--
David Bellamy.
Deparment of Commerce. The University of Queensland. Australia.
Internet: bel...@commerce.uq.edu.au
Phone: +61 7 336 56652
Achim Gratz
Smoke> Right, but the very fact that many of these technologies
Smoke> are *available* on the PC is impressive. Ever tried to do
Smoke> SMP on a Commodore 64? You can buy dual processor
Smoke> Pentium-133s *today* from Micron, for God's sake!
Yeah, I did: I computed all my Mandelbrot Sets in the 1541 floppy
drive, because the processor was twice as fast as the one in the
computer. I also had a good chunk of an ISAM file system directly
loaded and run within the floppy drive (only limited by the 2kByte
memory).
--
Achim Gratz
-+==##{([***Murphy is always right***])}##==+-
E-Mail: gr...@ite.inf.tu-dresden.de
Phone: +49 351 4575 - 325
Aaron Bilger
...
>You guys are living in a dream world. Major corporations
>who spend zillions of dollars just to figure out there
>logo are switching to NT. UNIX is an old tired OS that
>has served us well, but it is time to bury it. As desktops
>become more and more like mini's, clustering, and all that
>distributed stuff, NT is poised and has major support, but
>time will tell and its a strange world. I see that Taligent
>is being dumped.
When Microsoft provides source code with NT, then and only then
could I even consider recommending it over UNIX.
Farewell-
Aaron
Anthony D. Tribelli
: Yeah, Linux will probably spread faster than NT ever could imagine of...
Within the Unix/hacker community, absolutely.
Among programmers, maybe, but I expect most will dual-boot so they can
have both.
In the rest of the computing world, probably not. Linux is merely another
implementation of UNIX. The one with best price-performance ratio, but it
is limitted to those who want/need Unix. I think the general purpose
computing world has already voted on whether or not they want Unix.
Tony
--
------------------
Tony Tribelli
adtri...@acm.org
Gary Ferguson
>
>What about OS/2 SMP? That's been around about as long as NT if not longer...
>
Doesn't count 'cause it never really worked...
Gary
Anthony D. Tribelli
: What is happening right now under our noses:
: 1) Unix companies are pushing Win3 users to switch to win95 and NT
: by making browsers/languages that run under win95 and NT but not under
: win3.x (I should not have to spell this out. Perceptively
: challenged need not understand.)
Actually, Windows companies are pushing users to a Win32 environment.
Developing Win32 apps is much less painfull than developing Win16 apps.
BTW, I think you incorrectly characterize the "browser/language"
companies at "Unix" companies, they are more accuaractely "Net"
companies. Unix was merely the first of many targets.
: 2) Thus making the Windows programs portable to Unix.
: 3) Creating libraries to make all Windows 95/NT programs
: run natively under Unix (once compiled)
These libraries can add hundreds of dollars to the cost of a single unit
after all the licenses are paid. For example product X for Win32: $200,
for Unix: $400. How does this make Unix look attractive? These libraries
merely let you sell a few more programs to people who happen to have a
workstation rather than a PC.
: 4) Making free/cheap products (such as Netscape) to include
: things like Word/Excel/... thus eating into Microsoft
: market share.
Congradulation, you've come up with a truly amazing way to screw up Netscape.
: 5) MakeServers/Browsers provide things like Encarta etc....
: to eat into that market share by Microsoft.
Unless you are subsidized by a University :-), connect time costs people
money. CDs will probably remain more popular unless the online provider
can offer vastly better content, IMHO.
Sinan Karasu
Anthony D. Tribelli <a...@netcom.com> wrote:
>Sinan Karasu (si...@u.washington.edu) wrote:
>: 4) Making free/cheap products (such as Netscape) to include
>: things like Word/Excel/... thus eating into Microsoft
>: market share.
>
>Congradulation, you've come up with a truly amazing way to screw up Netscape.
Ooops. You are right.... Sorry...
Sinan
David B. Schmidt
[Damn near everything snipped]
>"Let's get rid of Novell and our Unix boxes. Everything those
>operating systems do can be under NT anyway and it's cheaper to buy,
>maintain, and operate and it will all be standardized."
Huh ? First - I don't know of anyone who runs Novell overtop (in lieu
of?) UNIX TCP/IP...
But what I respond to is ... Don't forget to count those of us that are
tired of MS, and their constant line of crap, and are heading back to
once we came - UNIX and building UNIX based apps.
Never should have left,
-David <dsch...@shadow.net>
Mark Davis
>Anyone really want to bet that everything you can do on a Unix box
>today and tomorrow won't be able to be done on an NT server in 2 or 3
>years? I wouldn't take that bet. And don't tell me about security on a
>Unix box vs. NT and the superiority of Unix. I would agree. Unix is a
>superior operating system to NT. Right now. And even in 2 or 3 years
>it probably will be. But so what? Unix is not what sells. Microsoft is
>what sells. As sad as that may be.
Unix is a moving target. It is *NOT* a static thing which has remained the
same for 25 years. In 2 or 3 years, it will still be ahead of Microsoft-
as long as development continues. SCO will have the 64 bit P7 Unix released by
then, for example.
--
/--------------------------------------------------------------------------\
| Mark A. Davis | Lake Taylor Hospital | Norfolk,VA (804)-461-5001x431 |
| Director/SysAdmin | Information Systems | ma...@taylor.infi.net |
\--------------------------------------------------------------------------/
Stanley T.H. Chow
Achim Gratz <gr...@ite.inf.tu-dresden.de> wrote:
>>>>>> "Smoke" == Smoke Crack and Worship Satan <m...@aplace.com> writes:
>
> Smoke> Right, but the very fact that many of these technologies
> Smoke> are *available* on the PC is impressive. Ever tried to do
> Smoke> SMP on a Commodore 64? You can buy dual processor
Symetric Multi-Processing
> Smoke> Pentium-133s *today* from Micron, for God's sake!
>
>Yeah, I did: I computed all my Mandelbrot Sets in the 1541 floppy
>drive, because the processor was twice as fast as the one in the
>computer. I also had a good chunk of an ISAM file system directly
>loaded and run within the floppy drive (only limited by the 2kByte
>memory).
Very clever hack; impressive; etc.
Never the less, not Symetric by a long leap of imgination. At msot,
I would call it very loosely coupled (considering the speed of the
link between the drive and the "real" CPU).
--
Stanley Chow; sc...@bnr.ca, stanley....@nt.com; (613) 763-2831
Bell Northern Research Ltd., PO Box 3511 Station C, Ottawa, Ontario
Me? Represent other people? Don't make them laugh so hard.
Neal P. Murphy
>m...@aplace.com (Smoke Crack and Worship Satan)
>>Cute, but you miss the point. UNIX will never be a mainstream operating system.
>>Would you like to walk your grandmother through installing Linux on her PC? And
>>installing X Windows? And purchasing shrink wrapped applications for her UNIX
>>based PC?
>I sure wouldn't like to show my grandma how to install any MS
>windows application. Having to run to every desktop with 20
>floppies and install each one. And even with CDROM installs,
.
. [fine oration editted for brevity]
.
>In the regional sites, we only have some PC techs who would change
>tapes or power cycle hosed machines for us. I cannot imagine
>having to support it if they were all PC's.
Well said!
Fest3er
--
The Microsoft thirty bladed swiss army knife with corkscrew and magnifying lens
[is] inappropriate for all of my word processing needs.
- Stephanie Evans
Edmond
>>When Microsoft provides source code with NT, then and only then
>>could I even consider recommending it over UNIX.
>
>typical NT system, requires fewer hardware resources, and is more stable. I'll
>take Linux over NT (and especially Winsux 95) any day. I can do a file server
>on a 386 machine running 8M with a 200M hard disk and keep everyone on my
>network happy. Try that with Gates Inc.
>
More stable than NT? Your kidding yourself Steve. I've used both long
enough to know this one. Linux is nice as a cheap UNIX workstation, but
no one would put there business up on it. Not enough support nor enough
critical applications.
--------------------------------------------------------------------
Edmond Underwood
Computing & Network Services (University of Colorado)
E-mail: unde...@Colorado.Edu
Ho Yoon
Jeff Bonwick <bon...@cathy.eng.sun.com> wrote:
>m...@aplace.com (Smoke Crack and Worship Satan) writes:
>
>And that means that NT "supports" remote login? Hmmm... the table
>beneath my workstation is "supporting" networking as we speak.
>Therefore, I claim that the table is "internet-ready".
>
>a Unix workstation can." All you have to do is buy a top-of-the-line
>P6 processor, lots of memory, a high-end networking card, a high-end
>disk controller, some fast disks, a whizzy graphics card, a high-res
>24-bit color monitor, a bunch of expensive networking software and
>drivers, some LAN and WAN administration utilities, backup software,
>performance monitoring tools, ftp, telnet, the Windows versions of
>the Unix shell and popular utilities from /usr/bin, and voila!
>Half a workstation for twice the price!
>
>> NT and 95 support TCP/IP and firewalls transparently.
>
>I'm talking about *being* the firewall. How secure would you feel
>knowing that the only thing between your corporate network and the
>hackers of the world was Win95 or NT?
My understanding was that NT, or for that matter, most desktop OSes,
didn't have the security holes which Unix had. And these security holes
related to the remote login protocols?
Christopher L. Estep
> I am resisting the urge to make a very large bet with you... I
>will instead just settle for laughing uproariously.
>
I should take the NT end, and spot you the entire SPARC market. Face it;
when CHRP PowerMacs begin shipping, most of them will NOT be running MacOS,
but NT. Why? For the SAME reason that Windows 95 has been kicking Warp all
over the planet: applications, applications, applications. NT ALREADY
outsells native versions of UNIX for all platforms except MIPS; is SGI gonna
sell THAT many SGI boxes?
Roger Lin
: In comparing NT copy count and Unix copy count, keep in mind that
: it is NT Server that hit Ingram Micro's top 10 list recently;
: it sells with 20 client access; many sites using NT Server put dozens
: or hundreds of users on per server. The number of NT-supported users
: is large and growing rapidly.
: Also the growth curve is steep upward for NT Workstation as well as
: Server, while at many sites the usage curve is downward for Unix.
: At my site:
: system usage 2 years ago usage now usage next year
: NTWS 2 occasional 2 occasional 15+
: NT Server 0 3 test 40
: Unix several 1 occasional 1 rare
: VMS hundreds ~20 ~10
: Novell 0 0 0
Total hundreds 20s 60s
either your figure is wrong or more than 80 people in your company stopped
using the computer recently and about 40 of them plan to start using again
next year.
Roger
Ken Kriesel, Physical Sciences Lab, UW-Madison
>
> How are you counting the copies of NT vs. copies of the various and
> sundry unix variants ??? Copies of OS sold or users supported by
> that 1 copy. I might agree with you that NT may have sold more
> copies than Unix ... might. But, I am certain that the number of
> users SUPPORTED by Unix vs. NT is a 10:1 ratio.
>
> Keep in mind the group I work with supports ~ 2500 Unix Users
> on 20 Copies of Unix. Number of NT stations: 0.
>
> You have 1 point right --- Win95/NT has desktop productivity
> applications on their side.
>
> So, while you are working on your spreadsheet (the 20th revison)
> I'll just get back to work building my network, WITH UNIX.
In comparing NT copy count and Unix copy count, keep in mind that
it is NT Server that hit Ingram Micro's top 10 list recently;
it sells with 20 client access; many sites using NT Server put dozens
or hundreds of users on per server. The number of NT-supported users
is large and growing rapidly.
Also the growth curve is steep upward for NT Workstation as well as
Server, while at many sites the usage curve is downward for Unix.
At my site:
system usage 2 years ago usage now usage next year
NTWS 2 occasional 2 occasional 15+
NT Server 0 3 test 40
Unix several 1 occasional 1 rare
VMS hundreds ~20 ~10
Novell 0 0 0
And finally, NTWS can also act as a file and print (and other) server
for 10 simultaneous users other than at the console. So do you count
it as 1 or 11 users?
NT is young and so has some shortcomings, which I expect to be resolved
with time as it matures. It already has the user base to fund this.
Ken
Ken Kriesel, Physical Sciences Lab, UW-Madison
>
> Total hundreds 20s 60s
>
> either your figure is wrong or more than 80 people in your company stopped
> using the computer recently and about 40 of them plan to start using again
> next year.
Yes the numbers do look a little odd, and probably contain some
errors. The trends however are right: VMS fading, Unix small and fading,
NT ahead of Unix already and sharply on the rise.
We evaluated buying Unix workstations for engineers and drafters but
concluded we could not cost-justify the Unix applications and higher-
priced hardware.
Subsequently, we reevaluated and determined that the real cost factor
was how complex an environment we were supporting. Reducing OS's will
help simplify and cost-reduce.
1) The 20s is too low. I think I left out the DOS & Win3.1 row.
Which will be replaced with Win95 & NT, on the same inexpensive
hardware.
2) A timeshare service on VMS went away, and in the process numerous
users closed accounts. Since it is no longer part of my department,
it is no longer counted. Maybe half quit for other platforms; some
went to PCs or MACs, while some went to Unix or their own VMS systems.
I don't have detail on that.
2) The figures were approximate, not exact counts.
3) Some individuals are rightly counted in more than one OS at a
given time, and usage patterns change with time.
4) Employment levels fluctuate (Minor effect)
Ken
John DiMarco
>bon...@cathy.eng.sun.com (Jeff Bonwick) wrote:
>> The various flavors of Unix are among the most secure operating systems
>> you can get because (1) Unix was designed to be secure, and (2) hackers
>> all over the globe have spent a couple of decades trying to break in,
>> so most of the bugs have been found and fixed.
>This is a good joke. Teenagers are breaking into phone company
>computers. Unix is still the number one topic of CERT
>advisories, by a wide margin. Many bugs are still being found and
>fixed. Security bugs or risks in other OS's often are introduced
>by the act of adding a Unix-type utility to the non-Unix OS. It
>seems to me that Unix is mentioned in more advisories, than all other
>OS's combined. Sometimes a bug is found in one Unix flavor, and
>then detected in numerous other flavors, but shown to not exist in
>some others, while additional flavors have not been tested when the
>advisory comes out. The tools (Satan and Courtney, etc) were written
>for Unix.
While I think Jeff's original assertion is a bit misleading, the mere fact
that most CERT advisories and security bulletins address UNIX is a consequence
of the fact that most of the server systems on the Internet run UNIX of one
flavour or another. There just aren't a whole lot of VMS, NT, OS/400, etc.
servers out there in comparison.
After all, bank robbers rob banks because "that's where the money is", not
because banks are more inherently insecure than other commercial buildings.
Regards,
John
--
John DiMarco <j...@cdf.toronto.edu> Office: EA201B
Computing Disciplines Facility Systems Manager Phone: 416-978-1928
University of Toronto Fax: 416-978-1931
http://www.cdf.toronto.edu/~jdd
Jay Martin
>The various flavors of Unix are among the most secure operating systems
>you can get because (1) Unix was designed to be secure, and (2) hackers
>all over the globe have spent a couple of decades trying to break in,
>so most of the bugs have been found and fixed.
What? Unix security has always been a widely recognized joke (ever
heard of the internet worm?) which is in my humble opinion
due to incompetent (hacked) design.
Andrew Welch
> turned off. That's more than 23% of ALL Web-surfers.
>
> So if you rely on Netscape tags and images to make your site, then only
> half of all web surfers can read your "standard" page.
How does "surf[ing] with images turned off" in any way negate the "Netscape-standard" html tags? Tables, etc. seem to work fine with images turned off...
+--------------------------------------------------------------+
| Andrew Welch - Thaumaturgist - Ambrosia Software, Inc. |
+-------------------------------+------------------------------+
| AOL-> Keyword: Ambrosia | eWorld-> Shortcut: Ambrosia |
| CIS-> GO word: Ambrosia | http://www.AmbrosiaSW.com/ |
+-------------------------------+------------------------------+
Andrew Welch
Jan Vorbrueggen
Satan) writes:
First of all, the consumer market naturally lags the high-end workstation
market.
Why "naturally"? Oh, in performance, that used to be the case. But in
functionality? Nope, that wasn;t even necessary at DOS 1.0 time, and certainly
not in 1990.
So bringing VMS and even a program monitor to the PC is hardly "recycling
old technology"-- it's permitting home users to take advantage of OS
features formerly reserved for workstations.
Duh. It's still recycling. Why should an OS feature be "reserved" to a
workstation? The first versions of VMS, with much more functionality then than
DOS has now, ran on a machine with 256 kB --- yes, 256 _kilobytes_ --- of main
memory. It's been a number of years now that an average PC has that amount of
cache. No, the reason is a bad technical decision by IBM that managed to
perpetuate itself; that's what you get from the chaotic, self-organizing
principles of capitalism.
Jan
Ian G Batten
Dan Stromberg <stro...@bingy.acs.uci.edu> wrote:
> Trivia: Who was it that said "Those who do not understand Unix are
> condemned to reinvent it, poorly."? It's surprising how often one
Henry Spencer.
ian
Steve Martin
Jeff Bonwick
>
> What? Unix security has always been a widely recognized joke
> (ever heard of the internet worm?)
Yes I have, and FYI, it didn't get through Sun's Firewall-1.
> which is in my humble opinion due to incompetent (hacked) design.
The design is fine; there are no architectural holes in Unix security.
There are bugs in the implementations, which are of varying quality.
Anyone who wants to be around in 2-3 years takes them quite seriously.
Jeff Bonwick
Solaris Performance
Ken Kriesel, Physical Sciences Lab, UW-Madison
>
> In article <m#bs...@quack.kfu.com> h...@quack.kfu.com (Ho Yoon) writes:
> >
> > My understanding was that NT, or for that matter, most desktop OSes,
> > didn't have the security holes which Unix had.
>
accessible by Samba client, for example. The WFWG problem was rather
old before it was discovered along with Win95's. Win95 logon security
at the keyboard is very weak.
> There are two components of security: architectural (was the system
> designed to be secure?) and practical (are there bugs you can exploit?).
>
> Security wasn't considered an issue for non-networked single-user PCs,
> so most of the early PC operating systems didn't have any. All of the
> "anti-virus" software you see now is an attempt to compensate for this.
> DOS and Windows 3.1, for example, are by design not secure.
Right. DOS & Win3.1 security is nonexistent. It was not a design
parameter or marketing point at the time they were being created.
Win95 also had security as a low priority in design considerations.
NT on the other hand was designed to have security as a selling point.
> The various flavors of Unix are among the most secure operating systems
> you can get because (1) Unix was designed to be secure, and (2) hackers
> all over the globe have spent a couple of decades trying to break in,
> so most of the bugs have been found and fixed.
This is a good joke. Teenagers are breaking into phone company
computers. Unix is still the number one topic of CERT
advisories, by a wide margin. Many bugs are still being found and
fixed. Security bugs or risks in other OS's often are introduced
by the act of adding a Unix-type utility to the non-Unix OS. It
seems to me that Unix is mentioned in more advisories, than all other
OS's combined. Sometimes a bug is found in one Unix flavor, and
then detected in numerous other flavors, but shown to not exist in
some others, while additional flavors have not been tested when the
advisory comes out. The tools (Satan and Courtney, etc) were written
for Unix.
> NT is designed to be secure*, but it's difficult to assess the practical
> side. Breaking into NT systems just hasn't become cool yet, probably
> for the simple reason that NT systems aren't protecting exciting stuff.
> Breaking into universities, banks and Pentagon labs is interesting;
> breaking into Bob's spreadsheet, not so much.
Sounds like you're saying we don't know how secure NT is because
noone's broken it yet, at least that we know of. (I've posted once or
twice asking for such anecdotes, but haven't heard of any. Perhaps
it's just lack of response.)
You don't think it would be exciting for the talented but ethically
impaired to break into Microsoft (which
runs NTWS on numerous processor types, as well as NT Server)? How
about some major corporations, which build computers for NT (DEC,
Intergraph, Micron, etc) or use NT Server where others might use VMS,
Unix, or Novell? How about a major university that might
be doing computer security research on multiple OS's?
NT has gotten C2 certification (without network connection I think).
Security is a relative term, so whether you consider that "secure"
depends on what security level is adequate for your purposes, and
what your alternatives are.
> * I've read that NT's support for older apps constitutes an architectural
> security hole because these apps have to be able to access physical
> resources directly. It sounds plausible, but does anyone out there
> know for sure? I'd think MS could get around this by running these
> apps in an isolated virtual machine (though it might kill performance).
This is incorrect referring to NT (but correct referring to Win95).
Win95 is not very secure because it supports basically everything
already installed on a DOS machine. That was a design goal.
There are many ways in which apps access the hardware.
Programs which use security dongles on the LPT port need new code,
precisely because NT will not allow them to access it directly, so
NT's protection of the physical resources is pretty effective for old
apps. Some communications apps also have to be upgraded because they
try to access the serial ports directly. NT owns all the hardware
resources and protects them at the expense of compatibility with
"ill-behaved" old applications.
DOS-extended apps do not in general run in NT, but require an upgrade
to do so; NT controls memory, but the DOS-extended apps try and fail
to control memory themselves.
Programs which do legal DOS manipulations of files can get at any file
on a FAT volume in NT. The FAT file structure does not include the
concept of ownership; it was designed when only one person used a PC
and he had full privileges. An NT administrator has the choice of
using FAT and taking a chance, or of using NTFS and regaining disk
space on large volumes as well as some protection of one user's files
from the actions of another user and his programs, at the cost of
higher administration effort.
NT3.51 does provide (as a property of a program item, see Program Manager,
File, Properties) the option to "Run in Separate Memory Space".
There are also other tools for security, either in the base license or
the NT resource kit. Get the latest version, 3.51a.
NT's security can be much weakened by implementing a lot of Unix-style
utilities and particularly by improperly configuring them.
It can be strengthened by running C2CONFIG from the NT resource kit.
NT is stable enough that it can run several weeks continuously;
this PC is a 2 year old Gateway and has been up 46.5 days to date.
I don't recall why it was shut down last.
> One thing is certain: security is going to become an absolutely critical
> issue in the coming years. Historically most break-ins were motivated
> simply by the intellectual challenge: can I do it? But real Internet
> commerce will motivate real Internet crime. Operating systems that
> haven't been shaken down by the innocuous propeller-heads of the world
> may be in for a real baptism by fire.
Some people try to persuade themselves that small businesses or small
departments within a university really
don't need to pay attemtion to this sort of thing. That's very
dangerous; a couple of occurrences of losing your accounts payable
file, or a few man weeks or months lost due to damage from and
countermeasures for viruses, or a competitor knowing what you're
going to bid on a job he wants to get (because he copied your bid
spreadsheet after breaking in to your system) can sink a small business.
Regardless of which OS is involved, more people can lose their jobs
than just the fellow who decided security wasn't that important.
In some cases you'll never know what hit you.
Ken
Nik Simpson
Jeff Bonwick <bon...@cathy.eng.sun.com> wrote:
>In article <m#bs...@quack.kfu.com> h...@quack.kfu.com (Ho Yoon) writes:
>>
>
>* I've read that NT's support for older apps constitutes an architectural
> security hole because these apps have to be able to access physical
> resources directly. It sounds plausible, but does anyone out there
> know for sure? I'd think MS could get around this by running these
> apps in an isolated virtual machine (though it might kill performance).
No Win16 application is allowed to touch hardware in NT. If the Win16 app
relies on directly accessing hardware then it will not run on NT.
--
|--------------------------------------------------|
| Nik Simpson Mail : ndsi...@ingr.com |
| Intergraph Computer Systems |
|--------------------------------------------------|
Mark Davis
>[Damn near everything snipped]
>>"Let's get rid of Novell and our Unix boxes. Everything those
>>operating systems do can be under NT anyway and it's cheaper to buy,
>>maintain, and operate and it will all be standardized."
>Huh ? First - I don't know of anyone who runs Novell overtop (in lieu
>of?) UNIX TCP/IP...
Novell is a company...
Ling Wang
>You guys are living in a dream world. Major corporations
>who spend zillions of dollars just to figure out there
>logo are switching to NT. UNIX is an old tired OS that
>has served us well, but it is time to bury it. As desktops
>become more and more like mini's, clustering, and all that
>distributed stuff, NT is poised and has major support, but
>time will tell and its a strange world. I see that Taligent
>is being dumped.
You must be living in your dream world. I admit we have tons
of NT boxes in our company, but all the NT techs are running
around to individual desktops to troubleshoot problems. I am
sitting here administering 11 regions in 8 cities with remote
logins, xhosting etc. And I can't imagine anyone using servers
where one cannot dynamically reconfigure a network interface
without a reboot. We have tons of servers with failover backups,
which pretty much does a ifconfig on its network interface and
automatically takes over a failed primary server.
--
### ### +------------------------
# ###### #### | Ling Wang
## ### ########## | ICST Unix Admin
# ###### ## ### | Voice: (212)236-2183
# # ## ## #### | Fax: (212)236-3565
# ## #### | Email: lwang@.ml.com
### ### ## +------------------------
## ## ### #### ### | Merrill Lynch
## ## ## ### | Enterprise Services
### #### ## ### | World Financial Center
## ###### ########## ## | South Tower,14th Floor
# ## ## # # ## ### ## | New York, NY 10080-6114
# ## #### # ## # ## +------------------------
##### #### ####
Ling Wang
>>In the regional sites, we only have some PC techs who would change
>>tapes or power cycle hosed machines for us. I cannot imagine
>>having to support it if they were all PC's.
>Well said!
The reason I left my last job was because I support a ratio of
60% UNIX BOXES 40% PC's, but 90% of the problems I encountered
was PC related. I was getting tired of running around to users's
desktops because there are no remote system error logging,
remote access capabilities, remote boot, etc.
Mr. Who
sundry unix variants ??? Copies of OS sold or users supported by
that 1 copy. I might agree with you that NT may have sold more
copies than Unix ... might. But, I am certain that the number of
users SUPPORTED by Unix vs. NT is a 10:1 ratio.
Keep in mind the group I work with supports ~ 2500 Unix Users
on 20 Copies of Unix. Number of NT stations: 0.
You have 1 point right --- Win95/NT has desktop productivity
applications on their side.
So, while you are working on your spreadsheet (the 20th revison)
I'll just get back to work building my network, WITH UNIX.
--
+-------------------------------------------+
| Mr. Who is: |
| jo...@sigg.com (Business) |
| jkot...@interaccess.com (Personal) |
| |
| New: The Mr. Who home page |
| http://homepage.interaccess.com/~jkotches |
+-------------------------------------------+
Edward Jung
>In article <m#bs...@quack.kfu.com> h...@quack.kfu.com (Ho Yoon) writes:
>>
>> My understanding was that NT, or for that matter, most desktop OSes,
>> didn't have the security holes which Unix had.
>There are two components of security: architectural (was the system
>designed to be secure?) and practical (are there bugs you can exploit?).
For what it's worth on the *operating system* security side, certain
Unix implementations and Windows NT have passed US government B
certification for security. This is largely an architectural
verification and documentation issue required by the Feds.
Typical Unix architecture does not specify enough to pass B; many
implementors (including Sun) do extra work to make it so. Thus, some
specific Unix *implementations* have security architectures that are
equivalent (or superior) to NT's architecture, but "Unix" in the
standard, generic use of the term does not. So "NT vs. Unix security"
isn't the right question -- maybe "NT vs. Solaris 2.2" is better.
In terms of *services* running on top of the OS (email, directory,
http, routing, authentication, etc.), there are various security
issues with the protocols (architecture, such as SMTP, PPP, TCP/IP,
http) and implementations (such as sendmail, Netscape server, NT RAS).
Until recently most of the Internet protocols were not very secure
(despite lots of hackers) mostly because it wasn't the main issue with
services (getting them to work on the heterogenous Internet and
rapidly folding in many improvements was). The Internet was pretty
easy to hack compared to, say, RACF on IBM mainframes. Heck, look at
the web protocols, or TCP/IP, etc. These were NOT architecturally
designed to be secure. Even things like Kerberos aren't as secure as
really paranoid designs (professors at CMU joke that breaking Kerberos
was a graduate student project). For reasons of compatibility with
Internet protocols this makes many otherwise secure operating systems
suffer. These days there is much more focus on security and Internet
protocols are more mature, so I expect there will be much improvement.
===== My thoughts are my own =====
Edward Jung, edw...@interramp.com
Terry Thiel
<jkot...@interaccess.com> wrote:
> christop...@idsonline.com (Christopher L. Estep) wrote:
> >In article <48icl3$p...@brtph500.bnr.ca>, man...@bnr.ca says...
> >
> >> I am resisting the urge to make a very large bet with you... I
> >>will instead just settle for laughing uproariously.
> >>
> >
> >I should take the NT end, and spot you the entire SPARC market. Face it;
> >when CHRP PowerMacs begin shipping, most of them will NOT be running MacOS,
> >but NT. Why? For the SAME reason that Windows 95 has been kicking Warp all
> >over the planet: applications, applications, applications. NT ALREADY
> >outsells native versions of UNIX for all platforms except MIPS; is SGI gonna
> >sell THAT many SGI boxes?
> >
> How are you counting the copies of NT vs. copies of the various and
> sundry unix variants ??? Copies of OS sold or users supported by
> that 1 copy. I might agree with you that NT may have sold more
> copies than Unix ... might. But, I am certain that the number of
> users SUPPORTED by Unix vs. NT is a 10:1 ratio.
>
> Keep in mind the group I work with supports ~ 2500 Unix Users
> on 20 Copies of Unix. Number of NT stations: 0.
>
> You have 1 point right --- Win95/NT has desktop productivity
> applications on their side.
>
> So, while you are working on your spreadsheet (the 20th revison)
> I'll just get back to work building my network, WITH UNIX.
I don't think so. There are still very few NT native apps. I seriously
doubt people will buy CHRP boxes to run a single operating system of any
type. The whole idea is that you can run whats best for a particular
project.
-Terry
Darren M. Greenwald
>In article <xjlraz1...@maroon.tc.umn.edu>, albr...@maroon.tc.umn.edu
>(Bruce Albrecht) wrote:
>
>>
>> So? Most of the features in Windows 95 (and then some) were available in
>> home computers back in the mid 1980's, either in the Mac or the Amiga,
>
>This is a ridiculous argument. Because Microsoft didn't throw out the "how to write an
>Operating System" textbook and create everything from scratch, their product is inferior?
Yes, I have to agree with you here. Also, OS/2 has been available
for 3+ years now. If the market was scrambling to install a 32-bit
protected multitasking OS on their desktops, IBM had a good head start.
The dynamics of why people are going to choose Windows 95 is a much
more complex one than who has the best technology.
>> Maybe so, but OS/2 runs 99% of all windows programs, without the braindead
>> requirements DOS/Windows (including 95) has with respect to the 640K memory
>
>This argument doesn't hold water post-Win95. OS/2 does *not* run 32-bit Windows applications.
It is also problematic for people to have to deal with three user
interfaces (well, two if we ignore DOS). But ultimately the real
issue here is going to come down to what system the applications are
being written for, and in this regard, OS/2's ability to run Windows
and DOS apps has not resulted in more OS/2 apps.
>> Actually, the OS most people buy is the one that's pre-installed on their
>> machine. If IBM were to have done the cutthroat marketing to get OS/2
>
>* There are virtually no quality OS/2 applications
>* The OS/2 environment has been shown time and time again to be incredibly confusing to end
>users, even compared to the aging Windows 3.1 metaphor
>* OS/2 has a terrible installation routine
>
>The reason it isn't pre-loaded is because there's no demand for it to be pre-loaded. OS/2
>can't compete with Windows 95 or Windows NT. It's in a no-man's land sort of middle ground.
>Not as robust or scalable as NT, and not as friendly or easy as 95. Not an enviable position.
I agree with you. My opinion is much the same.
Again, and I know this flames OS/2 fans, I have to agree. NT meets
the demands of a market that require near crash-proof behavior and
security, but for the small business or users at home, Win95 is
slick, polished, much less confusing, and I could go on. Yes,
I know it is popular to bash Win95, but my experience has been
that it has all of the right things going for it. Despite
some of the things people post, I have crashed it two times
since August 24th (my fault while programming) and once it
locked up during a tape backup (but never since). Ok, otherwise
it has happily trapped the compiler's own crashes, and a few
misc applications that access memory illegally without crashing
itself. That is a total of three times for a brand new OS,
two times are my fault, and I leave my computer at home on
all of the time.
There are a lot of dynamics that you have to factor in to
understand why the market has so quickly adopted Windows 95,
and not OS/2. But blaiming Microsoft because they do a better
job marketing is silly. If any company has had the muscle
to build a market for a new OS, it is IBM. They have had
at least a three year head start too. If at this point
Aptiva's (nice machines of course) are being sold with
Windows 95 installed instead of OS/2, that tells us something.
OS/2 has a lot of nice things going for it, but from a purely
subjective point of view, Win95 'feels' good to use. It
is very clear that a lot of time went in to polishing it and
making sure it is consistent everywhere. Even the applets
that come with it are nice and consistent with the system,
not some console-looking things with strange ugly fonts and
mixed size graphics images that look like a computer programmer
designed them because they fit in 8x8 cells :)
If quality is in the details, it would be entirely
unfair to Microsoft to say that Win95 is a loser product.
Sure, it is true that it borrows ideas from existing
systems, but duh, so what? Who cares? When was the last
time you refused to buy something on the grounds that some other
companies product has done something similar before???
--
-----------------------------------------------------------
Darren M. Greenwald | Scala Inc. R&D /
Senior Systems Engineer | dar...@scala.com o
-----------------------------------------------------------
Smoke Crack and Worship Satan
There are plenty of 32-bit native apps for NT on the Intel platform, courtesy of Windows 95.
Almost every Windows 95 logoed title will run on NT. As for the other platforms, well...
there ain't much.
Burner
> Here is a sampling of Netscape vs. other GUI Web browsers
>
>
> Oct 1995:
> NetScape 321000 87%
> NCSA Mosaic 18600 5%
> MS 8800 2.5%
> AOL 8300 2.5%
> AIR Mosaic 4467 1.2%
> Prodigy 2615
> WebExplorer 2435
> NetCruiser 2060
> MacWeb 738
> OmniWeb 710
> ------
> Total ~ 370000
>
>
> Seems like anything other than Netscape does not matter....
Just like MS-Windows/Office... :)
>
> Sinan
> --
> Redistribution by Microsoft Network is prohibited.
>
>
I know that's a GUI browser listing, but it does leave out lynx and WWW
(which includes stuff like slipknot and the like - graphical browsers
that don't need SLIP or ethernet). Just thought I'd add that.
-----------------------------------------------------------------------------
-=<The following is a .sig>=-
MRSH - A recording device, a corned beef sandwich with russian dressing, an
aquatic bird, and the foamy stuff at the top of a beverage
-Kerrly
bc8...@binghamton.edu
preferred addr. --> bur...@degobah.core.binghamton.edu
Check out my homepage!
http://www.core.binghamton.edu/~burner
_ _ __ _ _ _
|\ /| | \ / \ | |
| \ / | | | \ | |
| \/ | |_/ \ |---|
| | | \ \ | |
| | | \ \_/ | |
Jeff Bonwick
>
> Could this "success" have something to do with the fact
> that the "worm" struck some six years ago (long before
> the Firewall-1 existed)
Sorry, I should have been more precise. The worm failed to get through
sun-barr, Sun's main internet firewall. The firewall technology that
our internet and security gurus developed over the years for sun-barr
was recently made available as a product called Firewall-1. The worm
predates the commercial product offering, but not the technology.
My apologies; I should have made that distinction clear.
Jeff Bonwick
Solaris Performance
Hugh LaMaster
cop...@chem.wisc.edu (Stingo) writes:
|> You guys are living in a dream world. Major corporations
|> who spend zillions of dollars just to figure out there
|> logo are switching to NT.
I recall being told sometime in 1990-91 that everyone
was switching to NT, even though beta versions were
not even available. But, NT was just around the corner,
and once it was available, everyone would move to it pronto.
Finally, in 1993, NT actually arrived. Strangely enough,
all the "server" functions, including really simple things
like a telnetd, were not available for a simple "client"
workstation, I was told. No X Window System, and,
no remote windowing at all. And other "features" such
as really dumb routing. Now, I have not followed
the development of NT carefully, so maybe someone can
answer whether or not these basic capabilities have
finally been provided. In any case, the folks making
the prediction back in 1990-91 were a bit off, I would
say. I wonder if they might be off again?
UNIX is an old tired OS that
|> has served us well, but it is time to bury it.
Old, yes. But if it still has many features not
available, and perhaps even inexpressible in, the
"new" operating system being offered, perhaps the
new OS should be described as "new and tired".
Or maybe still just an immature baby.
As desktops
|> become more and more like mini's, clustering, and all that
|> distributed stuff, NT is poised and has major support,
So NT is poised to do most (but not all) of the things
that SunOS 3.5 could do 10 years ago? [And DEC VMS could
do by ~1988 ...] This is a great leap forward?
but
|> time will tell and its a strange world.
Yes, it is. I am frequently surprised by what short
memories people in this business often have. Most
of these OS features were conceived back in the
1960's, denigrated by the PC crowd back in the 80's
as "too complex", and are now being reinvented and
marketed as if they were brand new.
I see that Taligent
|> is being dumped.
Yes, and that means ... ?
Anyway, most folks running Unix are rather accustomed
to its capabilities and won't easily give them up.
OTOH, I could easily imagine that Windows users might
be attracted to NT, since, limitations aside, it is a
Real(tm) OS, possibly the first such OS most Windows
users will have experienced.
--
Hugh LaMaster, M/S 233-18, Email: Please send ASCII documents to:
NASA Ames Research Center Internet: lama...@ames.arc.nasa.gov
Moffett Field, CA 94035-1000 Or: lama...@george.arc.nasa.gov
Phone: 415/604-1056 Disclaimer: Unofficial, personal *opinion*.
Smoke Crack and Worship Satan
>
> So? Most of the features in Windows 95 (and then some) were available in
> home computers back in the mid 1980's, either in the Mac or the Amiga,
> available to Unix PC users for 5-10 years, and OS/2 users for several years
> on the PC (where they also could run their old DOS/Windows apps). Sounds to
> me like Microsoft is recycling old technology and calling it "NEW!!!!".
>
This is a ridiculous argument. Because Microsoft didn't throw out the "how to write an
Operating System" textbook and create everything from scratch, their product is inferior?
If you want to make a good product, you study the work of the people who went before you. You
recreate the things they did well, and you avoid the things they did wrong. There is no
perfect program or operating system.
>
> Maybe so, but OS/2 runs 99% of all windows programs, without the braindead
> requirements DOS/Windows (including 95) has with respect to the 640K memory
> limits. IMHO, dealing with the 640K limit has probably wasted more time than
> any other computer related time-lossages prior to the introduction of
> Netscape. Why didn't users flock to OS/2 in droves? Marketing. Stupid
> marketing by IBM, and cutthroat marketing by Microsoft.
>
This argument doesn't hold water post-Win95. OS/2 does *not* run 32-bit Windows applications.
That means the newest, coolest apps cannot be executed under OS/2. So if you want to run old
software, by all means use OS/2. The rest of us want to move forward. Nobody except Lotus is
developing for OS/2.. and they develop for Win32, too.
>
> Actually, the OS most people buy is the one that's pre-installed on their
> machine. If IBM were to have done the cutthroat marketing to get OS/2
> pre-installed on most major PC manufacturers machines, rather than being
> idiots and pricing it out of the marketplace, it would be the number one OS
> on the market today rather than DOS/Windows. It runs all the old user apps,
> allows them to build 32 bit native apps, etc., but because IBM in their
> infinite stupidity primarily sold OS/2 to Big Business as a aftermarket sale,
> lost out big time.
>
Never mind that:
Ken Kriesel, Physical Sciences Lab, UW-Madison
>
> ....There is a difference between the number of copies of NT sold and the
> number of NT systems actually used for useful work...I have NT running on
> my PC in the office. I wouldn't say that it's much more than a paper
> weight -- I do ALL my work (compiling, debugging, web surfing, netnews, etc.)
> on the workstation and only turn on the NT box when I need
> to create a Win32 executable of my app. It's pretty good for that.
Yes, usefulness to the end user is what they are for and what matters.
For the applications needed by that end user. Quite a valid point,
and one which means that all the Unix boxes at my work should be
subtracted from the Unix total.
Does it mean that the DOS boxes & NT boxes that work 2 shifts or
more should be counted twice? Don't claim the Unix box can do it in
1/n the time either, I've seen the benchmarks for the Unix boxes of
comparable cost to a current PC and they do not measure up for the
applications we run daily, overnight or through a weekend.
Ken
Michael Hancock
>In article <m#bs...@quack.kfu.com> h...@quack.kfu.com (Ho Yoon) writes:
>>
>> My understanding was that NT, or for that matter, most desktop OSes,
>> didn't have the security holes which Unix had.
[snip]
>NT is designed to be secure*, but it's difficult to assess the practical
>side. Breaking into NT systems just hasn't become cool yet, probably
>for the simple reason that NT systems aren't protecting exciting stuff.
>Breaking into universities, banks and Pentagon labs is interesting;
>breaking into Bob's spreadsheet, not so much.
>
>* I've read that NT's support for older apps constitutes an architectural
> security hole because these apps have to be able to access physical
> resources directly. It sounds plausible, but does anyone out there
> know for sure? I'd think MS could get around this by running these
> apps in an isolated virtual machine (though it might kill performance).
>
>One thing is certain: security is going to become an absolutely critical
>issue in the coming years. Historically most break-ins were motivated
>simply by the intellectual challenge: can I do it? But real Internet
>commerce will motivate real Internet crime. Operating systems that
>haven't been shaken down by the innocuous propeller-heads of the world
>may be in for a real baptism by fire.
Some MS Security Holes:
http://www-leland.stanford.edu/~llurch/win95netbugs/IP-Security-Bug.txt.a
sc
Though it is easy to setup basic networking with NT. It's bound to have
some costs. Underneath it all it is *still* LANMAN.
--
Mike Hancock
mich...@cet.co.jp
John DiMarco
>j...@cdf.toronto.edu (John DiMarco) wrote:
>> that most CERT advisories and security bulletins address UNIX is a consequence
>> of the fact that most of the server systems on the Internet run UNIX of one
>> flavour or another. There just aren't a whole lot of VMS, NT, OS/400, etc.
>> servers out there in comparison.
>What do you consider a lot? There are millions of NT licenses out
>there in use, many of them on the Internet.
Say, a substantial portion (>10%) of the accessible hosts on the internet?
>It isn't just the server systems that need to be secure, but also
>workstations.
Accessible (eg. unfirewalled) workstations, certainly.
>I suspect the level of security attacks on Unix has at least as much
>to do with the newness of NT vs. the age of Unix; there's been plenty
>of access to Unix systems and time to learn where the holes are. As
>an NT user, I wonder how many holes NT has that simply are not common
>knowledge yet.
Agreed.
>Or that the security is not as good as the US Treasury or the Cali
>cartel's; far more money there, but the cost of getting to it could
>be a bit high. As time goes on, NT will be increasingly "where the
>money is".
Certainly Microsoft would wish it so. If and when that time comes, NT will
probably be a more attractive target for malfeasance than it is now.
thefuzz
: In article <48sbt5$12...@saba.info.ucla.edu> jma...@cs.ucla.edu (Jay Martin) writes:
: >
: > What? Unix security has always been a widely recognized joke
: > (ever heard of the internet worm?)
: Yes I have, and FYI, it didn't get through Sun's Firewall-1.
Could this "success" have something to do with the fact
that the "worm" struck some six years ago (long before
the Firewall-1 existed) and was designed to strike VAXen?
: > which is in my humble opinion due to incompetent (hacked) design.
: The design is fine; there are no architectural holes in Unix security.
: There are bugs in the implementations, which are of varying quality.
: Anyone who wants to be around in 2-3 years takes them quite seriously.
As used today, there is a major architectural hole in UNIX security
in that it relys on trusting its peers. Without this trust,
the "Network IS the computer" slogan would be without value.
UNIX network security requires that all trusted systems (NFS & rlogin)
be equally secure.
Bob
Ken Kriesel, Physical Sciences Lab, UW-Madison
>
> In many ways, security is in the eyes of the beholder. For example, a
> Macintosh is traditionally the most secure HTTP server possible,
> simply because, unless you explicitly add them, it doesn't support any
> of the features that cause other servers' security holes.
>
> If you want to compare the "inherent" insecurity of one system over
> another, configure them with similar capabilities. For instance,
> disabling "sendmail" altogether on a UNIX box increases its security
> dramatically. fingerd similarly, and telnetd likewise. Certainly
> rshd and rlogind. Preferrably also take out X and NFS. You're left
> with a fairly secure system.
>
> Another reason, it seems to me, that "UNIX" security holes may appear
> more often, is that there are quite a few UNIXes out there. Comparing
> the security of a particular brand of UNIX to NT might be more
> enlightening, since you're probably only running one at a time.
Whether for the Mac, Unix, NT, or other OS, limiting the number of
services installed definitely improves security.
One of the criticisms leveled at MS operating systems recently is that
the default configuration was not very secure. It appears from your
comments that at least some flavors of Unix also have that
characteristic. (MS responded by providing C2CONFIG in the resource
kit, a simple-to-use tool which lets a user tighten various aspects
of NT security in a point-and-click way.)
On a particular machine, probably one to two OS's will be used. But
a business unit may be running several OS's on several HW platforms
simultaneously. This is particularly true of organizations that do
contract support or programming work. One might have any of DOS,
Win3.1, WFWG, Win95, NTWS, NTServer, MacOS, VMS, Ultrix, and other
Unix flavors; OS/2, Novell, etc.
This diversity derives from choosing office and other applications
from the mass market, and needing a seat or two to do development
on the same platform as the diverse preexisting choices of customers.
Then the question is, what to put on new machines and where to put
new or moved services, and which OS's to phase out or introduce.
The amount of effort needed for learning and handling the security
and other quirks of a particular OS certainly factors into those
choices.
Examining the holes present in a fully-loaded Unix system or other OS
is a useful way to learn where some holes may be in NT, and so how to
configure an NT system for acceptable security levels. Looking at
Unix is particularly effective because a number of the services being
made available for NT began in Unix, and seems to be the most often
discussed.
Counting Cert advisories listing a specific brand of Unix once (not
once for each flavor of Unix afflicted with the problem) and counting
NT once when both WS and Server are afflicted is a fair method.
It will still show Unix mentioned in more advisories, in my experience.
It is rare to see an advisory mention NT, but the usual advisory
mentions Unix.
Someone doing an impartial formal comparison and posting the result
would do the newsgroup a great service.
Is there a table that cross references CERT advisories to OS, for the
past year or two?
Ken
Andre April
: As used today, there is a major architectural hole in UNIX security
: in that it relys on trusting its peers. Without this trust,
: the "Network IS the computer" slogan would be without value.
: UNIX network security requires that all trusted systems (NFS & rlogin)
: be equally secure.
It is possible now to use UNIX systems in a network without trusting
the peers. There are lots of secure protocols that ensures that the
peer system is the one it claims to be and that encrypts the
communication in such a way that only the peer can decrypt.
Have a look at <http://www.cs.hut.fi/crypto/protocols.html>
Of course, these secure protocols use more resources. It is your
responsibility to choose whether they worth the overhead or not.
That's where the problem is. People are not conscious of security
problems or underestimate them. Security by obscurity is still to widely
accepted (and practiced).
: Bob
André.
-------------------------------------------------------------------
André April Département d'ingénierie informatique
E-mail: a...@info.ucl.ac.be Place Sainte Barbe, 2
Tel: +32 (0)10/47.31.13 1348 Louvain-la-Neuve
Fax: +32 (0)10/45.03.45 Belgium
Tom Holub
Ken Kriesel, Physical Sciences Lab, UW-Madison <kkri...@facstaff.wisc.edu> wrote:
)
)Counting Cert advisories listing a specific brand of Unix once (not
)once for each flavor of Unix afflicted with the problem) and counting
)NT once when both WS and Server are afflicted is a fair method.
Uh, no it's not. CERT advisories are released only for vendors which
cooperate with CERT and only for problems which have patches or workarounds
available.
-Tom
Robert Krawitz
In article <xjlraz1...@maroon.tc.umn.edu>, albr...@maroon.tc.umn.edu
(Bruce Albrecht) wrote:
> So? Most of the features in Windows 95 (and then some) were
> available in home computers back in the mid 1980's, either in the
> Mac or the Amiga, available to Unix PC users for 5-10 years, and
> OS/2 users for several years on the PC (where they also could run
> their old DOS/Windows apps). Sounds to me like Microsoft is
> recycling old technology and calling it "NEW!!!!".
This is a ridiculous argument. Because Microsoft didn't throw out
the "how to write an Operating System" textbook and create
everything from scratch, their product is inferior?
The problem is that DOS and Windows have lots of flaws that prevent it
from being a reliable and robust operating system.
If you want to make a good product, you study the work of the
people who went before you. You recreate the things they did well,
and you avoid the things they did wrong. There is no perfect
program or operating system.
And if you want to maintain a monopoly position, you deliberately do
things differently to try to prevent competitors from catching up.
This argument doesn't hold water post-Win95. OS/2 does *not* run
32-bit Windows applications. That means the newest, coolest apps
cannot be executed under OS/2. So if you want to run old software,
by all means use OS/2. The rest of us want to move forward. Nobody
except Lotus is developing for OS/2.. and they develop for Win32,
too.
There was someone else who was trolling the OS/2 newsgroups not long
ago who used much the same marketing-ese ("newest, COOLEST [emphasis
mine]," "packed with new features" and so forth). There's nothing
"cool" about choice of application; it's a matter of what gets one's
work done most efficiently at least cost. On that standard, the new
applications had better provide a BIG gain to justify their prices.
I notice, BTW, that you're posting from a nonexistent domain (my
nameserver replies authoritatively to that effect). I'm very curious
who you work for.
* There are virtually no quality OS/2 applications
There are more OS/2 applications than Windows '95 ones.
* The OS/2 environment has been shown time and time again to be
incredibly confusing to end users, even compared to the aging
Windows 3.1 metaphor
By whom? Microsoft?
* OS/2 has a terrible installation routine
Well, all right, the installation routine is a bit of a pain.
--
Robert Krawitz <r...@tiac.net> http://www.tiac.net/users/rlk/
Member of the League for Programming Freedom -- mail l...@uunet.uu.net
Tall Clubs International -- tci-r...@aptinc.com or 1-800-521-2512
James D Pollock
>Actually, the OS most people buy is the one that's pre-installed on their
>machine. If IBM were to have done the cutthroat marketing to get OS/2
>pre-installed on most major PC manufacturers machines, rather than being
>idiots and pricing it out of the marketplace, it would be the number one OS
>on the market today rather than DOS/Windows. It runs all the old user apps,
>allows them to build 32 bit native apps, etc., but because IBM in their
>infinite stupidity primarily sold OS/2 to Big Business as a aftermarket sale,
>lost out big time.
>
>
people didn't want to have OS/2. Why? The same reason they buy systems with Intel
processors, or with Creative Labs sound boards: They have the best chance of being
compatible, both with existing software and with new products. It really is a
self-regenerating spiral: Because it is the top-seller, people buy more of it.
Because people keep buying it, it stays the top seller.
This same principle worked in the home videogame market, too: Nintendo had the best
licensee software production, so they had the best games. They had the best games,
so they sold more game machines. They sold more game machines, so they got the best
licensees.
Both the TurboGrafx-16 and the Sega Master system were better technically than the
NES. But the NES sold rings around the others. Why? Because people don't buy
hardware to have the best hardware; they buy hardware to run the best software.
Now back to the original point: Windows 95 sold because MS had the huge market
weight to guarantee third-party support. This will sell to consumers. OS/2 sells
because it has features that a "technical elite" desire. This will NOT sell to
consumers. People buy an OS to run their software. If their software is all
Win/Win95, then they will buy Win95.
Robert Krawitz
In article <JAN.95No...@cora.neuroinformatik.ruhr-uni-bochum.de>,
j...@neuroinformatik.ruhr-uni-bochum.de (Jan Vorbrueggen) wrote:
> Duh. It's still recycling. Why should an OS feature be "reserved"
> to a workstation? The first versions of VMS, with much more
> functionality then than DOS has now, ran on a machine with 256 kB
> --- yes, 256 _kilobytes_ --- of main memory. It's been a number
> of years now that an average PC has that amount of cache. No, the
> reason is a bad technical decision by IBM that managed to
> perpetuate itself; that's what you get from the chaotic,
> self-organizing principles of capitalism.
Oh, I don't know. Perhaps because the workstation and home markets
are COMPLETELY DIFFERENT? Workstations start at $5,000. Home
computers start at $800. Think about it.
I don't know what that comment has to do with anything. In any event,
there's precious little difference between a workstation and a
high-end "home PC" these days. I have a Pentium 90 with 16 MB RAM
these days, and Linux on that PC blows away the Sparcstation 10 on my
desk at work. Certainly the graphics systems on PC's are superior to
many workstation display systems. The only substantive difference is
marketing.
Bryan Althaus
: Both these pointers discuss WFWG and Win95 holes. Providing that
: due caution is used to close these holes and prevent them being
: windows onto NT systems, there's nothing relating to NT here.
Explain windows onto NT systems? IF this means that NT security
would be compromised, then even though the bug is on WGWG or Win95,
it still shows a way to get at NT. You did not mention if NT received
a patch.
: Anyone have some NT security holes to report, other than the Unix-
: style-service ones?
It's still kind of early yet for NT and Net security. Our company is NT
happy but when it came to a Net server, they went with Solaris (UNIX)
not NT.
This from someone's recent sig:
"I don't know if security explains why the Win95 support Web servers run BSDI
2.0--an Intel-based Unix--rather than Windows NT, which Microsoft insists is
the ideal Web software solution. Does Redmond know something we don't know?"
-Robert X. Cringely, INFORWORLD, 9/11/95
Terry Lambert
]
] bon...@cathy.eng.sun.com (Jeff Bonwick) writes:
]
] >The various flavors of Unix are among the most secure operating
] >systems you can get because (1) Unix was designed to be secure,
] >and (2) hackers all over the globe have spent a couple of decades
] >trying to break in, so most of the bugs have been found and fixed.
] What? Unix security has always been a widely recognized joke (ever
] heard of the internet worm?) which is in my humble opinion
] due to incompetent (hacked) design.
Call me when NT gets a B2 rating and per process credentials
(instead of per connection credentials).
Terry Lambert
te...@cs.weber.edu
---
Any opinions in this posting are my own and not those of my present
or previous employers.
Jan Vorbrueggen
Oh, I don't know. Perhaps because the workstation and home markets are
COMPLETELY DIFFERENT? Workstations start at $5,000. Home computers start
at $800. Think about it.
Sure, they're different. But is the price of developing software the reason
that the quality of said software is so different on the prototypical PC and a
workstation? When I look at the relative rate of bug patches coming in for
Solaris and VMS, I don't think so.
Jan
Doug McDonald
r...@sunspot.tiac.net (Robert Krawitz) wrote:
> In article <48tftr$6...@nntp.crl.com> m...@aplace.com (Smoke Crack and Worship Satan) writes:
>
> In article <JAN.95No...@cora.neuroinformatik.ruhr-uni-bochum.de>,
> j...@neuroinformatik.ruhr-uni-bochum.de (Jan Vorbrueggen) wrote:
>
> > Duh. It's still recycling. Why should an OS feature be "reserved"
> > to a workstation? The first versions of VMS, with much more
> > functionality then than DOS has now, ran on a machine with 256 kB
> > --- yes, 256 _kilobytes_ --- of main memory.
We had an early VAX, in fact we got serial number 5 (though we never
used it, as it was dropped off the dock and replaced by SN 32).
The first version of VMS did **NOT** have in shape or form many features
anywhere as near as advanced as even available on PC many, many years ago.
1) It did not support in ANY way graphics. The only way to do graphics
was with a "graphics terminal" over a 9600 baud serial line (unlike
PC COM ports, the VAX ports really truly supported the spec: MAXIMUM
rate of 9600 baud ... absolutely, unequivocally, enforced by
banks of line-slowing capacitors.)
I remember well the campus demo of the first MircoVax: hoots of
laughter greated their attempt at a graphics demo.
2) It did not support in ANY way networking (networking, in the form of
the proprietary DECNet was promisde, but not delivered ... and the
promise fas only for file transfer .... a remote login was NOT to be
supported, by policy edict!)
It DID of course have many good features:
1) Virtual memory that actually worked well. In this it beat out
IBM mainframes.
2) Multiuser ... it supported, well, up to maybe 5 users doing something minor.
3) And one half-good, half bad, feature: record oriented IO that was less
painful than IBM mainframes, but still a royal pain compared to PC or Unix IO.
For people moving from $SYSIN DD * it was a pleasure, but only for those.
PCs have been able to do excellent multiuser stuff, as good as VMS from the
user viewpoint, since the PC AT and Xenix. This did not, of course, have
real virtual memory. PCs to this day do not suffer from the abomination of
record oriented IO, whether they run Windows 95 or Unix.
The big lag .. and it is TRULY AMAZING to me that people didn't demand it
earlier .. is virtual memory.
Doug McDonald
mcdo...@aries.scs.uiuc.edu
Smoke Crack and Worship Satan
wrote:
>
> In article <48uf8v$9...@nntp.crl.com> m...@aplace.com (Smoke Crack and Worship
> Satan) writes:
>
> The problem is that DOS and Windows have lots of flaws that prevent it
> from being a reliable and robust operating system.
>
Talk about flawed-- I know Mac users whose system crashes on a daily basis. You could argue
that the MacOS is flawed, not reliable, not robust.. but does it really make any difference?
Is this what's important to consumers?
> If you want to make a good product, you study the work of the
> people who went before you. You recreate the things they did well,
> and you avoid the things they did wrong. There is no perfect
> program or operating system.
>
> And if you want to maintain a monopoly position, you deliberately do
> things differently to try to prevent competitors from catching up.
So Netscape, by introducing new and incompatible HTML extensions, is doing just this?
Hypocrisy is the greatest luxury. Market share rules, so Netscape can make the rules. There's
very little substantive difference between what Netscape is doing and what Microsoft is
doing. It ain't pretty, but it's the way things get done in the computer industry.
>
> This argument doesn't hold water post-Win95. OS/2 does *not* run
> 32-bit Windows applications. That means the newest, coolest apps
> cannot be executed under OS/2. So if you want to run old software,
> by all means use OS/2. The rest of us want to move forward. Nobody
> except Lotus is developing for OS/2.. and they develop for Win32,
> too.
>
> There was someone else who was trolling the OS/2 newsgroups not long
> ago who used much the same marketing-ese ("newest, COOLEST [emphasis
> mine]," "packed with new features" and so forth). There's nothing
> "cool" about choice of application; it's a matter of what gets one's
> work done most efficiently at least cost. On that standard, the new
> applications had better provide a BIG gain to justify their prices.
>
The point is this: many companies are not developing 16-bit Windows 3.1 applications any
more. Much the same way that companies really don't develop DOS apps any more. It may be a
viable situation now, but a year from now, OS/2 users are going to be using some creaky old
software indeed.
>
> * The OS/2 environment has been shown time and time again to be
> incredibly confusing to end users, even compared to the aging
> Windows 3.1 metaphor
>
> By whom? Microsoft?
>
Actually, no.. by every type of usability study I could name. The most recent one that I
remember was in HomePC magazine last month. Look it up if you're that interested.
Arun Gupta
-arun gupta
Excerpts from :
Microsoft, Novell seek C2 approval. (C2 security level is a goal for
Microsoft Windows NT 3.5 and NetWare 4.1)
Fisher, Sharon; Jacobs, Marcia A.
CommunicationsWeek ISSUE: n573 Pagination: p5(2)
Publication date: Sept 4, 1995
.....
Two Types of Security
The NCSC, part of the National Security Agency, defines two types of C2
security. One, specified by the Trusted Computer System Evaluation Criteria
(commonly known as "the Orange Book"), covers stand-alone systems. The other,
specified by the Trusted Network Interpretation (commonly known as "the Red
Book"), includes networked components as well. C2 security covers
authentication, access control, auditing and object reuse.
Microsoft's Windows NT 3.5 was submitted for TCSEC evaluation, while
Novell's NetWare 4.1 was submitted for TNI evaluation. Novell was able to do
this by working with Cordant Inc., which makes a secure Ethernet card, said
Dave Clare, product line manager for Novell's NetWare core services.
With the addition of the secure network card, users can create a secure
network, not just a secure server, he said. Novell also will be able to work
with any other vendor that provides a secure network interface card, he added.
"It's a subtle but distinct difference," Pescatore said. "Typically, the
vendor gives [the NCSC] a software product, and they evaluate that standalone
thing. "What Novell did was they took NetWare and Cordant and built a little
LAN configuration, so they basically got a C2 evaluation for a LAN, as opposed
to a standalone product. Most of the C2 stuff, including the way that Windows
NT 3.5 is being certified, meets C2 specs, but when you connect it to a
network, you're on your own. Novell got it blessed in a network
configuration."
Vendors and analysts cautioned that users need to follow many policies and
procedures for a network to remain in C2 compliance.
.....
Windows NT Workstation received its C2 evaluation on Aug. 18, and the
network version will be certified within the next two months, said Megan
Bliss, Microsoft's product manager for the Windows NT workstation.
****
[So, has it ? -arun gupta]
Edmond
>| Typical Unix architecture does not specify enough to pass B; many
>| implementors (including Sun) do extra work to make it so. Thus, some
>| specific Unix *implementations* have security architectures that are
>| equivalent (or superior) to NT's architecture, but "Unix" in the
>| standard, generic use of the term does not. So "NT vs. Unix security"
>| isn't the right question -- maybe "NT vs. Solaris 2.2" is better.
>
>Are you trying to tell us that every sold copy of NT is passing
>B validation?
MS is submitting NT for B2 clearance, but that doesn't mean that it has
passed it yet.
--------------------------------------------------------------------
Edmond Underwood
OCS
E-mail: unde...@Colorado.Edu
Arun Gupta
Some more security related stuff.
-arun gupta
New Wang PC with secure OS earns NSA's B3 security rating. (Wang
Federal Inc's XTS-300 workstation) (Product Information)
McCarthy, Shawn P.
Government Computer News VOL: v14 ISSUE: n22 Pagination: p1(2)
Publication date: Oct 16, 1995
Abstract: Wang Federal Inc's XTS-300 workstation running its Secure Trusted
Operating Program (STOP), a Unix-like operating system, gets a B3 security
rating from the National Security Agency (NSA)> B3 is the highest B-level
security and indicates high resistance to tampering and penetration. The
XTS-300 succeeds the older XTS-200, the only other workstation to be
accorded the security rating. The XTS-300 is based on an Intel 486 50MHz
DX2 processor and costs $30,000 for a system with STOP, 16MB RAM, a 500MB
hard drive, a tape drive, a 3.5-in floppy drive and one parallel and two
serial ports. Another version costs $44,000, sits between two networks and
includes a coded security policy.
Kevin Burton
when one compares NT to 3.1, 3.1 always wins. NT cannot
communicate with other applications throughout the use of OLE
or DDE. However I do believe that some applications such as
those that are legacy can execute resources without going
through conventional means such as the _new_ APIs that have
been developed specifically for NT.
I think that in the future as NT becomes more of a venerable
competition for unix boxes that we will see more breakins.
To believe we live in a world that is safe from hackers is a
falsehood. As operating systems get more complex they become
more vulnerable. The perfectly secure operating system does
not exist. The only way to assure that someone from the
outside world cant get into your systems is to pull the plug.
Otherwise be very carefull.
Kevin Burton
(kbu...@umbc.edu)
Mike
> Funny you should mention VMS. DEC is moving it's VMS customers to NT.
<>------- j...@neuroinformatik.ruhr-uni-bochum.de (Jan Vorbrueggen)
> Did'nt you get that the wrong way 'round? I see DEC as moving
> its NT customers to VMS, what with the NT API support being
> integrated into VMS and VMSclusters being ported to NT.
from:
http://techweb.cmp.com:80/techweb/programs
/cmp_waisgate?RF=817150995.24417&num=13#head
Making A Move
Last month, HP rival Digital Equipment announced a
comprehensive program to migrate customers of its
proprietary Open-VMS-based Alpha systems to NT over
the next five years. The plan was a break in tradition
for the Maynard, Mass., vendor, which had maintained a
policy of not trying to sway customer migration decisions
to either Digital's Unix or NT-based Alpha platforms.
==================================================================
see also:
http://techweb.cmp.com:80/techweb/programs
/cmp_waisgate?RF=817150995.24417&num=24#head
http://techweb.cmp.com:80/techweb/programs
/cmp_waisgate?RF=817150995.24417&num=15#head
muzo
>* I've read that NT's support for older apps constitutes an architectural
> security hole because these apps have to be able to access physical
> resources directly. It sounds plausible, but does anyone out there
> know for sure? I'd think MS could get around this by running these
> apps in an isolated virtual machine (though it might kill performance).
>
this is hilarious. NT is criticized (sp?) for not running all dos/win16 apps
because it doesn't let them touch hardware and someone says something like this.
The fact that NT is not fully compatible with old programs because of security is
one of the most notorious facts about NT.
What have been reading ?
muzo
standard disclaimer
Jan Vorbrueggen
writes:
Funny you should mention VMS. DEC is moving it's VMS customers to NT.
Did'nt you get that the wrong way 'round? I see DEC as moving its NT customers
to VMS, what with the NT API support being integrated into VMS and VMSclusters
being ported to NT.
Jan
Terry Thiel
Worship Satan) wrote:
> In article <48icl3$p...@brtph500.bnr.ca>, man...@bnr.ca (Cathy Mancus) wrote:
> >
> > : And there will be more machines running NT than UNIX by 1997 IMHO.
> >
> > I am resisting the urge to make a very large bet with you... I
> > will instead just settle for laughing uproariously.
>
> You could be right. But consider the volume of machines running Windows
95, and
> what percentage of those users will opt for Windows NT. As for UNIX--
the largest
> selling RISC machine is the PowerMac. Let's make it 1998. :)
Whatever. The reality is that NT has had phenomenal growth, far
surpassing even Microsofts expectations. Corporations are replacing Unix
and Netware with NT in droves. Personanly I have mixed feelings about
this since I can't stand Netware and Unix is too fragmented but Microsoft
is not a company I like to rely on.
-Terry
Pete Clinch
: That's the idea. Using the same OS for different types
: of tasks.
Quite absurd to scale *everything* though. Why would someone writing
letters on a desktop PC want or need the complexities of UNIX or NT on
their machine?
Don't give people sledgehammers to crack nuts!
: If we want to run force field calculations, it
: has to be done in UNIX. I have used UNIX and I don't see
: what the big deal is. It takes way to much time to master.
: I know I shouldn't be asking this but - what are the top
: five attributes of UNIX that are not in NT?
these are just *my* preferences:
1-4: Multi User system. This week we've been running our 4 Suns (hence
reasons 1-4) to host X sessions for an image analysis course, with a
dozen people logging in from PCs. Tricky with a single user operating
system, and the program was only available for the workstations, not the
PCs, so needed a compute server.
5: Doesn't run on SPARC or PA-RISC (or PPC???) platforms, which are the best
selling technical workstation lines with the biggest installed base, which
people will not want to throw away.
6: Before you wanted to scale your OS to everything. "Everything" on NT
is limited to 4 processors right now. Solaris does 64 out the box.
"Everything" on NT is limited to 32 bit. OSF/1 is available as a full 64
bit OS. There's more choice on UNIX to fit what people specifically want
to do, as there are several UNIXes (a curse as well as a blessing...) If
you want a special job doing, get something that fits it well. The job
may be a general purpose OS, for which NT is fine, but UNIX has been in
interesting niches for longer, so often represents a better choice for
specialist jobs. We want to interface with MRI scanners: they run on
UNIX boxes, because their development lead time is so long that NT didn't
*exist* when they were designed. Much easier to run the same box that's
under the scanner (usually a SPARC 2, 10 or 20).
7: Turning your question around, what top 5 attributes does NT have that
UNIX doesn't? I can manage UNIX, but not presently NT. Why should I
learn NT when UNIX is doing what I want very well? If I didn't know
either, I might well go for NT, but why throw away my experience just for
the sake of some Big New Thing?
It isn't a case of being *better* particularly. UNIX is better for some
things, NT for others. Communications between all systems are
increasingly better, so there isn't *that* much problem running them side
by side.
Most desktops are inappropriate for either, as they just don't need that
sort of OS, so it's absurd saying "everyone will/should use
NT/UNIX/whatever".
Get a life people, not a flame war...
Pete.
--
Peter Clinch Dundee University & Teaching Hospitals NHS Trust
Tel 44 1382 660111 x 3637 Medical Physics, Ninewells Hospital
Fax 44 1382 640177 Dundee DD1 9SY Scotland UK
net p.j.c...@dundee.ac.uk http://www.dundee.ac.uk/MedPhys/
Arun Gupta
Somebody was saying ?
-arun gupta
C2 rating aside, NT isn't secure. (National Security Agency's C2
certification of Windows NT unjustified) (Technology Information)
Constance, Paul
Government Computer News VOL: v14 ISSUE: n19 Pagination: p1(1)
Publication date: Sept 4, 1995
Abstract: Despite the National Security Agency's (NSA) C2 certification of
Windows NT, anyone with rudimentary programming skills can easily breach
Intel-based computers running the Microsoft operating system. Copying and
deleting information without leaving behind an audit trail can be done by
booting from a floppy and using assembly-language commands to manipulate
data stored on the hard drive. Apparently, the NSA tested the operating
system on Compaq computers with disabled floppy disks, because it views
floppy drive security as a 'physical security' issue. The Department of
Defense (DOD) has installed more than 100,000 copies of Windows NT, and
the Defense Information Systems Agency (DISA) plans to use Windows NT on
its 6,500-node DISAnet, which contains sensitive but not classified
information.
*****