OpenStep, waining MACH and a candid review of NeXT (LONG)
Brian Wright
Howdy,
I don't often post to these sorts of groups, but this has been
churning for a while now and I thought it about time to put it into words.
NeXT Computers, Inc. definitely has goal related problems. Every
business wants to make money AND stay in business. NeXT Computers is no
exception. However, their way of doing this is questionable and, to say
the very least, downright terrible. Take, for example, NeXTStep. A good
example of a great OS with little or no support. Why? Well, it's really
pretty simple. Price. It's always been priced way out of the consumer's
hands. When NeXT Computers produced hardware, the prices were outrageous
and you see where that got them. They had to stop producing hardware the
costs were so high (among other things). It's also not been priced well
enough (or structured well enough) to offer high end solutions to
corporations. A Pentium PC running NeXTStep isn't going to be enough to
power thousands of users attempting to run Oracle queries.
Most businesses don't see NeXTStep as an alternative in doing
business on it. Well, it makes perfect sense. For the high end side,
would you want to run databases that need 7*24 coverage on a box without
things like failover and high availability capabilities? I don't think
so. As for low end desktops it's not the perfect solution because it
requires significant investment on a per seat basis. Most companies would
rather just throw a PC on someone's desk preloaded with Windows (any
flavor) and be done with it. Why jump through hoops to get NS working on
a PC? Not to mention the training necessary to get the PC techs up to
speed in dealing with a UNIX based OS.
And, of course, we come to Stability. NeXTStep running under Mach
is also not the most stable OS in the world. I mean, you open a shell
unmount and mount an already mounted filesystem and Workspace manager has
fits. Also, adding things like PPP in can sometimes cause sporadic
crashes. A business would be hard pressed to tolerate this for long.
Especially not knowing if NeXT will fix these problems.
Now they only produce the OS NeXTStep, which is about to change to
OpenStep. What is the price of this OpSys? $799 the last time I looked.
Of course, that doesn't include the price of the hardware to run it on.
Running it on an Intel means you have to basically jump through hoops to
get it to work or pay an outrageous price from a 3rd party vendor to
produce you a NeXTStep ready PC. You could always opt for relatively
unsupported black hardware (which is becoming outdated) and pay a lower
price (relatively speaking). You could obtain a SparcStation (not cheap
in itself) and run NeXTStep Sparc (which I've heard has a few problems).
Or finally you could obtain an HP 7xx system of some sort (not cheap in
itself either).
So basically, to get into NeXTStep, you're not talking about a
cheap investment. Unlike Win95 or WinNT and PC hardware (not that they
compare in quality, but definitely is cheaper in overall price). Even a
Mac fully loaded is still cheaper than NeXTStep after hardware + OS price.
Ok, I will admit if you go the educational route you can save some bucks
on the OS part, but you relenquish any rights to producing any for-sale
software if using NeXTStep Developer.
So now we come OpenStep. Will it be a savior or will it be a boat
anchor? I venture to guess, neither. It will flounder somewhere in the
middle and keep NeXT Computers running like they always have been. Making
just enough money to stay afloat, but not making enough to move it to the
next level. Of course, one major obstacle in the Win95/OpenStep solution
is Win95 itself. Win95 simply isn't robust enough to handle OpenStep
along with all of its associated overhead (along with tons of Windows apps
and NeXTStep applications). Placing something the size of OpenStep on top
of Win95 is bound to cause resource related problems, hitting GDI
ceilings, GPFs, etc (especially when programming). About the only OpSys
capable of handling this product will be Solaris (assuming this version
becomes available) and the Windows NT version. Of course, how many people
are running Windows NT? Probably as few as are running NeXTStep. Above
the technical problems associated with a product like OpenStep you've got
the acceptance level. The what-will-this-product-really-do-for-me kinds
of people. Well, to give an example we have to go back a few years. And I
won't go into the costs of obtaining Solaris capable systems for
tangential reasons.
A product on the Amiga called MUI came out and promised an easy to
use Object Oriented programming interface. The problem is, it adds a new
level of overhead to the OS that simply is, at times, unbearable. I can
see the same thing happening with OpenStep. The level of acceptance for
MUI to this day is a love-hate relationship. I realize that the scope of
MUI is MUCH smaller than that of OpenStep, but the concept is not new and
will go through much the same problems as did MUI. Some programmers will
think, "Why should I invest in OpenStep, when I can program the same thing
myself without all the overhead and large investments?" The users will
think, "Why should I invest in OpenStep just to get one or two
applications I don't already have?". That is knowing the costs to
purchase OpenStep may be significant. Especially if the applications may
already be available in a different form under Windows. And, of course,
NeXT has to decide who is going to pay for OpenStep. If I know NeXT,
they'll make both the users and the developers pay for it (and pay
significantly). This is really to the detriment of this product.
About the only people OpenStep will appeal to are those people who
NeXTStep already appeals to (which is kind of like advertising NeXTStep in
NEXTWORLD). Granted, I really like NeXTStep and its capabilities.
However, that is not enough for most people. They want to be able to
share things with their neighbor next door. It's hard to do when your
neighbor has a PC running Windows.
What's in store for NeXTStep/OpenStep in the future? Well,
probably not a whole lot. There are a lot of avid fans of NeXTStep who
will graduate slowly into OpenStep (because NeXT forces them to do this)
and griping will eventually subside. However, Jobs has his sites on the
entertainment industry at the moment while only glancing casually over at
NeXTStep to make sure things are still running smoothly. Now that Toy
Story has been such a success, there is no reason why he should spend
oodles and gobs of money on marketing for OpenStep when he can just let it
ooze its way around slowly like he has done for NeXTStep. He'll just keep
making movies for Disney and raking in the cash this way.
Is Jobs attempting to make OpenStep the OS of choice? No, I don't
think so. At this point, it as an alternate solution for those needing
(or wanting) an Object Oriented solution to their business dilemma.
Clearly he has not poised OpenStep to take over any market. And, it
seems, he has no intention to do so.
The one solution that seems to have elluded Jobs is the law of
supply and demand. If the price goes down, the demand goes up. He could
probably sell many more copies of NeXTStep and OpenStep by pricing that is
much more competative with todays other OSes. I don't hold much hope that
OpenStep will be the end-all-be-all solution, just another ploy by Jobs to
obtain a little piece of the Windows market. Even he sees the profits
that can be made from PC software.
There is a lot more that could be said about this entire
situation, but this is long enough already. Feel free to make any
comments you wish; however, all flames will be directed to /dev/null.
--
Brian Wright <wri...@extreme.intex.net>
NeXTMail accepted.
Joe Panico
Brian Wright (wright@extreme) wrote:
: Howdy,
: I don't often post to these sorts of groups, but this has been
: churning for a while now and I thought it about time to put it into words.
: capable of handling this product will be Solaris (assuming this version
: becomes available) and the Windows NT version. Of course, how many people
: are running Windows NT? Probably as few as are running NeXTStep. Above
I guess you don't read industry news. The results are in-- this year NT will
outsell all flavors of Unix combined. Many are predicting that it will overtake
Netware as the dominant server platform next year. I don't remember off the
top of my head, but I believe that NT will sell something like a half million
units this year.
: --
: Brian Wright <wri...@extreme.intex.net>
: NeXTMail accepted.
--
Joe Panico
NeXTStep/OpenStep Developer
BLaCKSMITH Inc.
William Grosso
In <jpanicoD...@netcom.com> jpa...@netcom.com (Joe Panico)
writes:
>I guess you don't read industry news. The results are in--
>this year NT will outsell all flavors of Unix combined. Many
>are predicting that it will overtake Netware as the dominant
>server platform next year. I don't remember off the
>top of my head, but I believe that NT will sell something like
>a half million units this year.
>
On the off chance that this isn't a joke...
"The results are in..."
For *this* year ???? Given press lag time, and
the problems gathering accurate data quickly, the
"results" are (at best) determined by data through
Feb 15.
The results are in on next year's Triple Crown too.
Moreover, If NT sells half a million copies, then it
will not outsell all flavours of Unix combined. Futhermore,
6 copies of OS/2 will be sold for every copy of NT
(at least. I'm not sure what the latest OS/2 figures are).
"Many are predicting that it [NT] will overtake
Netware as the dominant server platform."
Once again: THIS IS THE COMPUTER INDUSTRY. 4 years ago,
many were predicting that NT would be the dominant
platform by 1994. Many people thought that the Newton
would change computing forever (and maybe it will. But
certainly not yet). Many people predicted that Novell,
once it acquired WordPerfect, would successfully invade
Microsoft's Turf. Many people ......
Moreover, consider the OS/2 sales figures again. OS/2
is as much of a server OS as NT. Which would imply that,
if Novell were being toppled, it wouldn't be NT doing
the toppling.
The key, dear poster, is to stop and think when you
read the industry trade mags. Far more often than not,
the predictions are composed of equal portions of
shrill boosterism and shameless blue-skying.
Mindlessly repeating such vapid drool accomplishes nothing.
Cheers,
Andy
Joe Panico
William Grosso (apul...@ix.netcom.com) wrote:
: In <jpanicoD...@netcom.com> jpa...@netcom.com (Joe Panico)
: writes:
: >
: "The results are in..."
: For *this* year ???? Given press lag time, and
: the problems gathering accurate data quickly, the
: "results" are (at best) determined by data through
: Feb 15.
How do pollsters manage to predict, frequently with frightening accuracy,
the results of elections weeks in advance? If the numbers are large enough,
statistics and simple extrapolations can usually do the trick.
My source of information is not Computer Industry journalists, or press
releases from IT vendors (in particular the Dark Lord himself). I'm simply
quoting the results of *Market Research* firms. I'm not willing to plop down
the $300 it takes to buy the actually reports themselves (if you do this, you
will be providing us all with a very enlightening experience), but I have
read extracts of the reports from a number of sources.
: Moreover, If NT sells half a million copies, then it
: will not outsell all flavours of Unix combined. Futhermore,
: 6 copies of OS/2 will be sold for every copy of NT
: (at least. I'm not sure what the latest OS/2 figures are).
According to IDC you are wrong:
April 96 ComputerWorld Journal: p. 19, "Today, market research firms
*universally* agree that Windows NT will soon become the world's most popular
operating system in terms of units shipped."
p. 20 shows research from IDC (one of the two most respected sources
of this kind of information). There results show that in calendar '95,
Unix NT
1st 1/2 254K 143K
2nd 1/2 263K 270K
: "Many are predicting that it [NT] will overtake
: Netware as the dominant server platform."
: Once again: THIS IS THE COMPUTER INDUSTRY. 4 years ago,
: many were predicting that NT would be the dominant
The other big market research firm, DataQuest, reached a similar conclusion:
http://www.dataquest.com/register/abstracts/pcsw-ww/pcsw-ww-da-9601-ab-0001.html
San Jose, California; April 19, 1996GMicrosoft shipped 18.8 million units of Win
dows 95 in 1995, and the operating system is expected to more than triple in shi
pments to 62.7 million units in 1996, according to Dataquest's forecast of PC pl
atform operating systems. Windows 95 and Windows NT will make Microsoft the over
whelming leader in PC operating systems by the year 2000, surpassing the Macinto
sh OS, DOS, UNIX, NetWare (server licenses), and IBM's OS/2.
"Windows NT shows the most growth, and judging by the level of organizational in
terest, we can expect to see substantial Windows NT desktop installations by the
end of 1996," said Chris Le Tocq, director and principal analyst for Dataquest'
s Personal Computing Software service.
You can purchase the actual report for only $295. You would be doing a great
public service if you summarized it for us.
: The key, dear poster, is to stop and think when you
: read the industry trade mags. Far more often than not,
: the predictions are composed of equal portions of
: shrill boosterism and shameless blue-skying.
: Mindlessly repeating such vapid drool accomplishes nothing.
OK, now what are your sources of information to the contrary? Could you please
post references?
: Cheers,
Indeed.
: Andy
:
William Grosso
In <jpanicoD...@netcom.com> jpa...@netcom.com (Joe Panico)
writes:
>
>William Grosso (apul...@ix.netcom.com) wrote:
>: In <jpanicoD...@netcom.com> jpa...@netcom.com (Joe Panico)
>: writes:
>: >
>
>: "The results are in..."
>
>: For *this* year ???? Given press lag time, and
>: the problems gathering accurate data quickly, the
>: "results" are (at best) determined by data through
>: Feb 15.
>
>How do pollsters manage to predict, frequently with
>frightening accuracy, the results of elections weeks
>in advance? If the numbers are large enough, statistics
>and simple extrapolations can usually do the trick.
>
Uh. Joe ? They're not all that accurate. The
way newspapers have got around this in recent years
is by creating a large pool of "undecideds" and then
talking about "how the undecideds will break."
In First-year physics, you called "undecideds" a
"fudge factor."
Anyroad, predicting an election a week beforehand is not
analagous to predicting OS sales 10 months beforehand.
>I'm simply quoting the results of *Market Research* firms.
>I'm not willing to plop down the $300 it takes to buy the
>actual reports themselves (if you do this, you will be
>providing us all with a very enlightening experience), but
>I have read extracts of the reports from a number of sources.
Bad news here too, Joe. The "Market Research Firms" (IDC,
Dataquest, Gannet) are all pretty darn bad at predicting.
Since life is short, I'll simply focus on Dataquest (since
you cited them more than the other firms).
Let's look at their short-term work. According to Dataquest,
Windows 95 sold 18.8 million copies in 1995. I can recall
hearing predictions of 60 million copies. But maybe that
wasn't DQ.
DQ did predict 30 million. On August 23, 1995 (numerous sites
on the net have archived this fact-- I'd actually cut and pasted
a URL, but then Win95 crashed), DQ came out and with all the
force of their immense reputation, stated that Win95 would
sell 30 million copies.
Ooops. Off by a factor of 2, in August. I'd call that a
significant error, wouldn't you Joe ? One that, by itself,
is enough to cast serious doubt on their predictions
for the coming year ?
Now, let's look at their long term work. In 1992, DQ said
that, in 1996, Windows NT would have 40% of the OS market.
Ooops. Gosh. Those market research specialists (in business
for 24 years) got it wrong again. Off by a factor of 40.
(Cf: this month's Byte).
You see why the sober among us don't put much credence in
DQ's forecast for the year 2000 (which you so lovingly
quoted) ?
For more data on Dataquest, free for the asking, use dejanews
and search on dataquest. Th OS/2 people are mighty pissed--
it seems like each year is the last year for OS/2, according to
DQ. And, of course, as far as network operating systems go,
somehow OS/2 doesn't even register as a blip on the DQ charts.
The summary: you're completely wrong to trust market
research firms; they have a history of being totally
clueless; I liked the lemming analogy.
Cheers,
Andy
brian wright
Joe Panico (jpa...@netcom.com) wrote:
: I guess you don't read industry news. The results are in-- this year NT will
: outsell all flavors of Unix combined. Many are predicting that it will overtake
: Netware as the dominant server platform next year. I don't remember off the
: top of my head, but I believe that NT will sell something like a half million
: units this year.
My my... How we digress so quickly.
Ok, I probably should have qualified that. There are many companies out
there using NT, but at least the last 3 companies I've worked for, there
has only been a spattering of NT boxes. Further, many of those companies
who are using NT are basically using it to replace Novell within their
organization. Some are actually using NT for productivity. However, I
venture to guess you will find many more Win95 boxes in a given company
than NT boxes. And even further, how many people really run NT at home?
I would guess there are more OS/2 boxes in homes than NT. Which is what
I was referring to and should have qualified.
For the same reason that NeXTStep hasn't been accepted by the industry as
a full featured robust operating system for all out heavy production, so
NT will also fail in that area as well. NT will quickly move to replace
Novell because who wants a NOS when they can have a real OS. Beyond being
a glorified NFS equivilent and being able to run SOME windows apps, NT
will be searching for a market to hit. That is, until Gates decides to
outmode Win95 and its successors and force everyone to use NT Workstation
as their default environment. Plus, NT still doesn't have any kind of a
fat binary system to allow cross architecture portability of binaries.
However, this is tangential to my original post. This is very much par
for the course, though. Someone decides they have to pick out one
piece of information and attempt rip it to shreds when clearly
that wasn't even close to the original intent of the original article.
And need I mention that this is a comp.sys.NEXT.advocacy? As such, I
won't be replying to any other posts dealing strictly with the topic of
another OS when taken out of the context of my original posting.
As I mentioned before, all flames to /dev/null.
: Joe Panico
: NeXTStep/OpenStep Developer
: BLaCKSMITH Inc.
--
Brian C. Wright <wri...@extreme.intex.net>
NeXTMail Accepted.
--
--
Brian C. Wright
Southwestern Bell Mobile Systems -- UNIX Systems Admin/Support
Joe Panico
William Grosso (apul...@ix.netcom.com) wrote:
: In <jpanicoD...@netcom.com> jpa...@netcom.com (Joe Panico)
: writes:
: On the off chance that this isn't a joke...
: "The results are in..."
: For *this* year ???? Given press lag time, and
: the problems gathering accurate data quickly, the
: "results" are (at best) determined by data through
: Feb 15.
: The results are in on next year's Triple Crown too.
One more thing. I don't think the Triple Crown anology is a good one. Much
better would be a lemming analogy. You are standing by the ege of a cliff,
and a giant herd of lemmings are running towards. You happen to have observed
that every several years, on that same spot, a giant herd of lemming jumps
over the edge. The lemmings are quickly approaching the edge and you conclude
that....
...they will veer away at the last minute and run to safety.
Interesting conclusion.
--
Felipe A. Rodriguez
In article <4n5d5t$j...@dfw-ixnews1.ix.netcom.com> apul...@ix.netcom.com(William Grosso)
writes:
>In <jpanicoD...@netcom.com> jpa...@netcom.com (Joe Panico)
>writes:
>>
>>William Grosso (apul...@ix.netcom.com) wrote:
>>: In <jpanicoD...@netcom.com> jpa...@netcom.com (Joe Panico)
>>: writes:
>>: For *this* year ???? Given press lag time, and
>>: the problems gathering accurate data quickly, the
>>: "results" are (at best) determined by data through
>>: Feb 15.
>>
>>frightening accuracy, the results of elections weeks
>>in advance? If the numbers are large enough, statistics
>>and simple extrapolations can usually do the trick.
>>
>
>Let's look at their short-term work. According to Dataquest,
>Windows 95 sold 18.8 million copies in 1995. I can recall
>hearing predictions of 60 million copies. But maybe that
>wasn't DQ.
>
>DQ did predict 30 million. On August 23, 1995 (numerous sites
>on the net have archived this fact-- I'd actually cut and pasted
>a URL, but then Win95 crashed), DQ came out and with all the
>force of their immense reputation, stated that Win95 would
>sell 30 million copies.
>
>Ooops. Off by a factor of 2, in August. I'd call that a
>significant error, wouldn't you Joe ? One that, by itself,
>is enough to cast serious doubt on their predictions
>for the coming year ?
>
Well now that we're bashing DataQuest I'll toss in my two
cents worth.
Last year DQ announced that there would be a shortage of silicon
wafers this year. Oops. As it turns out we're into May and it
seems the Book to Bill is running at 9 year lows of .78.
Silicon Shortage Forecast to Hit Next Year, According to Dataquest_
(December 7, 1995)
http://www.dataquest.com/irc/press/ir-n9548.html
Another one of their grand predictions was that this year
would shape up to be one of the most profitable for the
drive business. So far? Guess again.
Dataquest Predicts Record Profits for Disk Drive Makers Next Year_
(December 11, 1995)
http://www.dataquest.com/irc/press/ir-n9549.html
I think the aformentioned are indicative of how innacurate the
fortune teller's crystal ball actually is.
Nonetheless, I won't argue anyone's numbers. I would like to
make a personal and highly subjective observation. It has
been my experience that NT usually displaces Win95, DOS,
NetWare, OS2 or the MacOS. In cirumstances where you need the power
and flexibility of UNIX, NT will probably fall short in most respects.
--
Felipe A. Rodriguez # Francesco Sforza became Duke of Milan from
Agoura Hills, CA # being a private citizen because he was
# armed; his successors, since they avoided
f...@ix.netcom.com # the inconveniences of arms, became private
(NeXTmail preferred) # citizens after having been dukes.
(MIMEmail welcome) # --Nicolo Machiavelli
William Grosso
In <4n5kj9$i...@dfw-ixnews3.ix.netcom.com> f...@ix.netcom.net(Felipe A.
Rodriguez) writes:
>
>Nonetheless, I won't argue anyone's numbers. I would like to
>make a personal and highly subjective observation. It has
>been my experience that NT usually displaces Win95, DOS,
>
Actually, you can do this with numbers. Take 1992
sales, convert them to percentages, and then
do something like
1992 1996
"Windows NT --> Windows NT"
"Macintosh --> Macintosh"
"Unix --> Unix"
"Windows 3 --> Windows upgrade (combination of
and DOS" 3.x and 95)"
Wanna bet that this "model" (percentages staying roughly
the same) does better than any of the "forecasts"
propounded by the market-research firms ?
Anybody interested in setting up a Web page to track
the market-research predictions ? I'd be willing to
help doublecheck the data etcetera (I'm actually
curious to see how well they do overall, instead of
just on their well-publicized predictions).
Cheers,
Andy
Donald R. McGregor
In article <4n5kj9$i...@dfw-ixnews3.ix.netcom.com>,
Felipe A. Rodriguez <f...@ix.netcom.net> wrote:
>Nonetheless, I won't argue anyone's numbers. I would like to
>make a personal and highly subjective observation. It has
>been my experience that NT usually displaces Win95, DOS,
>and flexibility of UNIX, NT will probably fall short in most respects.
Probably true for the most part. You'd be nuts to replace a nice solid
smtp or nntp server with NT. OTOH, NT does have, potentially, an advantage
that NeXT used to tout: ONE computer on every desk. Run productivity
apps and do workstationy-things at the same time. I suspect that
some traditional workstation CAD and graphics programs will migrate to
NT, whereupon the workstation versions will die. PCs are getting scary-
fast, so there's no real horsepower advantage to using a workstation.
The OS's are equivalent in the services they provide. Networking is in
the same ballpark for most applications.
PC programming environments
are getting pretty damn good now that the 'reboot' has been taken out of the
traditional PC edit-compile-link-test-reboot cycle. With gnu-32 there's
a half-decent command line. I think I'd take an NT environment for programming
over a generic Unix box today.
--
Don McGregor | The wages of sin is death--but after taxes it's
mcg...@crl.com | only a tired feeling at the end of the day.
Joe Panico
William Grosso (apul...@ix.netcom.com) wrote:
: In <jpanicoD...@netcom.com> jpa...@netcom.com (Joe Panico)
: writes:
: Bad news here too, Joe. The "Market Research Firms" (IDC,
: Dataquest, Gannet) are all pretty darn bad at predicting.
: Since life is short, I'll simply focus on Dataquest (since
: you cited them more than the other firms).
[snip, why dataquest can't forecast]
These arguments are reasonable as far as they go-- forecasting is inherently
difficult and market research firms sometimes do a crappy job at it. We
all know first hand that the farther out you try to forecast, the more
unreliable your results will be. But these arguments also belie the more
compelling information that I posted-- the numbers from IDC. Those numbers
are not derived from crystal ball gazing, they come from looking through
the *rear* view mirror. The IDC numbers show that NT outsold Unix in the
2nd half of 95, and that the derivative is clearly on the side of NT;
NT sales are accelerating while Unix sales have flattened.
Customarily, when trying to refute someone's numbers , you produce numbers
of your own. Is it not reasonable to accept someones numbers unless you
can produce a counter argument based on a different set of empirical results?
Do you have any real evidence that IDC's '95 numbers are wrong?
Garance A Drosehn
wright@extreme (Brian Wright) wrote:
> Howdy,
Hi.
> NeXT Computers, Inc. definitely has goal related problems.
This could be true.
> When NeXT Computers produced hardware, the prices were outrageous
> and you see where that got them. They had to stop producing
> hardware the costs were so high (among other things).
When NeXTstations came out, there were actually a pretty good
deal for the hardware you were getting. NeXT dropped hardware
in Feb 1993. In April of 1994 I paid more for Intel-based
hardware than I would have paid for a NeXTstation color turbo,
and that hardware (bought a year later) does not seem all that
much faster than the NeXTstation.
> It's also not been priced well enough (or structured well enough)
> to offer high end solutions to corporations. A Pentium PC running
> NeXTStep isn't going to be enough to power thousands of users
> attempting to run Oracle queries.
NeXT has never pictured itself as a server company (well, not until
WebObjects, at least). In fact, it went out of it's way to discourage
anyone who had the notion of running NeXTSTEP on servers.
> Most companies would rather just throw a PC on someone's desk
> preloaded with Windows (any flavor) and be done with it. Why
> jump through hoops to get NS working on a PC? Not to mention
> the training necessary to get the PC techs up to speed in dealing
> with a UNIX based OS.
The frustrating thing is those same companies have no trouble
jumping thru hoops to get Microsoft operating systems running
on the same hardware. The big advantage isn't the operating
system so much as the applications run on top of the operating
system.
> And, of course, we come to Stability. NeXTStep running under
> Mach is also not the most stable OS in the world. I mean, you
> open a shell unmount and mount an already mounted filesystem and
> Workspace manager has fits.
Yeah, like this is something that every PC user does every 10
minutes. Geez.
> Also, adding things like PPP in can sometimes cause sporadic
> crashes. A business would be hard pressed to tolerate this for
> long.
Who are the two biggest pieces of the personal computer pie?
MacOS and Windows. The MacOS certainly has had troubles once
you try to hookup PPP (which may finally be addressed by the
recent system update and the latest version of FreePPP), and
until Windows95 came along it too had plenty of headaches with
SLIP or PPP.
> Now they only produce the OS NeXTStep, which is about to change
> to OpenStep. What is the price of this OpSys? $799 the last
> time I looked. Of course, that doesn't include the price of the
> hardware to run it on. Running it on an Intel means you have to
> basically jump through hoops to get it to work or pay an outrageous
> price from a 3rd party vendor to produce you a NeXTStep ready PC.
Or you could get on the nextstep/intel homebrew mailing list.
> You could always opt for relatively unsupported black hardware
> (which is becoming outdated) and pay a lower price (relatively
> speaking). You could obtain a SparcStation (not cheap in itself)
> and run NeXTStep Sparc (which I've heard has a few problems).
> Or finally you could obtain an HP 7xx system of some sort (not
> cheap in itself either).
Unfortunately it'd be a real bad idea to buy into NeXTSTEP for HP
at this point. NeXT and HP have already stated that there will be
no NeXTSTEP 4.0 for HP's PA-RISC architecture. And NS/Sparc doesn't
work on all Sun's hardware either, so that option isn't looking as
wonderful as it once did. Given that the newest NeXTstation is at
least three years old, it'd be rather questionable that you'd want
to start something serious (such as a business) on that hardware
at this time -- even though it is fine hardware for the time it
was made.
Which tends to leave us with just NS/Intel.
> So now we come OpenStep. Will it be a savior or will it be
> a boat anchor? I venture to guess, neither. It will flounder
> somewhere in the middle and keep NeXT Computers running like they
> always have been. Making just enough money to stay afloat, but
> not making enough to move it to the next level.
I suspect this will be true.
> About the only people OpenStep will appeal to are those
> people who NeXTStep already appeals to (which is kind of like
> advertising NeXTStep in NEXTWORLD). Granted, I really like
> NeXTStep and its capabilities. However, that is not enough for
> most people. They want to be able to share things with their
> neighbor next door. It's hard to do when your neighbor has a PC
> running Windows.
OpenStep for WindowsNT will appeal to some people that NeXTSTEP
does not appeal to. Running OpenStep/WinNT for one application
does not prevent you from running other WindowsNT applications,
so you should be able to share things more than you would with
NeXTSTEP.
And for me, OpenStep for MachOS still appeals to me more than
anything on WindowsNT will. So, hopefully, NeXT will have a
somewhat larger market with the two products than it did with
just the MachOS product.
> Is Jobs attempting to make OpenStep the OS of choice?
> No, I don't think so. At this point, it as an alternate
> solution for those needing (or wanting) an Object Oriented
> solution to their business dilemma. Clearly he has not poised
> OpenStep to take over any market. And, it seems, he has no
> intention to do so.
OpenStep isn't an OS. OpenStep on MachOS does include an OS,
but the other OpenStep's are additions to an operating system
that the user has already chosen.
> The one solution that seems to have elluded Jobs is the
> law of supply and demand. If the price goes down, the demand
> goes up. He could probably sell many more copies of NeXTStep
> and OpenStep by pricing that is much more competative with todays
> other OSes. I don't hold much hope that OpenStep will be the
> end-all-be-all solution, just another ploy by Jobs to obtain a
> little piece of the Windows market. Even he sees the profits
> that can be made from PC software.
I do wish the prices were lower now that the products from NeXT
are software-only. It's better to make a little money over a lot
of users than trying to make a lot of money off a few users.
Given that NeXT seems to be putting more emphasis on their OpenStep
for WindowsNT product than the MachOS product, one might wonder
how much commitment they have to making future major upgrades to
the MachOS portion. I would like to see an OpenStep for FreeBSD
product. This would leave the operating system issues to people
who want to work on them (and who don't mind giving out the source
code), but pick up the user interface that has worked well (for
me) in NeXTSTEP. Note that this is different than NeXT upgrading
their OS to include the latest BSD kernel and networking.
---
Garance Alistair Drosehn = g...@eclipse.its.rpi.edu
Senior Systems Programmer (MIME & NeXTmail capable)
Rensselaer Polytechnic Institute; Troy NY USA
Charles William Swiger
Excerpts from netnews.comp.sys.next.advocacy: 13-May-96 Re: OpenStep,
waining MACH .. by Garance A Drosehn@eclips
> deal for the hardware you were getting.
NeXT's base hardware models were good value for the money, I certainly
agree. However, NeXT charged between 2 and 3 times the normal street
price for add-on items like RAM and hard drives.
>> It's also not been priced well enough (or structured well enough)
>> to offer high end solutions to corporations. A Pentium PC running
>> NeXTStep isn't going to be enough to power thousands of users
>> attempting to run Oracle queries.
>
> NeXT has never pictured itself as a server company (well, not until
> WebObjects, at least). In fact, it went out of it's way to discourage
> anyone who had the notion of running NeXTSTEP on servers.
Agreed. NeXT has long avoided making changes to Mach that would allow
real scalability improvements, such as SMP support, support for
user-space Mach memory objects, upgrades to NFS, upgrades to sendmail,
and so forth. NEXTSTEP/SPARC does not run on Sun's higher-end hardware,
like the Hyper- and UltraSPARC systems.
However, there have been a few signs that NeXT is working to improve
this. NS 4.0 will include an updated sendmail, at least, even if other
Mach improvements won't be there. And OPENSTEP running on other
operating systems like NT or Solaris should be able to run on a broader
range of hardware. It might be pretty cool to see OPENSTEP running on
higher-end parallel machines.
>> And, of course, we come to Stability. NeXTStep running under
>> Mach is also not the most stable OS in the world. I mean, you
>> open a shell unmount and mount an already mounted filesystem and
>> Workspace manager has fits.
>
> Yeah, like this is something that every PC user does every 10
> minutes. Geez.
I've had NS 3.3p1 be completely rock-stable on black hardware and some
Intel hardware (if you've got well-supported Intel hardware, anyway).
30 day uptimes are pretty common, with the machines being voluntarily
rebooted more often than they crash.
[ ... ]
> Given that NeXT seems to be putting more emphasis on their OpenStep
> for WindowsNT product than the MachOS product, one might wonder
> how much commitment they have to making future major upgrades to
> the MachOS portion. I would like to see an OpenStep for FreeBSD
> product. This would leave the operating system issues to people
> who want to work on them (and who don't mind giving out the source
> code), but pick up the user interface that has worked well (for
> me) in NeXTSTEP. Note that this is different than NeXT upgrading
> their OS to include the latest BSD kernel and networking.
How about OPENSTEP/Linux? Or perhaps OPENSTEP (or, failing that,
GNUSTEP) over the HURD kernel? HURD is a decendant of Mach....
-Chuck
Charles Swiger | cs...@andrew.cmu.edu | standard disclaimer
----------------+---------------------+---------------------
I know you're an optimist if you think I'm a pessimist.
mmalcolm crawford
On 05/11/96, Brian Wright wrote:
> NeXT Computers, Inc. definitely has goal related problems. Every
> business wants to make money AND stay in business. NeXT Computers is no
> exception. However, their way of doing this is questionable and, to say
> the very least, downright terrible. Take, for example, NeXTStep. A good
> example of a great OS with little or no support. Why? Well, it's really
> pretty simple. Price. It's always been priced way out of the consumer's
> hands. When NeXT Computers produced hardware, the prices were outrageous
> and you see where that got them.
>
If you look at what you got for your $3995 way back when (1991 or
thenabouts), compare the price of a similarly-speced Mac (if you could find
such a thing), and made price your sole criterion, you'd have bought the NeXT
every time.
Some people were prepared to pay a lot more for Suns round this time too.
This argument has been rehashed more times than I can remember, and everyone
has their own take on it. Mine is that the problem was lack of intelligent
marketing -- it seemed to me on this side of the pond that the people NeXT
signed up as resellers just didn't understand the product, wanted to push
boxes as fast as they could as if they were PCs or Macs, and invested no time
or effort in educating the market into what NeXT was about (NeXT's own lack
of understanding of what NeXT was about didn't help here, of course).
Others will have different ideas, and I suspect things have got to the stage
where noone is likely to change their opinion, so let's leave this one.
> They had to stop producing hardware the costs were so high (among other
> things).
>
Again, I have a different take on this. As far as I understand it, the cost
of the hardware was actually remarkably low -- the old factory was, I
believe, one of the most efficient on the planet. And just look inside the
box -- not a lot of wastage there. One of the most elegant designs I've ever
seen.
> It's also not been priced well
> enough (or structured well enough) to offer high end solutions to
> corporations. A Pentium PC running NeXTStep isn't going to be enough to
> power thousands of users attempting to run Oracle queries.
>
If you're trying to get the Pentium to act as an Oracle *server*, then you
deserve what you'll get. NeXT has said consistently that they don't do
servers. A network of 1000 Pentium PCs running NEXTSTEP as Oracle *clients*,
now that makes sense.
> Most businesses don't see NeXTStep as an alternative in doing
> business on it. Well, it makes perfect sense. For the high end side,
> would you want to run databases that need 7*24 coverage on a box without
> things like failover and high availability capabilities? I don't think
> so.
>
See my point above -- NeXT isn't trying to sell in this market.
> As for low end desktops it's not the perfect solution because it
> requires significant investment on a per seat basis. Most companies would
> rather just throw a PC on someone's desk preloaded with Windows (any
> flavor) and be done with it. Why jump through hoops to get NS working on
> a PC? Not to mention the training necessary to get the PC techs up to
> speed in dealing with a UNIX based OS.
>
Sure, most companies would rather do the "easy" thing, accepting the lowest
common standard demoninator. That's the tragedy. If businesses looked for
quality, Macs would have won out first. Instead they've bought cheap garbage
at the outset, and paid for it in recurrent costs of maintenance, support,
and in particular employees' time. I would dearly love to see a cost/benefit
analysis of NEXTSTEP vs. Windows from the *user's* perspective -- I believe
at least one was carried out in London, and it confirmed my opinion that
people are a lot more productive using NS. Now, cost $599 for s/w vs
$10-20,000 for salary...
> And, of course, we come to Stability. NeXTStep running under Mach
> is also not the most stable OS in the world. I mean, you open a shell
> unmount and mount an already mounted filesystem and Workspace manager has
> fits. Also, adding things like PPP in can sometimes cause sporadic
> crashes. A business would be hard pressed to tolerate this for long.
> Especially not knowing if NeXT will fix these problems.
>
Umm, whilst claiming "At least we're not as bad as the other guys" is not the
best marketing slogan in the world, at least in my experience NS is a *lot*
more stable than Windows. Have you had different experiences?!
> Now they only produce the OS NeXTStep, which is about to change to
> OpenStep.
>
Umm, no, they also do (or will do) OpenStep for Windows(NT and maybe 95,
others?), PDO, NetInfo, WebObjects...
> What is the price of this OpSys? $799 the last time I looked.
> Of course, that doesn't include the price of the hardware to run it on.
> Running it on an Intel means you have to basically jump through hoops to
> get it to work or pay an outrageous price from a 3rd party vendor to
> produce you a NeXTStep ready PC.
>
No more outrageous than a PC capable of sensibly running Windows95. NS and
W95 have approximately the same system requirements, as far as I can tell.
[...]
> So basically, to get into NeXTStep, you're not talking about a
> cheap investment. Unlike Win95 or WinNT and PC hardware (not that they
> compare in quality, but definitely is cheaper in overall price).
>
You've really made my argument for me here... if you want quality, go for
NEXTSTEP. Trouble is, not many people seem to be sufficiently Promethean to
invest in quality (and thereby lower their overall costs).
I may get round to responding to the rest of your message later, but it's a
bit too woolly to get to grips with at the moment.
Best wishes,
mmalc.
--
William Grosso
>produce numbers of your own. Is it not reasonable to accept
>someones numbers unless you can produce a counter argument
>based on a different set of empirical results?
No, it is not reasonable. If a given firm has a horrible
track record, I am not required to produce alternative
numbers before I reject their projections.
Especially since my claim, over this and other threads,
has been that long-term forecasting is, by and large, a
futile endeavor. In another thread, I pointed out that
"projecting" current market-shares (e.g. take the status
quo and upgrade the version numbers) has done better than
DQ's forecasts.
>Do you have any real evidence that IDC's '95 numbers are wrong?
Depends on what you mean by "real." I'm thinking this
is a tad unfair of you, though. You posted, I picked
a market-research firm and, by merely browsing the
web for 20 minutes, was able to completely refute their
numbers. You respond: "Yeah. Well, that was the easy case."
One obvious flaw in the IDC numbers: They're measuring
sales, not use. Which means that, up to some fairly small
percentage, they're really measuring "sales of high-end
PCs."
If you want to buy a high-end PC to run Unix, there's a
good chance that it's gonna come with NT already on the
drive. Ditto for OS/2 (which sold a million copies last
December, none of them bundled).
I've been looking into buying a fast pentium for home
use. moe companies give me a choice: W95 or NT. Most
default to NT (I'd have to ask to get W95).
On them other hand, the overwhelming majority of Unix
purchases involve people who are going to use Unix.
Another obvious flaw: They're not counting all the Unixes.
FreeBSD and Linux aren't counted in those numbers.
And while you might think "Linux ? That's not a server OS.
No truly mission-critical stuff runs on freeware" the fact
of the matter is that I can think of a dozen small companies
here in the valley that use Pentiums with Linux as their
fileservers (and, if I can think of a dozen, it means the
practice is fairly widespread among small tech companies).
Wanna guess what percentage of web-sites are running Linux ?
That's right, folks: Unix is the entry-level server OS.
Bodes quite well for Unix (if not for, say, Solaris).
The point: The IDC evidence is amazingly thin. I'd say it
points towards the (already well-known) trend of Intel
boxes replacing HP/SUN/DEC/IBM workstations more than
anything else.
Moving on to more general topics: The "future of servers."
Because, really, I replied to your original post because
you cited Dataquest's absurd prediction that NT will
be the dominant server OS by the year 2000.
I've now pointed out that DQ has yet to produce an accurate
forecast.
I've now pointed out that there are severe flaws in the
way IDC measures usage (unless I've misunderstood the
IDC numbers).
I've pointed out before, and will reiterate here, that
OS/2 deserves to be considered. It sold more copies
last December than NT has sold, period.
Etcetera. Maybe NT has momentum. Maybe it doesn't.
We just can't tell from the numbers you've posted.
Cheers,
Andy
d...@misckit.com
I basically agree with Chuck's points, but there's one I'd
like to comment on a little:
Charles William Swiger <cs...@andrew.cmu.edu> wrote:
> NeXT's base hardware models were good value for the money, I certainly
> agree. However, NeXT charged between 2 and 3 times the normal street
> price for add-on items like RAM and hard drives.
While this is true, it must also be remembered that NeXT was only
pricing the disk and RAM the same way HP and Sun were doing at the
time. I was at BYU and saw the HP price lists, and their prices
were typically inflated 33% or more above what NeXT's rates were!
(And those were edu prices...) In the workstation world, it was
more common to buy disk and RAM from third parties. That's what I
did then, and I still shop around now. It pays off. I guess the
whole point of the pricing model was that NeXT wanted to sell
machines and not have to bother with disk and RAM. So they'd
price things that way to encourage you to buy the commodity
products from someone who specialized in them.
So, anyone who had any sense about them would get the most stripped
down box they could and then beef it up themselves. Worked great
for me and I don't regret a penny of what I spent. (I have three
NeXT boxes of my own, FWIW.)
> I've had NS 3.3p1 be completely rock-stable on black hardware and some
> Intel hardware (if you've got well-supported Intel hardware, anyway).
> 30 day uptimes are pretty common, with the machines being voluntarily
> rebooted more often than they crash.
Yup. My web and file server simply doesn't crash. I reboot it
occasionally on purpose, but other than that it is rock solid. The
only forced reboots have to do with power outages (Utah power is
really flakey). In fact, my ISP, running Solaris on Sun machines,
tends to crash hard about weekly as of late. In that light,
NEXTSTEP seems to be doing well for me... :-)
--
Later,
-Don Yacktman
d...@misckit.com
<a href="http://www.misckit.com/don.html">My home page</a>
Uli Zappe
> OpenStep, waining MACH .. by Garance A Drosehn@eclips
> > product. This would leave the operating system issues to people
> > who want to work on them (and who don't mind giving out the source
> > code), but pick up the user interface that has worked well (for
> > me) in NeXTSTEP. Note that this is different than NeXT upgrading
> > their OS to include the latest BSD kernel and networking.
>
> GNUSTEP) over the HURD kernel? HURD is a decendant of Mach....
Well, the problem with GNUSTEP so far would be that it DOES NOT implement
"the user interface that has worked well [...] in NeXTSTEP" but rather makes
use of X11.
A GNUSTEP that would reproduce the NEXTSTEP GUI could well be THE solution,
but alas...
NeXT offering a "full NEXTSTEP w/o the Mach kernel" to run on FreeBSD (or
Linux, for that matter) might be a good alternative, but who knows if they
make up their minds?
By
Uli
--
_____________________________________________________________________
Uli Zappe
Lorscher Strasse 5
D-60489 Frankfurt
Germany
Phone: +49 (69) 9784 0007
E-Mail: u...@tallowcross.uni-frankfurt.de (NeXTMail - Mime - ASCII)
PGP: public key on request
Member of WiNG (Wiesbaden NEXTSTEP Group)
_____________________________________________________________________
Ronp
In article <4n7il3$n...@bignews.shef.ac.uk>,
mmalcolm crawford <m.cra...@shef.ac.uk> wrote:
>If you're trying to get the Pentium to act as an Oracle *server*, then you
>deserve what you'll get. NeXT has said consistently that they don't do
>servers. A network of 1000 Pentium PCs running NEXTSTEP as Oracle *clients*,
>now that makes sense.
It probably made sense at the time for NeXT to avoid competing with
big iron (Sun) and the DataBase vendors (Oracle), but if NeXT is
unwilling to competer with SCO, NeXT won't have enough volume to
establish a three tier object oriented architecture. If there are
going to be distributed object servers, will they be running Windows
NT?
I thought that the advantage of using Mach was that messaging between
objects was more efficient. If we use a more "primitive" operating
system kernel, object messaging will be slower, making it more
difficult to use object technology which allows scaling by adding
more loosely coupled processors.
Ron
Ken Stagg
In article <4n2pf4$a...@animal.intex.net> wright@extreme (Brian Wright) writes:
....
> The one solution that seems to have elluded Jobs is the law of
>supply and demand. If the price goes down, the demand goes up.
No, No, No! If the price goes down the QUANTITY DEMANDED goes up. This
is NOT just an academic distinction. I would dearly love to see something
done to increase the demand for NeXTSTEP/OpenSTEP, but that is more likely
to come about due to a technological change or a change in consumer
preferences. Despite your desparaging remarks concerning NeXTSTEP's
stability it is miles ahead of the preeminent desktop OS's. If consumers
get fed up enough with Winsnooze crashes then we might see an increase in
demand for NeXTSTEP/OpenSTEP. And pigs will fly...
....
>Brian Wright <wri...@extreme.intex.net>
>NeXTMail accepted.
-Ken
Paul Lynch
> business wants to make money AND stay in business. NeXT Computers is no
> exception. However, their way of doing this is questionable and, to say
> the very least, downright terrible.
Good so far. Biggest problem is that they move onto a new product and
drop everything else before it is released. They drop hardware four
months before releasing NS/I; they stop actively selling NeXTSTEP over a
year before OpenStep/NT is ready, and now they are concentrating on
WebObjects.
{price argument deleted}
This whole pricing thing isn't true. People were saying that price is the
problem before NeXT dropped hardware, and it wasn't true then, either.
There may be a perception problem relating to price, but price wasn't the
issue then. mmalc has just replied to this, and he's covered it well.
> And, of course, we come to Stability. NeXTStep running under Mach
> is also not the most stable OS in the world. I mean, you open a shell
> unmount and mount an already mounted filesystem and Workspace manager
has
> fits. Also, adding things like PPP in can sometimes cause sporadic
> crashes. A business would be hard pressed to tolerate this for long.
NeXTSTEP has been the stablest OS by far that most of my customers have
used on the desktop. In many cases it has been a significant factor in
why they bought NeXTSTEP. I also haven't had problems with PPP, although
I recognise that some people say otherwise.
> Especially not knowing if NeXT will fix these problems.
Sloppy support has been a problem, true, but not in the way that you mean.
My experience has been that once you have the attention of an engineer at
NeXT, they will work hard for you. Getting their attention is a bizarre
problem, as NeXT will refuse to talk to you and put other obstacles in
your way.
> Of course, that doesn't include the price of the hardware to run it on.
> Running it on an Intel means you have to basically jump through hoops to
> get it to work or pay an outrageous price from a 3rd party vendor to
> produce you a NeXTStep ready PC.
HP and SPARC aren't really supported; HP won't have 4.0, and the latest
SPARC chips won't run NeXTSTEP.
I don't consider the prices you have to pay to get a well sorted NeXT
machine outrageous. It is hard work requiring a good knowledge of PC
hardware. Assembling things yourself is reasonably easy, if you have that
knowledge; if you don't, then you shouldn't grudge the extra cost.
Margins in (Intel) hardware are very tight at the best of times.
> Ok, I will admit if you go the educational route you can save some bucks
> on the OS part, but you relenquish any rights to producing any for-sale
> software if using NeXTStep Developer.
I consider the academic deal to be excellent value for money. In fact,
for an effectively unsupported OS, it is very reasonable. I don't
consider the full developer price to be justified, for the reasons that
you claim. The market has moved on, and it did it a long time ago.
> Of course, how many people
> are running Windows NT? Probably as few as are running NeXTStep.
More than you think, it would seem, at least in the corporate market. NT
seems to be doing quite well. And the corporate market has always been
the one to have money to spend on PC software; look at 1-2-3 and Notes,
for example. Not a personal market product, but doing OK.
> Some programmers will
> think, "Why should I invest in OpenStep, when I can program the same
thing
> myself without all the overhead and large investments?" The users will
> think, "Why should I invest in OpenStep just to get one or two
> applications I don't already have?".
NeXTSTEP has always been for custom apps. Nobody is about to buy
OpenStep/NT for applications. The answer to the overhead question won't
be answered until the release, as all the people who could say how much of
a drag it is are under NDA.
The cost overhead is another matter, and that can be answered now.
OpenStep/NT will not be viable for a horizontal product of any nature. It
might be possible to bundle it with a very high cost vertical product, but
that's all. And the latest cost reports of OpenStep/Solaris look like it
is heading the same way.
> About the only people OpenStep will appeal to are those people who
> NeXTStep already appeals to (which is kind of like advertising NeXTStep
in
> NEXTWORLD).
This is quite correct; and it is interesting that this will have come
about because of marketing considerations, not technical ones.
> Now that Toy
> Story has been such a success, there is no reason why he should spend
> oodles and gobs of money on marketing for OpenStep when he can just let
it
> ooze its way around slowly like he has done for NeXTStep. He'll just
keep
> making movies for Disney and raking in the cash this way.
Except that reports have been that Jobs is very hands off at Pixar, and
still spends most of his time with NeXT. Marketing at NeXT has always
been very conservative, and totally unsuccessful as a result. Big dollar
spends would just be an extension of the conservative marketing strategy,
and in my opinion would be equally misplaced.
> Is Jobs attempting to make OpenStep the OS of choice? No, I don't
> think so. At this point, it as an alternate solution for those needing
> (or wanting) an Object Oriented solution to their business dilemma.
NeXT have effectively dropped NeXTSTEP to play with the Web. They have
dropped retail models to sell direct; they might even make more money (as
a smaller company) that way. What scares me (still!) is the opportunities
missed.
NeXTSTEP is a viable OS that some people say is the best in the world. It
doesn't get any attention from NeXT, and hasn't done for a long time.
NeXT have other fish to fry. At one stage it looked like they might have
ported it on top of BSD 4.4, but that seems to have been dropped. An
OpenStep for FreeBSD would be great, but not just the pathetically minimal
OpenStep for NT; it needs all the apps and the GUI. But I don't see NeXT
doing this; the best thing that could happen would be for them to let it
go.
Paul
--
Paul Lynch (NeXTmail)
http://www.plsys.co.uk/~paul
Garance A Drosehn
Charles William Swiger <cs...@andrew.cmu.edu> wrote:
> Excerpts from netnews.comp.sys.next.advocacy: 13-May-96 Re: OpenStep,
> waining MACH .. by Garance A Drosehn@eclips
> > deal for the hardware you were getting.
>
> certainly agree. However, NeXT charged between 2 and 3 times
> the normal street price for add-on items like RAM and hard drives.
NeXT, and just every other manufacturer of personal computers and
workstations. Apple was charging absurd prices for hard disks and
additional RAM. If I remember right, you used to be able to buy
an external 1-gig disk for less than Apple charged if you wanted
an extra 200-meg for your internal disk. Apple got really burned
on RAM once, and for years after that their RAM prices were utterly
absurd. NeXT played the same game, but it's not like they were
the only ones doing it at the time.
> >> And, of course, we come to Stability. NeXTStep running under
> >> Mach is also not the most stable OS in the world. I mean,
> >> you open a shell unmount and mount an already mounted filesystem
> >> and Workspace manager has fits.
> >
> > Yeah, like this is something that every PC user does every 10
> > minutes. Geez.
>
> I've had NS 3.3p1 be completely rock-stable on black hardware
> and some Intel hardware (if you've got well-supported Intel
> hardware, anyway). 30 day uptimes are pretty common, with the
> machines being voluntarily rebooted more often than they crash.
I'd like to echo this sentiment. I've had NeXTstations where the
most common reason for a system reboot was that we lost power to
the building the NeXTstation is in. Ones where the *average*
uptime is over 50 days -- and this is on a machine which *is*
providing various campus-wide services (and anon-ftp for anyone
to connect to).
Of course, some of those service daemons finally got large enough
that I now run into the dreaded ever-increasing swapfile problem,
so now my uptime is more like four weeks. My it sucks to have a
machine that I have to reboot it every month. Oh the pain, oh the
pain. Yes indeed, I just wish I had something as stable as Windows
[<insert coughing spasm here>].
I wasn't doing quite that well on my NS/Intel box, but now that
I've got patch 1 for NS-3.3 that machine also seems to be staying
up for more than a month at a time. The last few reboots have been
just to add or remove devices to it, except for that time I tried
to open a 150dpi 24-bit color postscript file that was 34 inches by
50 inches in Tailor. Ran out of swap space on *that* one...
> > Given that NeXT seems to be putting more emphasis on their
> > OpenStep for WindowsNT product than the MachOS product, one
> > might wonder how much commitment they have to making future
> > major upgrades to the MachOS portion. I would like to see an
> > OpenStep for FreeBSD product. This would leave the operating
> > system issues to people who want to work on them (and who don't
> > mind giving out the source code), but pick up the user interface
> > that has worked well (for me) in NeXTSTEP. Note that this is
> > different than NeXT upgrading their OS to include the latest
> > BSD kernel and networking.
>
> How about OPENSTEP/Linux? Or perhaps OPENSTEP (or, failing that,
> GNUSTEP) over the HURD kernel? HURD is a decendant of Mach....
My understanding is that Linux is more SysV than BSD, so I was
thinking it'd be easier to move OpenStep for MachOS on top of
FreeBSD. I could be wrong on that, I don't really know enough of
the details. Certainly I'd be happy to see OpenStep on top of
Linux as well. I guess I'm not quite sure of HURD yet, I don't
know if it has a big enough userbase to make it worth the effort
to get OpenStep running on it. I think I'd go with Plan9 or
BeOS before HURD (yeah, as if it's my decision to make...).
Basically I don't mind too much if NeXT wants to get out of the
operating system business, just as long as OpenStep is running on
some operating system which is being supported and somewhat higher
quality than anything Microsoft will throw together as an excuse
for a system. I will not be comfortable if NeXT's fortunes depend
on the good will of Microsoft, and Microsoft can someday decide
that maybe it wants to own whatever market NeXT has success with.
Charles William Swiger
Excerpts from netnews.comp.sys.next.advocacy: 14-May-96 Re: OpenStep,
waining MACH .. by Garance A Drosehn@eclips
>>> deal for the hardware you were getting.
>>
>> certainly agree. However, NeXT charged between 2 and 3 times
>> the normal street price for add-on items like RAM and hard drives.
>
> NeXT, and just every other manufacturer of personal computers and
Regardless of what the other workstation vendors were doing, you can't
claim that paying ~$2000 to upgrade a 8MB slab to a 32MB slab is "a
pretty good deal for the hardware you were getting".
24 MB of RAM isn't worth $2000, ya know.... :-)
[ ... ]
>> How about OPENSTEP/Linux? Or perhaps OPENSTEP (or, failing that,
>> GNUSTEP) over the HURD kernel? HURD is a decendant of Mach....
>
> My understanding is that Linux is more SysV than BSD, so I was
> thinking it'd be easier to move OpenStep for MachOS on top of
> FreeBSD. I could be wrong on that, I don't really know enough of
> the details.
Whether an OS is BSD or SysV shouldn't really matter to OPENSTEP. It
would have mattered to NEXTSTEP, but NeXT has obviously made things more
portable: they've got OPENSTEP for NT, after all.
Linux is a SysV/BSD hybrid like a lot of other operating systems,
including NEXTSTEP (in parts). Linux uses GNU tools all over the place,
and you can often configure the GNU tools to behave either way...
Andrew Lindesay
> > How about OPENSTEP/Linux? Or perhaps OPENSTEP (or, failing that,
> > GNUSTEP) over the HURD kernel? HURD is a decendant of Mach....
>
> "the user interface that has worked well [...] in NeXTSTEP" but rather makes
GNUSTEP running on X should be OK for starters and give the sort of
drivers and stuff that xf86 provides for various PC cards; then you can
of course run bowman? They are also writing a stand-alone DGS window
server are they not? I too hope they keep it simple and mimic the
obvious window system!
Actually the best solution would be for NeXT to continue to produce and
develop NeXTSTEP :-) That would be ideal as far as I'm concerned and I
do hope it will happen.
Andrew (a...@kcbbs.gen.nz)
Alan A. Barhorst
I too priced the black beauties against Macs or PC of the time. There was
no comparison, especially with third party hard drives, and Sun and
Apollo/HP were not in a grad students budget or were as completely
capable. The price/performance curves of the time all showed what I
experienced. As today, NeXT pricing may seem high but when you find what
you get, you quickly realize that you saved money. Of course I speak from
an academic seat.
I'll tell you what....If I tell two friends about NS....and they tell two
friends.....and so on.....maybe we won't all be driving MS products :-).
It really is funny the variety we as consumers demand for cars, shampoo,
etc, and as computer users we all want one design (and my design did not
win)....go figure. Or is it that we really don't have the diversity in
other products and the computer industry is following a natural
progression. I sure like the difference I see in the cars I shop for. I
sure hope to have the same option in the future of software purchases.
In article <4n992u$p...@news.xmission.com> d...@misckit.com writes:
>
> I basically agree with Chuck's points, but there's one I'd
> like to comment on a little:
>
> Charles William Swiger <cs...@andrew.cmu.edu> wrote:
> > NeXT's base hardware models were good value for the money, I certainly
> > agree. However, NeXT charged between 2 and 3 times the normal street
> > price for add-on items like RAM and hard drives.
>
> While this is true, it must also be remembered that NeXT was only
> pricing the disk and RAM the same way HP and Sun were doing at the
> time. I was at BYU and saw the HP price lists, and their prices
> were typically inflated 33% or more above what NeXT's rates were!
> (And those were edu prices...) In the workstation world, it was
> more common to buy disk and RAM from third parties. That's what I
> did then, and I still shop around now. It pays off. I guess the
> whole point of the pricing model was that NeXT wanted to sell
> machines and not have to bother with disk and RAM. So they'd
> price things that way to encourage you to buy the commodity
> products from someone who specialized in them.
>
> So, anyone who had any sense about them would get the most stripped
> down box they could and then beef it up themselves. Worked great
> for me and I don't regret a penny of what I spent. (I have three
> NeXT boxes of my own, FWIW.)
>
> > I've had NS 3.3p1 be completely rock-stable on black hardware and some
> > Intel hardware (if you've got well-supported Intel hardware, anyway).
> > 30 day uptimes are pretty common, with the machines being voluntarily
> > rebooted more often than they crash.
>
> Yup. My web and file server simply doesn't crash. I reboot it
> occasionally on purpose, but other than that it is rock solid. The
> only forced reboots have to do with power outages (Utah power is
> really flakey). In fact, my ISP, running Solaris on Sun machines,
> tends to crash hard about weekly as of late. In that light,
> NEXTSTEP seems to be doing well for me... :-)
>
> --
> Later,
>
> -Don Yacktman
> d...@misckit.com
> <a href="http://www.misckit.com/don.html">My home page</a>
--
AB
______________________________________________________________
Alan A. Barhorst | al...@osci.me.ttu.edu
Mechanical Engineering | http://www.osci.ttu.edu/
Texas Tech University | NeXT, MIME and ASCII mail.
______________________________________________________________
Uli Zappe
a...@kcbbs.gen.nz (Andrew Lindesay) wrote:
> > Well, the problem with GNUSTEP so far would be that it DOES NOT implement
> > "the user interface that has worked well [...] in NeXTSTEP" but rather
>
Which starters? Programmers? May be. Users? I think not, definitely, because
for them it's the GUI that matters most.
> then you can of course run bowman?
Bowman may do as good a job as is possible under X11; unfortunately, that's
far from being good enough :-(
> They are also writing a stand-alone DGS window server are they not?
So far, I don't think so.
> I too hope they keep it simple and mimic the obvious window system!
Feaar of copyright violation seems to prevent this :-(((
Bye
Eric A. Dubiel
Andrew Lindesay wrote:
> GNUSTEP running on X should be OK for starters and give the sort of
> drivers and stuff that xf86 provides for various PC cards; then you can
> of course run bowman? They are also writing a stand-alone DGS window
> server are they not? I too hope they keep it simple and mimic the
> obvious window system!
Isn't this also a great idea, because currently you can't control the
NEXTSTEP UI via X?! This way you could have the NeXT UI on MAC OS, Lynx,
Win 3.1...any OS that you can get an X client for...
I've been itching for this functionality for some time now...and there's
DPS for X!
------------------------------------------
Eric A. Dubiel; http://www.ilstu.edu/~eadubie
mailto:ead...@rs6000.cmp.ilstu.edu ASCII, MIME or NeXT Mail
Instructional Technology Services- Illinois State University
Understanding is best learned via experience
ALL VIEWS EXPRESSED REPRESENT MYSELF ONLY
Eric A. Dubiel
Uli Zappe wrote:
> NeXT offering a "full NEXTSTEP w/o the Mach kernel" to run on FreeBSD (or
> Linux, for that matter) might be a good alternative, but who knows if they
> make up their minds?
Or how about on the new PowerPC Macintosh Linux on MACH?!
This is what I want- NEXTSTEP/OPENSTEP- whatever and MAC OS on the same
box...with PPCP you can add a bunch more too...
see http://mklinux.apple.com
John Kheit
u...@tallowcross.uni-frankfurt.de (Uli Zappe) wrote:
> a...@kcbbs.gen.nz (Andrew Lindesay) wrote:
> Feaar of copyright violation seems to prevent this :-(((
Barring blatent copying of actual artwork (pixel for pixel copies of
recyclers and screen/icon elements) there is no copyright issue to look and
feel. The supreme court was quite clear in the Lotus case, look and feel
is NOT protectable, no matter what NeXT or any other company may want you
to believe...
--
Thanks, be well, take care, later, John Kheit )^> %^) =^)
monoChrome, Inc. | New York Law School
NEXTSTEP Developer | Opinions expressed represent me only
MIME, SUN, & NeXTmail OK | http://cnj.digex.net/~jkheit
mailto:jkh...@cnj.digex.net | Telepathy...It's coming...
Uli Zappe
John Kheit <jkh...@cnj.digex.net> wrote:
> Barring blatent copying of actual artwork (pixel for pixel copies of
> recyclers and screen/icon elements) there is no copyright issue to look and
> feel.
May well be, but who will be able apart from blatant copying to create icons
of the quality Keith Olfs did?
John Kheit
u...@tallowcross.uni-frankfurt.de (Uli Zappe) wrote:
> John Kheit <jkh...@cnj.digex.net> wrote:
> > Barring blatent copying of actual artwork (pixel for pixel copies of
> > recyclers and screen/icon elements) there is no copyright issue to look
and
> > feel.
> May well be, but who will be able apart from blatant copying to create
icons of the quality Keith Olfs did?
There are many artists out there... Moreover, the overall look-and-feel
can be achieved even if the icon/elements are not exact copies.
But I tend to agree, Keith has done some incredible work w/ NS...
The Shrike
In article <4nnq9m$g...@news3.digex.net>, jkh...@cnj.digex.net wrote:
] u...@tallowcross.uni-frankfurt.de (Uli Zappe) wrote:
] > a...@kcbbs.gen.nz (Andrew Lindesay) wrote:
] > Feaar of copyright violation seems to prevent this :-(((
]
] Barring blatent copying of actual artwork (pixel for pixel copies of
] recyclers and screen/icon elements) there is no copyright issue to look and
] feel. The supreme court was quite clear in the Lotus case, look and feel
] is NOT protectable, no matter what NeXT or any other company may want you
] to believe...
unfortunately, the supreme court decision was tied 4 to 4 (one justice had
to recuse himself because of potential conflict of interest), which means
that the lower court decision is upheld, but another company could raise
the issue again and get all the way to the supreme court with it, where we
could see something as bad as a 5-4 decision for look-and-feel protection,
or even sentences that span 6 lines.
s.
John Kheit
I think the above is wrong.
LEVEL 1 - 1 OF 10 CASES
LOTUS DEVELOPMENT CORP. v. BORLAND INTL., INC.
94-2003
SUPREME COURT OF THE UNITED STATES
116 S. Ct. 1062; 1996 U.S. LEXIS 1581; 134 L. Ed. 2d 206; 64
U.S.L.W. 3592
March 4, 1996, Decided
JUDGES: [*1] Rehnquist, Stevens, O'Connor, Scalia, Kennedy, Souter,
Thomas, Ginsburg, Breyer.
OPINION: The petition for rehearing is denied. Justice Stevens took no part
in the consideration or decision of this petition.
As you can see, rehearing is denied...which means that the previous cases
were properly held in the opinion of the Supremes...
The previous Borland case still stands...And so do the Apple v. MS cases.
Look and feel is NOT protectable.
Andrew Lindesay
In message <<4noeeg$2...@tallowcross.uni-frankfurt.de>> u...@tallowcross.uni-frankfurt.de writes:
> > Barring blatent copying of actual artwork (pixel for pixel copies of
> > recyclers and screen/icon elements) there is no copyright issue to look and
> > feel.
>
> of the quality Keith Olfs did?
Working on it :-)
http://www.iconz.co.nz/~iapl/gnustep_apps
Hopefully moving to the GNUSTEP page soon.... :-) Keith Olfs is BLOODY
GOOD at what he does, but it's not impossible to reach those hights :-)
I am going to a lot of bother to get these done so that GNUSTEP icons
are not ugly.
Andrew (a...@kcbbs.gen.nz)
Charles F. Waltrip
Andrew Lindesay wrote:
>
> In message <<4noeeg$2...@tallowcross.uni-frankfurt.de>> u...@tallowcross.uni-frankfurt.de writes:
> >
> > May well be, but who will be able apart from blatant copying to
> > create icons of the quality Keith Olfs did?
>
> Working on it :-)
>
> http://www.iconz.co.nz/~iapl/gnustep_apps
Very nice.
>
> Hopefully moving to the GNUSTEP page soon.... :-) Keith Olfs is
> BLOODY GOOD at what he does, but it's not impossible to reach those
> hights :-)
>
For me it is 8^} That doesn't stop me from appreciating the work of
others and I am especially pleased to see the appreciation for Keith
Ohlfs whose work has made the NEXTSTEP environment so aesthetically
pleasurable to work in.
> I am going to a lot of bother to get these done so that GNUSTEP icons
> are not ugly.
>
Thank you.
> Andrew (a...@kcbbs.gen.nz)
Chuck Waltrip
Opinions expressed are my own.
E-mail: wal...@zephyr.jhuapl.edu
NeXTmail OK.
The Shrike
In article <4o05a5$j...@news4.digex.net>, jkh...@cnj.digex.net wrote:
] shr...@batnet.com (The Shrike) wrote:
] > In article <4nnq9m$g...@news3.digex.net>, jkh...@cnj.digex.net wrote:
] > unfortunately, the supreme court decision was tied 4 to 4 (one justice
] had to recuse himself because of potential conflict of interest), which
] means that the lower court decision is upheld, but another company could
] raise the issue again and get all the way to the supreme court with it,
] where we could see something as bad as a 5-4 decision for look-and-feel
] protection, or even sentences that span 6 lines.
]
] I think the above is wrong.
]
] LEVEL 1 - 1 OF 10
CASES
] LOTUS DEVELOPMENT CORP. v. BORLAND INTL., INC.
] 94-2003
] SUPREME COURT OF THE UNITED STATES
] 116 S. Ct. 1062; 1996 U.S. LEXIS 1581; 134 L. Ed. 2d 206; 64
] U.S.L.W.
3592
] March 4, 1996, Decided
] JUDGES: [*1] Rehnquist, Stevens, O'Connor, Scalia, Kennedy, Souter,
] Thomas, Ginsburg, Breyer.
] OPINION: The petition for rehearing is denied. Justice Stevens took no part
] in the consideration or decision of this petition.
]
] As you can see, rehearing is denied...which means that the previous cases
] were properly held in the opinion of the Supremes...
as you can see, justice stevens took no part in the decision, and it was a
tie. everything i read (which i of course believe) said that this
particular suit was settled, but a similar case involving different
parties could reach the supreme court. i hope you're right, but i don't
think so.
s.
John Kheit
shr...@batnet.com (The Shrike) wrote:
> ] LEVEL 1 - 1 OF 10 CASES
> ] LOTUS DEVELOPMENT CORP. v. BORLAND INTL., INC. ]
94-2003 ] SUPREME COURT OF THE UNITED STATES ]
116 S. Ct. 1062; 1996 U.S. LEXIS 1581; 134 L. Ed. 2d 206; 64 ]
U.S.L.W.
> 3592 ]
March 4, 1996, Decided ] JUDGES: [*1] Rehnquist, Stevens, O'Connor,
Scalia, Kennedy, Souter, ] Thomas, Ginsburg, Breyer.
> ] OPINION: The petition for rehearing is denied. Justice Stevens took no
part ] in the consideration or decision of this petition. ]
> ] As you can see, rehearing is denied...which means that the previous
cases ] were properly held in the opinion of the Supremes...
> as you can see, justice stevens took no part in the decision, and it was
a tie. everything i read (which i of course believe) said that this
particular suit was settled, but a similar case involving different parties
could reach the supreme court. i hope you're right, but i don't think so.
What I see is the big DENIED. When something is denied by the supremes,
the net effect is upholding the court below.
Look and feel is not protected.
William Grosso
In <4og2fm$1...@news3.digex.net> John Kheit <jkh...@cnj.digex.net>
writes:
>What I see is the big DENIED. When something is denied by
>the supremes, the net effect is upholding the court below.
>
I think what may be confusing here is the "4-4" tie.
As I understand it (liable to be wrong; I ain't no
Judge Wapner), getting to the Supreme Court has 2
phases:
1) Will they even listen to you.
2) Making your case.
The "denied" happened on step 1. 4 Judges said "naaah.
Ain't nowhere near close enough to bother with." 4 judges
said "Well...maybe it's worth listening to."
Even if, in some future similar case, the Supremes were to
decide to listen to arguments (unlikely, having set a
precedent), the fact that 4 justices thought it wasn't
worth bothering with in this case means protecting look
and feel is an uphill struggle.
Cheers,
Andy
The Shrike
In article <4og2fm$1...@news3.digex.net>, jkh...@cnj.digex.net wrote:
] shr...@batnet.com (The Shrike) wrote:
] > ] LEVEL 1 - 1 OF 10
CASES
] > ] LOTUS DEVELOPMENT CORP. v. BORLAND INTL., INC.
]
] 94-2003 ] SUPREME COURT OF THE UNITED STATES
]
] 116 S. Ct. 1062; 1996 U.S. LEXIS 1581; 134 L. Ed. 2d 206; 64
]
] U.S.L.W.
] > 3592 ]
] March 4, 1996, Decided ] JUDGES: [*1] Rehnquist, Stevens, O'Connor,
] Scalia, Kennedy, Souter, ] Thomas, Ginsburg, Breyer.
] > ] OPINION: The petition for rehearing is denied. Justice Stevens took no
] part ] in the consideration or decision of this petition. ]
] > ] As you can see, rehearing is denied...which means that the previous
] cases ] were properly held in the opinion of the Supremes...
]
] > as you can see, justice stevens took no part in the decision, and it was
] a tie. everything i read (which i of course believe) said that this
] particular suit was settled, but a similar case involving different parties
] could reach the supreme court. i hope you're right, but i don't think so.
]
] What I see is the big DENIED. When something is denied by the supremes,
] the net effect is upholding the court below.
]
] Look and feel is not protected.
ok, you leave me no choice - you don't know what you're talking about.
s.
John Kheit
> OK, shrike. I'm wrong. I don't know what I'm talking about. So, I want
you to say what is the correct holding. If when I say "Look and feel is
NOT protected," I am wrong... Then I want you to say what is right. I
want to see it in type. Tell me what is right.
Just one other thought... If look and feel is copyrightable, I guess every
GUI, every app that ever used a menu structure would be beholding to
Lotus... Spreadsheets and forms of all types would be in essence patented
There would be no need for checking into 'derivative' works on copyright,
b/c anything that remotely seemed to look like another earlier copyrighted
work would be in violation. The supremes would in essence reverse the
constitution's general principle to provide copyrights so that more
information is produced so that people may build on other people's
innovations (I.E. stand on the shoulders of giants). In short, the entire
history of all limits on copyright would be reversed.
Wow, how the heck did I ever pass that copyright exam?
John Kheit
shr...@batnet.com (The Shrike) wrote:
> In article <4og2fm$1...@news3.digex.net>, jkh...@cnj.digex.net wrote:
> ] What I see is the big DENIED. When something is denied by the
supremes, ] the net effect is upholding the court below. ]
> ] Look and feel is not protected.
> ok, you leave me no choice - you don't know what you're talking about.
OK, shrike. I'm wrong. I don't know what I'm talking about. So, I want
you to say what is the correct holding. If when I say "Look and feel is
NOT protected," I am wrong... Then I want you to say what is right. I
want to see it in type. Tell me what is right.
--
David Evans
In article <4oshhh$l...@news4.digex.net>,
John Kheit <jkh...@cnj.digex.net> wrote:
>shr...@batnet.com (The Shrike) wrote:
>> In article <4og2fm$1...@news3.digex.net>, jkh...@cnj.digex.net wrote:
>> ] What I see is the big DENIED. When something is denied by the
>supremes, ] the net effect is upholding the court below. ]
>> ] Look and feel is not protected.
>
>> ok, you leave me no choice - you don't know what you're talking about.
>
>OK, shrike. I'm wrong. I don't know what I'm talking about. So, I want
>you to say what is the correct holding. If when I say "Look and feel is
>NOT protected," I am wrong... Then I want you to say what is right. I
>want to see it in type. Tell me what is right.
>
Calm down, John. I've started taking der Shrike much less seriously lately.
Perhaps he has returned from the infirmery--who knows. Think of him as comic
relief.
--
David Evans (NeXTMail OK) dfe...@bbcr.uwaterloo.ca
Computer/Synth Junkie http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo "Default is the value selected by the composer
Ontario, Canada overridden by your command." - Roland TR-707 Manual
William Grosso
In <DsEop...@novice.uwaterloo.ca> dfe...@bbcr.uwaterloo.ca (David
Evans) writes:
>
> Calm down, John. I've started taking der Shrike much less
>seriously lately. Perhaps he has returned from the infirmery--
>who knows. Think of him as comic relief.
>
This implies that at some point you were taking him
(shrike is a him ? I hadn't realized that) seriously ?
I'm scared.
Cheers,
Andy
David Evans
In article <4p04v4$m...@dfw-ixnews7.ix.netcom.com>,
Hmmm...you're right perhaps it's not a him.
But, remember a while ago he posted these vitriolic scathings of NS? Sort-of
like Crispin without the eloquence. Now he just seems to spout one-liners.
d...@misckit.com
dfe...@bbcr.uwaterloo.ca (David Evans) wrote:
> [...].
> But, remember a while ago he posted these vitriolic scathings of NS?
> Sort-of like Crispin without the eloquence. Now he just seems to spout
> one-liners.
Actually, without Crispin's content as well. Crispin has some valid
arguements, even if he was a bit acerbic in tone. This guy reminds
me more of Sean Eric Fagan (remember SEF? gee, that was a long time
ago, wasn't it!). I don't think anyone missed SEF.
The funniest SEF incident was when Jayson put a hardcoded image of
the devil into NewsGrazer that appeared whenever one of SEF's postings
was displayed. :-)
Gregory H. Anderson
d...@misckit.com writes
> dfe...@bbcr.uwaterloo.ca (David Evans) wrote:
> > [...].
> > But, remember a while ago he posted these vitriolic scathings of NS?
> > Sort-of like Crispin without the eloquence. Now he just seems to
> > spout one-liners.
As Nigel Tufnel (of Spinal Tap) used to say, "There's such a fine line
between clever and .... stupid."
> Actually, without Crispin's content as well. Crispin has some valid
> arguements, even if he was a bit acerbic in tone. This guy reminds
> me more of Sean Eric Fagan (remember SEF? gee, that was a long time
> ago, wasn't it!). I don't think anyone missed SEF.
The SEF wars were just before my time, but I had the special version
of NG. I was absolutely ROTFL the first time Mr. Devil Head popped up.
ObShrike: Perhaps this is a nothing more than a horrible spelling
accident? Transpositino si a fairyl commno errro. Yuo haev ot admti,
teh naem makse moer senes teh othre wya. Tyr ti yoursefl.
--
Gregory H. Anderson | "I wander'd off by myself, In the
Crystal Ball/Star Gazer | mystical moist night-air, and from
Anderson Financial Systems | time to time, Look'd up in perfect
gr...@afs.com (NeXTmail OK) | silence at the stars." Walt Whitman
William Grosso
In <4p22gr$f...@shelob.afs.com> Greg_A...@afs.com (Gregory H.
Anderson) writes:
>
>teh naem makse moer senes teh othre wya. Tyr ti yoursefl.
>
Did anybody else see this and wonder "Why's Greg typing
with a Scottish accent ?"
Andy
(at the end of a looong day)
David Evans
In article <4p1v9f$a...@news.xmission.com>, <d...@misckit.com> wrote:
>dfe...@bbcr.uwaterloo.ca (David Evans) wrote:
>> [...].
>> But, remember a while ago he posted these vitriolic scathings of NS?
>> Sort-of like Crispin without the eloquence. Now he just seems to spout
>> one-liners.
>
>arguements, even if he was a bit acerbic in tone. This guy reminds
>me more of Sean Eric Fagan (remember SEF? gee, that was a long time
>ago, wasn't it!). I don't think anyone missed SEF.
>
I'm afraid SEF was long before my time, although I still enjoyed his
"mall" encounter.
Mark's arguments were at least educational on occasion.
>The funniest SEF incident was when Jayson put a hardcoded image of
>the devil into NewsGrazer that appeared whenever one of SEF's postings
>was displayed. :-)
>
Ahhh....what times I've missed. :-)
Gregory H. Anderson
William Grosso writes
> In <4p22gr$f...@shelob.afs.com> Greg_A...@afs.com (Gregory H.
> Anderson) writes:
> >
> >teh naem makse moer senes teh othre wya. Tyr ti yoursefl.
> >
> Did anybody else see this and wonder "Why's Greg typing
> with a Scottish accent ?"
Aye t'ye, Rrrob Rrroy MacGrrregorrr. If it's not Scottish, it's crrrap.
(apologies to SNL)
Jonathan W. Hendry
William Grosso (apul...@ix.netcom.com) wrote:
: In <4p22gr$f...@shelob.afs.com> Greg_A...@afs.com (Gregory H.
: Anderson) writes:
: >
: >teh naem makse moer senes teh othre wya. Tyr ti yoursefl.
: >
: Did anybody else see this and wonder "Why's Greg typing
: with a Scottish accent ?"
Looks a bit Welsh, too. Perhaps he's been spending too much
time in Bryn Mawr, PA?
---
Jonathan W. Hendry Views expressed herein do
Steel Driving Software, Inc. not represent those of
stee...@ix.netcom.com Steel Driving Software, Inc.
j...@exnext.com or Lexis-Nexis
mmalcolm crawford
On 06/05/96, Jonathan W. Hendry wrote:
> William Grosso (apul...@ix.netcom.com) wrote:
> : In <4p22gr$f...@shelob.afs.com> Greg_A...@afs.com (Gregory H.
> : Anderson) writes:
> : >
> : >teh naem makse moer senes teh othre wya. Tyr ti yoursefl.
> :
> : Did anybody else see this and wonder "Why's Greg typing
> : with a Scottish accent ?"
>
> Looks a bit Welsh, too. Perhaps he's been spending too much
> time in Bryn Mawr, PA?
>
memory serves me correct, though it's eight years since I was there, so I may
have it wrong!)? Geez, I can't see why that'd make you talk Welsh!
How about Bryn Mawr, Chicago? Does that have a sizeable Celtic population?
If Greg's typing has anything to do with anything this side of the pond, it
looks to me more like he's been eating British beef... ;-)
Best wishes,
clamm.
--
EADEANS
Isn't this the month (finally!) for OpenStep v. 4 for Mach?
Thanks,
Ed Deans
TeamOS2 & ClubOpenDOC
Ian Stephenson
In article <4p4dcf$c...@mailgate.lexis-nexis.com> uhe...@meaddata.com
> William Grosso (apul...@ix.netcom.com) wrote:
> : In <4p22gr$f...@shelob.afs.com> Greg_A...@afs.com (Gregory H.
> : Anderson) writes:
> : >
> : >teh naem makse moer senes teh othre wya. Tyr ti yoursefl.
> : >
> time in Bryn Mawr, PA?
Can't possible be welsh - it's impossible to write taht much welsh
without a double l:
"llyrachwyn!"
Now that looks welsh!
$an
Gregory H. Anderson
Ian Stephenson writes
> Can't possible be welsh - it's impossible to write taht much welsh
> without a double l:
See, now Ian's doing it too. It's catching!
One of my favorite things about "The Englishman (Who Went Up a Hill
But Came Down a Mountain)" was the name of the peak in question.
I couldn't possibly reproduce it from memory, but it had about 6 l's.