RUMOUR ALERT: Novell dumps UnixWare on SCO
Dan Busarow
floor at Unix Expo In New York)
Well, now they've gone and done it.
According to reasonably reliable rumours and rumblings around here, it looks
like Novell will be handing UnixWare, both the product and the people who
make it, to the folks in black and teal. SCO, long criticized for not having
a Unix 4.2 platform, has been handed the 4.2 code on a silver platter, in
return for a 20% piece of SCO.
In a three-way deal, SCO gets UnixWare, the Florham Park devbelopment team,
licencing deals for IPX/SPX and a player to be named later. HP gets the ability
to license Unix to other companies, the charge for developing the planned
64-bit Unix extension, and basically anything to do with non-Intel Unix.
More details to come later, as we're sure you all will hear by the time
we return to California and Toronto.
Have a nice day.
Dan & Evan
Bob Stewart
: (This article is being co-written by Dan and Evan Leibovitch, from the show
: floor at Unix Expo In New York)
: Well, now they've gone and done it.
: According to reasonably reliable rumours and rumblings around here, it looks
: like Novell will be handing UnixWare, both the product and the people who
: make it, to the folks in black and teal. SCO, long criticized for not having
: a Unix 4.2 platform, has been handed the 4.2 code on a silver platter, in
: return for a 20% piece of SCO.
If true, this will go a long way towards creating a united front against
NT, etc. If not true, you'll roast in net-hell :=)
Later.
--
Ecological adversity ensures survival of the fittest...
Pleasurable reproductive processes ensure biodiversity...
Neat!
Bob Stewart (KB9ZW) | Who's gonna pay the bills
wk USA (310) 335-7152 | when everyone's on welfare?
Dave Parker
>(This article is being co-written by Dan and Evan Leibovitch, from the show
>floor at Unix Expo In New York)
>
>Well, now they've gone and done it.
>
>According to reasonably reliable rumours and rumblings around here, it looks
>like Novell will be handing UnixWare, both the product and the people who
>make it, to the folks in black and teal. SCO, long criticized for not having
>a Unix 4.2 platform, has been handed the 4.2 code on a silver platter, in
>return for a 20% piece of SCO.
>
>licencing deals for IPX/SPX and a player to be named later. HP gets the ability
>to license Unix to other companies, the charge for developing the planned
>64-bit Unix extension, and basically anything to do with non-Intel Unix.
>
>More details to come later, as we're sure you all will hear by the time
>we return to California and Toronto.
>
>Have a nice day.
>
>Dan & Evan
Nice. Real nice. Is old debbil Billygate behind this? This would tempt me
more than anything yet to give in to the 'dark side', make a deal with...
etc, etc... How they handle those folks like me who have shelled out major
bucks in the last couple of months to pass the required UnixWare reseller
certification and start out on the UnixWare CNE track will determine whether
I do indeed sell my soul to Microsloth. My dependents and I have to eat,
and those little red and white (but VERY expensive) study guides don't taste
very good.
Could this also explain why, after having been an authorized UnixWare
reseller since fall/winter '92, and having been in their data base at
least as recently as last April or May, I should suddenly and mysteriously
disappear from their data base after I shell out bucks, pass their exams,
and fax in the exam results, per their instructions???
I'm not angry or anything like that, but I'd better stop now before I have
to kill again...
--
Dave Parker dlpa...@dlpinc00.com
alternate address: dpa...@mail.coin.missouri.edu
W. Kent Rutan
Happy Browsing,
Kent
--
========================================================================
ru...@cat.com Caterpillar Inc.
Tel (309) 578-2278 Kent Rutan Technical Center
Fax (309) 578-2354 Peoria, Illinois USA
========================================================================
The opinions expressed herein are mine, not Cat's [etc., ad infinitum].
Steve Ward
Merge in the next UnixWare relase (codenamed Eiger).
Sounds like the 'roadmap' is going to turn some of us into 'roadkill' if we
can't afford SCO's exhorbitant pricing structure.
Steve
--
Steve Ward, Jr., Advanced Systems Specialist
School of Engineering, University of Portland
Portland OR
st...@up.edu
Craig Nordin
Doesn't this mean that if you are on SCO right now, you'll have
one helluva move up to SVR4.2, no matter what SCO does to make
it more friendly?
--
W. Kent Rutan
> While reading, I noticed that part of SCO's 'roadmap' includes unbundling
> Merge in the next UnixWare relase (codenamed Eiger).
But notice they mention Windows support is included. WABI?
>
> Sounds like the 'roadmap' is going to turn some of us into 'roadkill' if we
> can't afford SCO's exhorbitant pricing structure.
UnixWare definitely has the most bang for the buck (especially with
the developer's bundle).
The other iteresting thing I noticed was that the "Merged UNIX" will
use X.Desktop. In the interests of Unix unification, I think they
should ship CDE as the default, with X.desktop as an option.
Martin Sohnius
: (This article is being co-written by Dan and Evan Leibovitch, from the show
: floor at Unix Expo In New York)
: Well, now they've gone and done it.
: According to reasonably reliable rumours and rumblings around here, it looks
: like Novell will be handing UnixWare, both the product and the people who
: make it, to the folks in black and teal. SCO, long criticized for not having
: a Unix 4.2 platform, has been handed the 4.2 code on a silver platter, in
: return for a 20% piece of SCO.
: In a three-way deal, SCO gets UnixWare, the Florham Park devbelopment team,
: licencing deals for IPX/SPX and a player to be named later. HP gets the ability
: to license Unix to other companies, the charge for developing the planned
: 64-bit Unix extension, and basically anything to do with non-Intel Unix.
: More details to come later, as we're sure you all will hear by the time
: we return to California and Toronto.
: Have a nice day.
Well, what can I say. You were surprisingly accurate in the details,
considering that you posted about five hours before even the broadcast to
employees (which I tried to listen to on a very crackly cellphone from a
hotel room in France at about midnight local time).
Everybody expected the HP deal, nobody expected the SCO one. It sounds
very much as if SCO have entered into some sort of obligation in fact to
build upon the UnixWare 'Eiger' release they are acquiring (the 2.1
currently ramping up to beta stage), rather than dump it under the heading
of "one more competitor out of the way". If everything that is said in the
various press releases and announcements is correct, we have a situation
where at least three "flavours" of UNIX will merge: SCO UNIX,
SVR4.2/UnixWare, and HP-UX. Notice that there is definite commitment to
have the same APIs for the (non-Intel) HP-UX and for the SCO product.
On the whole, and leaving aside such nasty little questions as "do my friends
still have jobs", or "what's gonna happen to the small businesses who trusted
Novell", I would say that not having deadly competition between the Santa
Cruz Operation and Novell can only be A Good Thing. Assuming that the name
HP does not run the shivers down the spines of all other hardware
manufacturers (as "Sun" once did), the same could be said for the third leg
of this tripod.
I believe the one bit that is inaccurate in the above is that "SCO get the
Florham Park development team". Whatever the corporate types may have done,
they DID have control over the expenditure that went into that site.
I may not make friends by saying this, but it's perhaps the UnixWare
marketing folks in San Jose that we should ship across the Sierra towards
the ocean...
--
*******************************************************************
Martin F. Sohnius msoh...@novell.co.uk
Novell IS & T, Bracknell, England +44-1344-724031
*******************************************************************
* if (status = UNDER_NUCLEAR_ATTACK) *
* launch_full_counterstrike(); *
*******************************************************************
(C) 1995 M.F.Sohnius -- free distribution on USENET
(Not a spokesperson -- just a cyclist!)
Mark Tovey
>(This article is being co-written by Dan and Evan Leibovitch, from the show
>floor at Unix Expo In New York)
>
>Well, now they've gone and done it.
>
>According to reasonably reliable rumours and rumblings around here, it looks
>like Novell will be handing UnixWare, both the product and the people who
>make it, to the folks in black and teal. SCO, long criticized for not having
>a Unix 4.2 platform, has been handed the 4.2 code on a silver platter, in
>return for a 20% piece of SCO.
>
>In a three-way deal, SCO gets UnixWare, the Florham Park devbelopment team,
>licencing deals for IPX/SPX and a player to be named later. HP gets the ability
>to license Unix to other companies, the charge for developing the planned
>64-bit Unix extension, and basically anything to do with non-Intel Unix.
>
>More details to come later, as we're sure you all will hear by the time
>we return to California and Toronto.
>
>Have a nice day.
>
>
Oh great! I spent the last two years trying to convince everyone
around here that UnixWare was the way to go, that SCO is junk and we
should move away from it. Now that I have finally succeded and
started porting all of our applications to UnixWare, I find myself
back in bed with SCO again! Not my favorite bedfellow!
Mark Tovey
--
Mark Tovey |
Sr. Project Engr. | In the land of the dark, the Ship of the Sun is driven
HK Systems | by the Grateful Dead.
t...@netcom.com | -- Egyptian Book of the Dead
801-530-4287 |
Ludovic Valois
>around here that UnixWare was the way to go, that SCO is junk and we
>should move away from it. Now that I have finally succeded and
>started porting all of our applications to UnixWare, I find myself
>back in bed with SCO again! Not my favorite bedfellow!
Same here, ole buddy. We're an old SCO account, and one of the top 5 in
France. I just spent months going through a huge RFP to renew the entire app
server scheme, inc. hardware, OSes and vendors. The ink's not even dry on the
UnixWare deal, with training and ports about to debut, and here come the SCO
people again with huge grins on their faces...
The one consolation: almost no-one, in Europe at least, knew anything about
this deal prior to Wednesday. I was at the Novell headquarters in Paris just
that morning, chitchatting away when the French CEO busted in with the draft
press release that came out later in the US. The guy from USG in London almost
fell off his chair, _and_ he didn't know all the details yet (obviously Novell
tried to put a happy face on it, but it's bloody treason all the same,
whatever the beneficial long-term consequences to Unix itself...). I talked to
the SCO-France people today, they were deep in the dark as well...
Hey, remember the old Chinese saying ? 'May you live in interesting times ' ?
Oh, sorry: it's not a saying. It's a curse :-)
W. Kent Rutan
Sohnius) writes:
[...]
> Everybody expected the HP deal, nobody expected the SCO one. It sounds
> very much as if SCO have entered into some sort of obligation in fact to
> build upon the UnixWare 'Eiger' release they are acquiring (the 2.1
> currently ramping up to beta stage), rather than dump it under the heading
> of "one more competitor out of the way".
I doubt if there is any tangible obligation on SCO's part. Generally
the acquiring company thinks their way is best and that's what rules.
> If everything that is said in the
> various press releases and announcements is correct, we have a situation
> where at least three "flavours" of UNIX will merge: SCO UNIX,
> SVR4.2/UnixWare, and HP-UX. Notice that there is definite commitment to
> have the same APIs for the (non-Intel) HP-UX and for the SCO
> product.
It is nice to see the Unix industry consolidating a little more but it
sounds like a real rats nest of backward compatibility. But I guess
Spec1170 is too. And so is SVR4 for that matter. So the final merged
product will really be Xenix + SVR3 + 4.2 BSD + SunOS + SCO UNIX +
HP-UX. Am I leaving anything out? After having gone through all that
merging I don't see how they could miss being Spec1170 compliant. Or
even Spec2340.
For the graphical stuff, it looks like they could care less about
merging. Looks like everything is OpenServer stuff from their
statement. It would seem to me that CDE would be a better choice for
a merged Unix, or at least include pieces of the UW desktop. But
their choices probably have something to do with my first statement
above.
[...]
> I believe the one bit that is inaccurate in the above is that "SCO get the
> Florham Park development team". Whatever the corporate types may have done,
> they DID have control over the expenditure that went into that site.
Martin, I don't quite understand what you're saying here. Can you
elaborate?
Martin Sohnius
: more than anything yet to give in to the 'dark side', make a deal with...
: etc, etc... How they handle those folks like me who have shelled out major
: bucks in the last couple of months to pass the required UnixWare reseller
: certification and start out on the UnixWare CNE track will determine whether
: I do indeed sell my soul to Microsloth. My dependents and I have to eat,
: and those little red and white (but VERY expensive) study guides don't taste
: very good.
I am beginning to get the impression that SCO actually mean it when they
say that they are initially going to ship UnixWare and OpenServer
simultaneously, with the first "unified" product going into beta next
summer. Now, if that is so, you may well find it worthwhile to talk to
them NOW. They have, at present, very little skills in UnixWare, I'd say.
(They probably do have an archive of this newsgroup, somewhere, I'd think.)
:-)
Steve Ward
>While reading, I noticed that part of SCO's 'roadmap' includes unbundling
>Merge in the next UnixWare relase (codenamed Eiger).
After receiving further details, I must amend the above statement.
Sources within both SCO and Novell are telling me that this unbundling was
already planned by Novell for UW2.1. Apperently there are some sort of
issues with Locus that resulted in the decision.
I can't say I'm thrilled about it, but I can't blame it on SCO, it was
already in the works.
As for how it'll be priced, I guess we'll just have to wait and see...
Greg Smith
: Oh great! I spent the last two years trying to convince everyone
: around here that UnixWare was the way to go, that SCO is junk and we
: should move away from it. Now that I have finally succeded and
: started porting all of our applications to UnixWare, I find myself
: back in bed with SCO again! Not my favorite bedfellow!
I've been using this as a way to proclaim that I was right all along.
Something like "see, SCO is junk; even they know it, that's why they had
to buy UnixWare--so they could have a good product to sell."
* Greg Smith gsm...@westnet.com Baltimore, MD
* Computer consultant, speaker builder, and endless warehouse of
* classic rock music information
Evan Leibovitch
W. Kent Rutan <ru...@cat.com> wrote:
>It's not a rumor.
At the time it was first posted, it sure was.
Dan and I put out the "rumour alert" late Tuesday; the official
announcement did not break until *exactly* 10:30am EST on Wednesday.
Security was *very* tight, and there were very solid NDAs agreed to
by anyone involved. As soon as the statement was made, there was
plenty of PR info available. By the time Kent posted his reply,
getting the official statements was easy.
One detail that I did not flesh out until the next day, and I don't
believe have been detailed in any of the official reports, was the
exact nature of the ownership swap:
Novell gets roughly 18% of SCO, compared to about 13% currently held
by Microsoft. As far as I can tell, no cash was involved; only SCO
stock. Novell also gets a spot on the SCO board.
Computing Canada magazine is actually paying me to write something, so
they get first dibs on my report on the happening of this week. After
that, I'll post one of my usual long-winded analyses. Consider yourselves
warned.
Here's the executive summary of my spin (though much depends on SCO's
handling of its new property):
Big winners: SCO, SCO users, Caldera, ISVs, IHVs
Slight winners: UnixWare users, HP
Slight losers: Sunsoft, Microsoft, IBM
Big losers: Novell
Unsure: UnixWare OEMs, VARs & resellers
Details to follow.
--
Evan Leibovitch, Sound Software Ltd., located in beautiful Brampton, Ontario
SCO & Novell Unix Master Reseller / ev...@telly.org / (905) 452-0504
There are two types of people: those who divide people into two types,
Dave Parker
>Dave Parker (dlpa...@dlpinc00.com) wrote:
>
>: more than anything yet to give in to the 'dark side', make a deal with...
>: etc, etc... How they handle those folks like me who have shelled out major
>: bucks in the last couple of months to pass the required UnixWare reseller
>: certification and start out on the UnixWare CNE track will determine whether
>: I do indeed sell my soul to Microsloth. My dependents and I have to eat,
>: and those little red and white (but VERY expensive) study guides don't taste
>: very good.
>
>I am beginning to get the impression that SCO actually mean it when they
>say that they are initially going to ship UnixWare and OpenServer
>simultaneously, with the first "unified" product going into beta next
>summer. Now, if that is so, you may well find it worthwhile to talk to
>them NOW. They have, at present, very little skills in UnixWare, I'd say.
>
>(They probably do have an archive of this newsgroup, somewhere, I'd think.)
"Would you like some fries with your crow, Mr. Parker?"
"Yum! Tastes GREAT!"
I flew off the handle. It was a bad day. Things are looking better now.
After having put off taking the UW2 Installation and Configuration exam
and finally having done that last Saturday, faxing in my test results to
reseller certification, getting the CNE agreement to sign and mail back
to "ECNE/CNE Administration" (they need to update their forms), and then,
on the same day, getting a callback saying that since I'm not a UnixWare
reseller (been one since late '92) and hearing from some pretty knowledgeable
folks (thanks, Dan and Evan) that Novell was indeed selling UnixWare, I
think people can maybe understand how I got a little rattled.
The mixup at reseller certification was a very minor one, very easily
resolved. And the woman I spoke with at Novell Ed. this morning said that
for the foreseeable future nothing would change with the UnixWare CNE track.
That is consistent with what I read on the SCO web site. Thanks for the tip.
I'll continue on the UW CNE track and give SCO a call.
And apologies to all for any injudicious remarks I might have made about any
of this.
Jim Vlcek
> The other iteresting thing I noticed was that the "Merged UNIX" will
> use X.Desktop. In the interests of Unix unification, I think they
> should ship CDE as the default, with X.desktop as an option.
CDE is not an example of UNIX unification.
SCO's acquisition of UnixWare, on the other hand, is.
--
Jim Vlcek I don't understand...
vl...@byteware.com ...it's all UNIX to me.
The Black Box of Lowertown
Beautiful downtown St. Paul
Anthony W. Southworth
: In article <DF9G4...@novell.co.uk> msoh...@novell.co.uk (Martin
: Sohnius) writes:
: [...]
: > Everybody expected the HP deal, nobody expected the SCO one. It sounds
: > very much as if SCO have entered into some sort of obligation in fact to
: > build upon the UnixWare 'Eiger' release they are acquiring (the 2.1
: > currently ramping up to beta stage), rather than dump it under the heading
: > of "one more competitor out of the way".
: I doubt if there is any tangible obligation on SCO's part. Generally
: the acquiring company thinks their way is best and that's what rules.
I agree that there's probably not must strict obligation there, but...
SCO has taken much criticism for not have an SVR4 kernel, and being
built on "old" technology. Also, all the recent benchmarks have shown
the SVR4.2 MP kernel to scale much better on multiple processors, and
generally outperform SCO's latest release. I just don't see how basing
the combined OS on the older SVR3.2 kernel would be a good business
decision. Then again, I'm not SCO... B-)
: [snip]
: It is nice to see the Unix industry consolidating a little more but it
: sounds like a real rats nest of backward compatibility. But I guess
: Spec1170 is too. And so is SVR4 for that matter. So the final merged
: product will really be Xenix + SVR3 + 4.2 BSD + SunOS + SCO UNIX +
: HP-UX. Am I leaving anything out? After having gone through all that
: merging I don't see how they could miss being Spec1170 compliant. Or
: even Spec2340.
It's all the "backward compatibility" in the SCO press releases that
bothers me. (I'm sure none of us will forget the discussion in this
group about Novell dropping the older device driver interface in UW 2.)
While backwards compatibility sounds good on the surface, it always
seems to contribute to a larger, slower kernel. Someone has to draw the
line somewhere. Who knows, after SCO is done with the UnixWare kernel,
UnixWare might be on par speed wise with OpenServer. B-( If all the
compatibility does cause a decrease in speed, and SCO ups the price to
OpenServer's level, you can kiss all those UnixWare price/performance
records goodbye...
: For the graphical stuff, it looks like they could care less about
: merging. Looks like everything is OpenServer stuff from their
: statement. It would seem to me that CDE would be a better choice for
: a merged Unix, or at least include pieces of the UW desktop. But
: their choices probably have something to do with my first statement
: above.
SCO has said before that it didn't need CDE; SCO is "Windows Friendly",
remember? After all, no one needs a Unix desktop, just access the
Unix server from your MS-Windows desktop... (sorry, couldn't resist B-)
I do hope SCO considers keeping at least some of the UnixWare graphical
admin tools. As for the desktop itself, it's okay, but I can take it
or leave it.
: ========================================================================
: ru...@cat.com Caterpillar Inc.
: Tel (309) 578-2278 Kent Rutan Technical Center
: Fax (309) 578-2354 Peoria, Illinois USA
: ========================================================================
: The opinions expressed herein are mine, not Cat's [etc., ad infinitum].
--
Anthony W. Southworth Email: ant...@carsinfo.com
CARS Information Systems Corp. WWW: http://www.carsinfo.com/~anthony
4000 Executive Park Drive
Cincinnati, OH 45241
W. Kent Rutan
> W. Kent Rutan writes
> > The other iteresting thing I noticed was that the "Merged UNIX" will
> > use X.Desktop. In the interests of Unix unification, I think they
> > should ship CDE as the default, with X.desktop as an option.
>
> CDE is not an example of UNIX unification.
>
> SCO's acquisition of UnixWare, on the other hand, is.
So you think that we should have just one Unix vendor? Sure, let's
just forget this whole competition thing and let one vendor dictate
what we get. I, for one, don't agree with the Microsoft mentality.
--
Clarence Dold
: So you think that we should have just one Unix vendor? Sure, let's
What a small world we live in.
To my knowledge, the largest corporate user of Unix is KMart Corporation.
There are over 6000 Unix computers installed nationwide.
They are dropping SVR4, supplied by Unisys, some of which was destined to be
passthrough Unixware, in favor of Windows NT.
I don't know how important 6000 copies of UnixWare is, but with support
contracts, it would have been substantial.
--
---
Clarence A Dold - do...@rahul.net
- Pope Valley & Napa CA.
Martin Sohnius
: > I believe the one bit that is inaccurate in the above is that "SCO get the
: > Florham Park development team". Whatever the corporate types may have done,
: > they DID have control over the expenditure that went into that site.
: Martin, I don't quite understand what you're saying here. Can you
: elaborate?
It is becoming clearer now that the site at FPK will remain as a Novell site,
and will still house most of the people currently there. My comment arose
out of the very recent, and very expensive, move from Summit to the new
facility. Somebody in the know would probably have pulled the plug on it,
if they had intended to move out of NJ lock stock and barrel.
Jim Pickering
SCO ACQUIRES UNIX BUSINESS FROM NOVELL
AND LICENSES NETWARE TECHNOLOGY
Novell Takes 17% Equity Position in SCO
NEW YORK, NY (September 19, 1995) -- The Santa Cruz Operation, Inc.
(NASDAQ:SCOC) and Novell, Inc. (NASDAQ: NOVL) today announced a
definitive agreement for SCO to purchase the UNIX business from Novell.
Under the agreement, Novell will receive approximately 6.1 million shares
of SCO common stock, resulting in an ownership position of approximately
17% (post transaction) of the outstanding SCO capital stock. SCO will also
license Novell's NetWare Directory Services and other NetWare 4
technologies as the basis for future networking services. SCO plans to
merge the SCO OpenServer Release 5 and UnixWare 2 product lines to
create a standard high-volume UNIX operating system that contains
integrated NetWare networking services, and expects to release this
merged product in 1997.
Alok Mohan, president and CEO of SCO, said, "This extends SCO's
leadership position in the Business Critical Server market. Our customers
and resellers not only get a powerful UNIX operating system, but also the
most advanced network services in the world. Novell's advanced network
services, such as NetWare Directory Services, are setting the standard for
business networking. Our customers will be able to integrate their Business
Critical Servers with their existing workgroups to provide their people with
greater access to corporate data."
"SCO's Business Critical Server focus and worldwide distribution channel
makes them an ideal partner for taking UNIX application servers forward on
the Intel platform," said Robert J. Frankenberg, chairman and CEO of
Novell. By focusing on our areas of expertise, and working to integrate our
technologies, Novell and SCO together will meet the application server
needs of customers in a networked world."
According to the terms of the agreement, SCO will acquire Novell's
UnixWare business and UNIX intellectual property. In order to meet
customer support needs and protect development requirements, SCO
intends to hire a number of Novell employees. In addition to the SCO stock,
Novell will receive a revenue stream back from SCO based on revenue
performance of the purchased UNIX business. This revenue stream will
end in the year 2002 and is not to exceed $84 million net present value.
SCO and Novell expect the agreement to close on or about December 1,
1995, conditional upon several items including U.S. government approval
under the Hart-Scott-Rodino Act. Shareholder approval is not required by
either party.
SCO has outlined a product roadmap in which both SCO OpenServer and
UnixWare will initially continue to be individually supported and enhanced.
Upcoming releases include enhanced SCO OpenServer and UnixWare
products, containing integrated NetWare services, in the first half of 1996.
In the summer of 1996, SCO expects to release a beta version of the
product that merges SCO OpenServer and UnixWare, with a complete
software migration toolkit. This beta kit will enable developers to begin
developing their applications to a single merged product line that contains
the best capabilities of both environments. SCO expects to release the final
version of this merged product, containing integrated NetWare services, in
1997. The merged product will offer binary compatibility with existing SCO
OpenServer and UnixWare applications, as well as a full set of migration
tools to ensure that developers can easily develop for the new line from
either predecessor.
SCO will continue to offer UnixWare source code to existing and new OEM
licensees worldwide. Existing Certified Novell Engineers (CNEs) for
UnixWare are fully certified to support UnixWare products from SCO.
In order to smooth the transition for existing customers of SCO OpenServer
and UnixWare to the future merged product, SCO plans to offer a number
of enhancements to the product lines over the coming months. These
include UNIX 95 compliance; NetWare file, print, and directory services;
and tools that allow software developers to quickly create a single binary
that will run on SCO OpenServer Release 5, UnixWare 2, and the next-
generation merged operating system.
SCO is the world's leader for UNIX System servers and multiuser hosts.
SCO Business Critical Servers run the critical, day-to-day operations of
large branch organizations in retail, finance, and government, as well as
corporate departments and small to medium-sized businesses of every kind.
SCO is also a leading supplier of client-integration software that integrates
Windows PCs and other clients with UNIX servers from all of the major
vendors. SCO sells and supports its products through a worldwide network
of distributors, resellers, systems integrators, and OEMs.
The business of Novell, Inc. is connecting people with other people and the
information they need, enabling them to act on it anytime, anywhere. Novell
is the world's leading network software provider. The company's software
products provide the distributed infrastructure, network services, advanced
network access and network applications required to make networked
information and computing an integral part of everyone's daily life.
# # #
SCO, The Santa Cruz Operation, the SCO logo, and SCO OpenServer are registered
trademarks of The Santa Cruz Operation, Inc. in the USA and other countries.
NetWare and UnixWare are registered trademarks of Novell, Inc.. UNIX is a
registered trademark in the US and other countries, licensed exclusively
through X/Open Company Limited. All other brand or product names are or may be
trademarks of, and are used to identify products or services of, their
respective owners.
Arthur Emerson III
> Doesn't this mean that if you are on SCO right now, you'll have
> one helluva move up to SVR4.2, no matter what SCO does to make
> it more friendly?
> --
An even better question would be: Did our Unixware skills just become
much more valuable??????
This SCO deal makes me think about the unpublished Microsoft credo - If
you can't write your own, slander someone who can until they depreciate
enough to buy them........
-Arthur
Evan Leibovitch
>> Everybody expected the HP deal, nobody expected the SCO one. It sounds
>> very much as if SCO have entered into some sort of obligation in fact to
>> build upon the UnixWare 'Eiger' release they are acquiring (the 2.1
>> currently ramping up to beta stage), rather than dump it under the heading
>> of "one more competitor out of the way".
>I doubt if there is any tangible obligation on SCO's part. Generally
>the acquiring company thinks their way is best and that's what rules.
Yes, but there's little doubt that SCO was losing some ground to
UnixWare on new purchases, and that they'd wanted to get into SVR4
without paying blackmail-type licensing.
It should be noted that the evening before the announcement (Tuesday),
the UnixWare Technology Group had held a small champagne celebration
at its booth to honour UnixWare's standings in the latest AIM Technology
report (Fall 1995).
This report not only bestowed nine out of a total of 16 "Hot Iron"
awards to systems running UnixWare, but showed the results of AIM's
apples-to-apples head-to-head comparisons with SCO OpenServer in
which UW won every category. Depending on the test and the load,
on identical hardware UnixWare was between 1% and 5% better on file
server tests and between 8% and 10% better on the "Multiuser/Shared
System Mix" tests.
SCO can hardly just toss UnixWare.
Rampant speculation from those I spoke to (based on little behind it
than wishful thinking and what passed for common sense at the time) is
that the merged release planned by SCO for 1997 will have mainly
UnixWare innards, but will be called OpenServer for matters of pride
and stability. And make no bones about it, there are elements of SCO's
approach -- especially regarding IHVs -- that UnixWare can definitely
benefit from.
Still. there are many question markes. And believe me, I saw the
lobbying start the minute after the announcement was made.
>> If everything that is said in the
>> various press releases and announcements is correct, we have a situation
>> where at least three "flavours" of UNIX will merge: SCO UNIX,
>> SVR4.2/UnixWare, and HP-UX.
That's *generally* true; there are some features of HP/UX that will
creep into the SCO/UW product. But as HP makes their transition away
from PA/RISC and onto the next generation of Intel processors, it is
indeed safe to say that HP's OS will be the same as SCO's short of
any specific HP value-added.
>So the final merged
>product will really be Xenix + SVR3 + 4.2 BSD + SunOS + SCO UNIX +
>HP-UX.
Well, look at the backwards compatibility we have already...
SVR3.2 = System V + SCO Xenix
SVR4 = SVR3.2 + BSD/SunOS
SVR5 = SVR4 + small bits of HP/UX + SCO 3.2.4
^ (just guessing...)
>> I believe the one bit that is inaccurate in the above is that "SCO get the
>> Florham Park development team". Whatever the corporate types may have done,
>> they DID have control over the expenditure that went into that site.
>Martin, I don't quite understand what you're saying here. Can you
>elaborate?
From what I can gather, as the Keeper of the Holy Unix Source, SCO is
going to need most of the Bell/USL/USG engineers. They will finally
have the ability to do things in-house that until now had to sub-license.
As Martin suggested, it's likely that while most technical UnixWare people
will find work at HP and SCO, most of the Novell UW non-technical folk
probably won't.
Evan Leibovitch
Martin Sohnius <msoh...@novell.co.uk> wrote:
>Everybody expected the HP deal, nobody expected the SCO one.
I remember byself saying to myself at the time, "nobody expects the
Santa Cruz Operation", in the same tone/voice as in the Monty Python
skit, "Nobody expects the Spanish Inquisition".
Try it.
>Assuming that the name
>HP does not run the shivers down the spines of all other hardware
>manufacturers (as "Sun" once did),
This is the real grey area. OEMs are not going to be comfortable taking
their lead from HP; I received *distinctly* different signals from the
official HP position and an interview I conducted later with an AT&T-GIS
exec.
One of the keys to making this whole deal work will be SCO's willingness
to work with a supporting consortium such as UTG. If OEMs are unwilling
to "partner" with a direct competitor, and SCO isn't interested in them,
they may look elsewhere for their platforms...
Justin M. Collins
>Bye bye,
>Simple product line
>Great pricing
>Hello,
>Option this, and option that for more $$$
>Odd SCOisms (just when I was getting used to the UnixWareIsms)
all of that depends if you actually decide to stay with this product
line. We have already started looking into alternate options just
in case SCO really messes this up.
>Oh, well. Life goes on.
Just not business.
What I am MOST concerned about is that until SCO takes over that no
ptf's will be released for UW 2.0.x. And given the state of the mail
system that is just not an option.
-justin m. collins
---------------------------------------------------------------------------
Senior Systems/Network Administrator email:jcol...@firestorm.servtech.com
ServiceTech, Inc. http://www.servtech.com v:(716)263-3360 f:(716)423-1596
---------------------------------------------------------------------------
Darren R. Davis
|> In article <DF681...@news.beach.net> d...@dpcsys.com (Dan Busarow) writes:
|> >floor at Unix Expo In New York)
|> >
|> >Well, now they've gone and done it.
|> >
|> >According to reasonably reliable rumours and rumblings around here, it looks
|> >like Novell will be handing UnixWare, both the product and the people who
|> >make it, to the folks in black and teal. SCO, long criticized for not having
|> >a Unix 4.2 platform, has been handed the 4.2 code on a silver platter, in
|> >return for a 20% piece of SCO.
|> >
|> >In a three-way deal, SCO gets UnixWare, the Florham Park devbelopment team,
|> >licencing deals for IPX/SPX and a player to be named later. HP gets the ability
|> >to license Unix to other companies, the charge for developing the planned
|> >64-bit Unix extension, and basically anything to do with non-Intel Unix.
|> >
|> >More details to come later, as we're sure you all will hear by the time
|> >we return to California and Toronto.
|> >
|> >Have a nice day.
|> >
|> >Dan & Evan
|>
|> etc, etc... How they handle those folks like me who have shelled out major
|> bucks in the last couple of months to pass the required UnixWare reseller
|> certification and start out on the UnixWare CNE track will determine whether
|> I do indeed sell my soul to Microsloth. My dependents and I have to eat,
|> and those little red and white (but VERY expensive) study guides don't taste
|> very good.
|>
|> reseller since fall/winter '92, and having been in their data base at
|> least as recently as last April or May, I should suddenly and mysteriously
|> disappear from their data base after I shell out bucks, pass their exams,
|> and fax in the exam results, per their instructions???
|>
|> I'm not angry or anything like that, but I'd better stop now before I have
|> to kill again...
As you may have seen from my other posting, it is true that Novell
has sold SCO the UnixWare Business. The details of the deal are on:
http://www.sco.com
http://www.novell.com
--
Darren R. Davis
Senior Software Engineer
Novell Developer Support
Olav Tollefsen
ru...@cat.com (W. Kent Rutan) wrote:
> So you think that we should have just one Unix vendor? Sure, let's
> just forget this whole competition thing and let one vendor dictate
> what we get. I, for one, don't agree with the Microsoft mentality.
In think Microsoft Windows NT has caused a lot of healthy competition in the
UNIX world. The UNIX vendors have been able to sell their overpriced
systems because of hardware and software incompatibilities. You still have to
pay a lot more for a Digital Alpha box if you want UNIX compared with
Windows NT. Why is Oracle becoming more price competitive? Could it
be caused by MS SQL Server? I think so.
--
- Olav Tollefsen
- Microsoft Norway
-
- Nothing I have written here is in any way, shape, or form official
- Microsoft policy or opinion, and should not be represented as such
- when quoted here or in any other forum.
Jim Vlcek
> In article <DFAG9...@byteware.com> vl...@byteware.com (Jim Vlcek) writes:
> > CDE is not an example of UNIX unification.
> >
> > SCO's acquisition of UnixWare, on the other hand, is.
>
In a word, yes.
> Sure, let's
> just forget this whole competition thing and let one vendor dictate
> what we get. I, for one, don't agree with the Microsoft mentality.
We have Microsoft. We can have a single UNIX vendor. We can't have multiple
UNIX vendors, all peddling similar-but-maddeningly-different-in-small-ways
flavors of UNIX. We have a decade and a half of experience to prove this
point.
Justin M. Collins
>: all of that depends if you actually decide to stay with this product
>: line. We have already started looking into alternate options just
>: in case SCO really messes this up.
>Justin, what options are you guys looking into? Come to think of it,
>we wound up with UnixWare because there really weren't any Credible
>options. Everything else was too expensive, or too kludgy, or just too...
We are looking into just about everything. ie keeping our minds open
to different ways of doing thing. The one good thing is that
we will probably stay with 2.0.x (and 2.1 (maybe)) for a couple of
years to come, so in 3 years or so, who knows what will be available.
But what I have seen so far of SCO has not impressed me, and while
the pricing does not concern me so much, a poor product does.
>I really hope something good comes out of this. Somehow I don't think
>I'll be able to afford it, though :-(
That can be a problem (pricing).
>So, it comes down to Solaris X86 or Linux? I'm sorry, but SCO is still
>just not an option, for many of the same reasons that I won't buy from
>MicroSoft.
The only good thing is that Unixware already exists and I hope SCO
can take a decent product and make it better. If they were starting
off from scratch I would give it little or no chance of success.
Justin M. Collins
>In article <43uju4$f...@ns.oar.net>,
>Anthony W. Southworth <ant...@carsinfo.com> wrote:
>>
>>I agree that there's probably not must strict obligation there, but...
>>SCO has taken much criticism for not have an SVR4 kernel, and being
>>built on "old" technology. Also, all the recent benchmarks have shown
>>the SVR4.2 MP kernel to scale much better on multiple processors, and
>>generally outperform SCO's latest release. I just don't see how basing
>>the combined OS on the older SVR3.2 kernel would be a good business
>>decision. Then again, I'm not SCO... B-)
>This is one part that scares the heck out of me! I remember many
>times in the past SCO saying they had no need for SysVr4.x technology,
>as there 3.2 technology had been refined with all the needed business
>enhancements. I can see everything converting over to the SCO way of
>thinking, and if so I am out of this game..
This is also the part that scares the crap out me as well. The only
reason I can see for SCO saying we do not need SVR4.x technology
is because they did not want to pay the licensing fees. Now that
they do not have to, let us hope and pray that they use what is
at their disposal.
>Overall this is all part of what I worry about, I can see all the SCOisms
>being forced upon us. I have worked with many different UNIXes, and the
>main reason I avoid SCO at all costs is due to the fact the take most of
>the know interfaces/configuration methods and change them around. This
>just makes it this much harder to configure/debug there OS, so the easiest
>solution is to just not use it...
Agreed as well. It would be nice if SCO listened to its forthcoming
customer base, otherwise I have a feeling they may end up losing a
significant portion of this base to competition.
Evan Leibovitch
>> CDE is not an example of UNIX unification.
>> SCO's acquisition of UnixWare, on the other hand, is.
>So you think that we should have just one Unix vendor? Sure, let's
>just forget this whole competition thing and let one vendor dictate
>what we get. I, for one, don't agree with the Microsoft mentality.
Over the years it has become clear to me that there is more than enough
competition in this area. On the one hand you have the RISC vendors,
Sun and DEC and SGI and IBM, who have succeeded in one-Unix-on-this-CPU
environments (except for DEC, porting Linux in the hopes that someone
will actually notice the Alpha chip). Nobody seems to consider the
fact that you can only get AIX on RS/6000s to be anti-competitive.
On the other hand, there's NT, which is a far greater threat to Unix on
Intel systems than it is to the RISC and large-system vendors.
Me, I'm tired of Unix wars. Microsoft is at least partially right when
depicting Unix as the Bosnia of the computer world. Having seen the
comings and goings of the Dells and UHCs and Bell Technologies and
Consensys and Esix and two different Microports, I've had enough.
All this diversity has gotten us nowhere. While it could be argued that
UnixWare has put significant competitive pressures on SCO to improve its
product, I believe the onslaught of NT is having the same effect.
Then there's the rear assault of commercial Linux. If the choice of
commercial Unix on Intel evolves to either SCO or Caldera, that's enough
for me.
Having been the eternal optimist regarding Novell's commitment to Unix,
I am nonetheless glad that the company is getting out of a product which
was clearly out of its league. While I too have my reservations about SCO,
having three Unix variants (SCO, UW, HP/UX) evolve into one is something I
am ready for.
The big question mark IMO is: how much will SCO learn from the things
UnixWare did right?
--
Evan Leibovitch, Sound Software Ltd., located in beautiful Brampton, Ontario
SCO & Novell Unix Master Reseller / ev...@telly.org / (905) 452-0504
There are two types of people: those who divide people into two types,
and those who don't.
Evan Leibovitch
>SCO has taken much criticism for not have an SVR4 kernel, and being
>built on "old" technology.
At the announcement, this issue came up again, and Doug Michaels gave
what must by now be a canned answer:
"Yes, SCO is based on SVR3.2. So is SVR4."
SCO has indeed extended SVR3.2, indeed in many ways as much as SVR4 has
done. While SCO has much to learn from SVR4's implementations of various
things, it's *always* been an unfair knock on SCO to claim its stuff
is ancient.
Compare what the products do, how robust they are, how fast they are,
and how flexible they are. That's what counts -- harping on pedigree
to the exclusion of practical concerns is a waste of breath.
>SCO has said before that it didn't need CDE; SCO is "Windows Friendly",
>remember? After all, no one needs a Unix desktop, just access the
>Unix server from your MS-Windows desktop... (sorry, couldn't resist B-)
Having tired of X-Windows myself, I can't say that I'm bothered by this
anymore. IMO the "Windows-Friendly" strategy was a good one.
>I do hope SCO considers keeping at least some of the UnixWare graphical
>admin tools. As for the desktop itself, it's okay, but I can take it
>or leave it.
And I'm looking forward to having UnixWare on-line documentation in HTML
and system administration done through TCL rather than FMLI.
There were some things -- a lot of things -- that SCO did right. On-line
docs were a case where Novell's problem of backwards compatibility, in
this case with existing NetWare documentation, was a greater hinderence
to Novell than to SCO.
One can only hope that SCO will take the best of UnixWare and the best
of SCO and combine them in a single package that incorporates the best
of both. If this happens, combined with the fact that ISVs and IHVs will
only have to make a single Intel Unix distribution, will be the best
this that could haooen right now to this market.
Howard D. Leadmon
>the SVR4.2 MP kernel to scale much better on multiple processors, and
>generally outperform SCO's latest release. I just don't see how basing
>the combined OS on the older SVR3.2 kernel would be a good business
>decision. Then again, I'm not SCO... B-)
This is one part that scares the heck out of me! I remember many
times in the past SCO saying they had no need for SysVr4.x technology,
as there 3.2 technology had been refined with all the needed business
enhancements. I can see everything converting over to the SCO way of
thinking, and if so I am out of this game..
>It's all the "backward compatibility" in the SCO press releases that
>: For the graphical stuff, it looks like they could care less about
>: merging. Looks like everything is OpenServer stuff from their
>: statement. It would seem to me that CDE would be a better choice for
>: a merged Unix, or at least include pieces of the UW desktop. But
>: their choices probably have something to do with my first statement
>: above.
>
>SCO has said before that it didn't need CDE; SCO is "Windows Friendly",
>remember? After all, no one needs a Unix desktop, just access the
>Unix server from your MS-Windows desktop... (sorry, couldn't resist B-)
>
>I do hope SCO considers keeping at least some of the UnixWare graphical
>admin tools. As for the desktop itself, it's okay, but I can take it
>or leave it.
Overall this is all part of what I worry about, I can see all the SCOisms
being forced upon us. I have worked with many different UNIXes, and the
main reason I avoid SCO at all costs is due to the fact the take most of
the know interfaces/configuration methods and change them around. This
just makes it this much harder to configure/debug there OS, so the easiest
solution is to just not use it...
-------------------------------------------------------------------------------
Internet : how...@wb3ffv.ampr.org | Howard D. Leadmon
UUCP : wb3ffv!howardl | ABSnet Internet Services
PHONE : (410)-361-8160 | 200 E. Lexington Street
FAX : (410)-381-8162 | Suite-1602
PACKET : WB3FFV @ WB3FFV.MD.USA.NA | Baltimore, MD 21202
WWW : http://www.abs.net |
GOPHER : gopher.abs.net |
Evan Leibovitch
>I remember many
>times in the past SCO saying they had no need for SysVr4.x technology,
>as there 3.2 technology had been refined with all the needed business
>enhancements.
What were they going to say? Were they going to admit publicly that
they were inferior?
SCO would have been hit with fairly steep licensing fees to go to SVR4,
and they chose instead to do their own enhancements. You could say the
above argument was an attempt to save face.
>I can see everything converting over to the SCO way of
>thinking, and if so I am out of this game..
Having seen the latest AIM benchmarks where UW2 clearly beat out OS5
in every performance category, it's highly unlikely that SCO will ignore
this acquisition.
My prediction is that the "merged" product will be an SVR4.2MP-based
product (with a ton of enhancements) under the name "OpenServer".
Joaquin Menchaca
> Bye bye,
>
> Simple product line
> Great pricing
>
> Hello,
>
> Option this, and option that for more $$$
> Odd SCOisms (just when I was getting used to the UnixWareIsms)
>
Exactly.
SCO = $$$$$$
SCO = SCOisms, not UNIXisms.
I hope SCO just dumps their current line of crapola, and sticks with
UnixWare and standards, and stop adding their SCOism features.
I think Sun said it before, "Just Say No to SCO". I may just do that.
Last comment: Novell, you stink. I'll never recomend investing into
Novell ever again, with exception of WordPerfect devision.
Joaquin Menchaca
> just not an option, for many of the same reasons that I won't buy from
> MicroSoft.
"Say no to SCO", Sun
Unfortunately, I'm recommending WindowsNT for a server. Novell servers
are too lilmiting, because the OS does nothing except act a as Network
server and maybe more with some expensive NLMs. UnixWare had promise, but
not anymore.
I gave up on SCO as a viable product 5+ years ago. I never heard anything
positive about SCO, only negative. I expected people to move over to NT,
Linux, Solaris, or UnixWare, when they could. $4K servers are a tad to
high, especially with the garbage, SCOisms, you get with it.
Paul Beardsell
> This SCO deal makes me think about the unpublished Microsoft credo - If
> you can't write your own, slander someone who can until they depreciate
> enough to buy them........
>
> -Arthur
I agree.
Some other thoughts:
* The SCO - Unixware competition was a good thing.
* Unixware was neglected, mismarketed and misunderstood by Novell.
And was misunderstood in the industry. Managers and marketing
types were never going to understand that Novell owned the dog's
bollocks version of Unix. And the worst mistake was naming it
"Unixware". It should have been called "SVR4" or at least had that
in its name.
* Unixware had great promise. I was looking forward to support of
enhanced mode MS Windows apps in particular. Then I would have argued
long and hard for the installation of Unixware PE on all our clients'
PCs.
* SCO overprice their products. But Unixware competition was beginning
to tell. E,g. this month's special offer of upgrade to latest SCO
from previous releases for peanuts.
* SCO support sucks but Unixware was worse except here on the net
where it was brilliant.
--
Paul Beardsell SAM Business Systems Ltd
~~~~~~~~~~~~~~ 21 Finn House, Bevenden St,
pbear...@cix.compulink.co.uk London, N1-6BN, UK.
p...@sambusys.demon.co.uk (+44 or 0)171 608-2391
Raymond Shwake
>SCO has said before that it didn't need CDE; SCO is "Windows Friendly",
>remember? After all, no one needs a Unix desktop, just access the
>Unix server from your MS-Windows desktop... (sorry, couldn't resist B-)
>I do hope SCO considers keeping at least some of the UnixWare graphical
>admin tools. As for the desktop itself, it's okay, but I can take it
>or leave it.
SCO essentially sat on the CDE sidelines - initially paying lip-service
to the effort, but not participating in the development work. That was done
by HP, IBM, Novell and Sun. During that effort, SCO effectively said "never
mind", and went its own way. Yes, SCO's "Windows-Friendly" strategy is more
comprehensive and developed than Novell's own musings. I suspect though that
SCO simply doesn't want to pay CDE licensing fees, especially since it's got
an in-house alternative in IXI. If HP, IBM and Sun stick with CDE, SCO will
remain at odds not only with the UNIX majors, but with one of their own
partners (HP).
Dan Busarow
: them, too. Should make for exciting board meetings.
After all the people who had signed non-disclosures (anyone at Novell
who knew _anything_ about this) were freed to talk, we learned that
while MS has a board seat it is filled by a junior accountant who must
leave the room whenever strategy is discussed.
Also, as Kent has pointed out via the URL for SCO's web page, UnixWare
as a product will continue at least until 1997. It stays with Novell
through December 1995 so if you have any deals coming up around then
you may want to push them through to get current pricing.
The licensing issues are what really worry me. Will UW be the low priced
entry model (even though clearly superior) for the Open Server 5 star or
will it be the other way around?
Interesting times.
Dan
--
Dan Busarow
DPC Systems
Dana Point, California
Leslie Mikesell
Evan Leibovitch <ev...@telly.telly.org> wrote:
>The big question mark IMO is: how much will SCO learn from the things
>UnixWare did right?
My prediction is that it will be about the same as what Novell learned
from what Dell had done right. How long did it take for UnixWare
to have a command to build a tape that could be loaded directly
from boot disks? And you still can't modify those boot disk to
load the way you want...
Les Mikesell
l...@mcs.com
Roby Sherman
Justin M. Collins <jcol...@firestorm.servtech.com> wrote:
>
>What I am MOST concerned about is that until SCO takes over that no
>ptf's will be released for UW 2.0.x. And given the state of the mail
>system that is just not an option.
Agreed. WHAT *IS* going to happen to the ptfs STILL NEEDED by the 1.1.x and
2.x base?
I've been waiting for a fix to 1.1's MBUF problem forever and all Novell has
done is beat-around-the-bush. Which leads me to believe that they are NOT
working on any more patches at this time.
Roby
--
--r...@nexus.interealm.com--------------------------------------Aurora, CO--
| * Hopeful Romantic | "Mom, I find it disturbing that you refer to the |
| * Poet | Weekly World News *AS* the paper. The paper |
| * Imagineer | contains facts..." -- So I married an Axe Murderer|
Ed Hall
: . . . .
: It's all the "backward compatibility" in the SCO press releases that
: bothers me. (I'm sure none of us will forget the discussion in this
: group about Novell dropping the older device driver interface in UW 2.)
: While backwards compatibility sounds good on the surface, it always
: seems to contribute to a larger, slower kernel. Someone has to draw the
: line somewhere. Who knows, after SCO is done with the UnixWare kernel,
: UnixWare might be on par speed wise with OpenServer. B-( If all the
: compatibility does cause a decrease in speed, and SCO ups the price to
: OpenServer's level, you can kiss all those UnixWare price/performance
: records goodbye...
This might not be a problem if they figure out how to make all that
backwards compatibility and make it /optional/. That's one of the
less commonly stated advantages of the "give 'em source" argument the
Linux crowd uses--with sourcecode, it's easy to excise unwanted
features entirely rather than leaving stubs everywhere. Perhaps SCO
should spend some time and let those folks who don't want to suffer
for other people's past sins avoid them. It might mean breaking things
up into a few more object files and perhaps gluing a few more things
together with source (part of the system call interface, for example)
or perhaps it means making a smarter ld. But there's got to be a way
for someone who wants to Keep It Simple, Stupid to Keep It Simple,
Stupid. Functionality stopped its linear relationship to kernel
size and complexity long ago...
-Ed
Dave A. Johnson
Time to start reading the Solaris news group.
In further support for solaris - I just read that Sunsoft has internally
ported Solaris to a power PC -based system.
Iain Hardcastle
p...@sambusys.demon.co.uk "Paul Beardsell" writes:
>
> * SCO support sucks but Unixware was worse except here on the net
> where it was brilliant.
I know someone who is in the know on this sort of thing.
he reckons that SCO support in the UK is naff. Only 2 distibutors can sell SCO
here. Frontline and Sphinx. both of whom (apparently) know little of what they
are on about!
you can buy a support contract from SCO for loads of wonga though.
alternatively buy UW from AZLAN and get the best support in the galaxy ;-)
--
Iain Hardcastle
please note: all opinions expressed here are mine and not AZLAN's
Martin Sohnius
: In article <43ualh$d...@firestorm.servtech.com>,
: Justin M. Collins <jcol...@firestorm.servtech.com> wrote:
: >
: >What I am MOST concerned about is that until SCO takes over that no
: >ptf's will be released for UW 2.0.x. And given the state of the mail
: >system that is just not an option.
: Agreed. WHAT *IS* going to happen to the ptfs STILL NEEDED by the 1.1.x and
: 2.x base?
: I've been waiting for a fix to 1.1's MBUF problem forever and all Novell has
: done is beat-around-the-bush. Which leads me to believe that they are NOT
: working on any more patches at this time.
Well, until Tuesday last week, none of the people working on patches had
the faintest idea of what was going on (I know; I was there and talked to
them). I would admit, however, that since then they may well have been
busy mainly talking to each other in the canteen, and re-vamping their
resumes.
However, whatever complaints you may have about the timeliness or otherwise
of bug fixes, so far that would have been unaffected by recent
developments. And the plan is that this activity should go on unperturbed,
and eventually be taken over by SCO, possibly with the same people. (This
is my reading of the announcements, not any inside info.)
Martin Sohnius
: In article <811844...@sambusys.demon.co.uk>
: p...@sambusys.demon.co.uk "Paul Beardsell" writes:
: >
: > * SCO support sucks but Unixware was worse except here on the net
: > where it was brilliant.
: I know someone who is in the know on this sort of thing.
: he reckons that SCO support in the UK is naff. Only 2 distibutors can sell SCO
: here. Frontline and Sphinx. both of whom (apparently) know little of what they
: are on about!
: you can buy a support contract from SCO for loads of wonga though.
: alternatively buy UW from AZLAN and get the best support in the galaxy ;-)
Absolutely, Iain! :-)
Well, from my reading of the small print in the announcements, one of the
UK distributors of SCO-UnixWare would initially be Azlan (Frontline is
another), because the existing distributor/reseller relationships will stay
what they are.
Apart from that, I am surprised to hear your poor opinion of SCO support
here in the UK. Not because I know any better, but because they didn't
pick me up when I was in the market a couple of years ago. Hence, I
naturally thought there must be a flood over here of good ACEs.
<More grins>.
Bryan Althaus
: Time to start reading the Solaris news group.
: In further support for solaris - I just read that Sunsoft has internally
: ported Solaris to a power PC -based system.
Solaris for PowerPC will be offered both by SunSoft and IBM. Last I
heard it had entered beta and all developer tools have also been moved to
the PowerPC.
Solaris 2.5 is to arrive in December of this year and will support the
new UltraSPARC 64bit chip. At the end is What's new in Solaris 2.5,
which comes with CDE and WABI 2.1.
Having moved from UnixWare to Solaris x86, I found since both are SVR4
the move was fairly easy, as were the software ports. Any PD software
out for UnixWare is also out for Solaris. Also, major companies like
Oracle and Sybase release on Solaris first before other OS's, Oracle
and Sun have a joint operation to sell Oracle/Solaris services.
And best of all, Netscape considers Solaris x86 a supported platform!!
As far as Solaris on PPC goes check out thses URL's:
http://www.Sun.COM/cgi-bin/show?sunsoft/catflash/powerpc.html
(3K)
(Introducing Solaris PowerPC Edition by Laura Mishima, Marketing Manager,
Solaris Products Group. One of the many exciting Solaris developments is
the long-awaited port to the PowerPC microprocessor Solaris PowerPC.
http://www.Sun.COM/cgi-bin/show?sunsoft/solaris.new/PPC/arch.html
(5K)
(Solaris and the PowerPC Architecture: A New - Standard for Enterprise
Client/Server Computing. Solaris(TM) is the developer's choice for mission
critical client-server development. It is the world's leading 32-bit
UNIX®)
http://www.Sun.COM/cgi-bin/show?sunflash/.sunflash/november.1994/sunflash/71.11.firmware.html
(4K)
(SMCC's Firmware Adopted by IBM and Apple. Sun Microsystems Computer
Company (SMCC)announced major third-party and IEEE standards support for
its OpenBoot firmware that promise to make the Sun technology an industry
standard.
---
From: bon...@Eng.Sun.COM (Jeff Bonwick)
Newsgroups: comp.unix.solaris
Subject: Re: What's new in Solaris 2.5?
Date: 3 Jun 1995 08:13:14 GMT
Here's a brief summary of the high points:
New Features:
Hardware support:
UltraSPARC platform support
PCI support
Networking:
NFS Version 3
NFS over TCP
Dynamic IP addressing for PPP
Multithreaded automounter
4.4BSD-compatible telnet
Version 8 sendmail
Faster, cleaner, *much* more robust network file locking (lockd/statd)
Security:
ACLs (Access Control Lists)
NIS+ Password Aging
Standards:
POSIX threads (P1003.1c)
Commands and utilities for SPEC-1170 (XCU4.2)
CDE (Common Desktop Environment)
Kodak Color Management System
X/Open Federated Naming
Performance analysis and debugging tools:
Integrated kernel/user-level tracing tools (see prex(1))
Process analysis tools (see proc(1))
Thread library debugging (libthread_db)
Scoping/versioning linker
Administration:
All-new admintool
Cache-only Clients
cachefs statistics (see cachefsstat(1M))
Features that were formerly unbundled or patched that are now standard:
PCMCIA support
SPARCstorage Array support
Kernel Asynchronous I/O
In-kernel telnetd/rlogind to support hundreds of users
(smaller, faster, and better in 2.5)
Performance Improvements:
Pipes: new pipe implementation 5 times faster than 2.4.
Standard I/O: fread(3S) and fwrite(3S) 4 times faster than 2.4.
Timesharing: dramatically improved due to low-level VM rewrite,
in-kernel telnet/rlogin support, per-processor kernel memory
allocation, and breakup of global locks in ufs, tmpfs and VM.
Name service cache: speeds up name service lookups -- whether from
NIS, NIS+, DNS, or just local files like /etc/passwd -- by two
orders of magnitude (even more for large files or slow networks).
Kernel memory comsumption: reduced by roughly a megabyte on
most platforms (sun4m, sun4d, and i86pc).
SPARC hardware multiply and divide support: most recent SPARC CPUs
provide this, but existing code doesn't take advantage of it.
In 2.5 calls to .mul and .div (etc) are transparently turned
into multiply and divide instructions if the CPU supports it.
NFS: roughly 10% faster.
Window system: 10-20% faster due to fast pipes and reduced
kernel memory consumption.
Compatibility: (bcopy is back!)
Source code:
Popular library routines from SunOS 4.x are now first-class
interfaces in libc:
memory:
bcmp, bcopy, bzero, index, rindex
random numbers:
random, srandom, initstate, setstate
process control:
killpg, getpriority, setpriority, ualarm, usleep,
wait3, wait4, getrusage, getwd, setregid, setreuid
regular expresssions:
re_comp, re_exec
standard i/o:
setbuffer, setlinebuf
miscellaneous:
ftime, getdtablesize, gethostod, gethostname,
sethostname, getpagesize, reboot
Shell scripts:
Popular commands from SunOS 4.x are now in /usr/bin:
arch, hostid, hostname, mach, pagesize
Makefiles:
/usr/ccs/bin/ranlib provided as a simple "exit 0" command.
SunOS 4.x applications:
Dynamically linked apps have been supported since 2.0;
support for statically linked apps was added in 2.3;
support for mixed dynamic/static apps is new in 2.5.
Ship dates:
Preliminary Solaris 2.5 Beta CDs were distributed at the
developer's conference in May 1995. Official Beta CDs
(the same thing plus a few last-minute bug fixes) are
in the works now.
The exact date for Solaris 2.5 FCS (the final product)
has not been set. It should be sometime this Fall.
Disclaimer:
This is not an official Sun document. This is my current
best estimate of the feature content and performance of
Solaris 2.5 based on the features we have integrated
thus far and the benchmarks we have run most recently.
Any or all of this is subject to change without notice.
Jeff Bonwick
Solaris Performance
Martin Sohnius
[...]
: It should be noted that the evening before the announcement (Tuesday),
: the UnixWare Technology Group had held a small champagne celebration
: at its booth to honour UnixWare's standings in the latest AIM Technology
: report (Fall 1995).
If any Novell employees were present, this would have been a quite
astonishing display of what they call "stiff upper lip" over here. After
all, the announcement of the sale of UnixWare to SCO was made to employees
via conference call at 5:30 p.m. EST, Tuesday.
: >> I believe the one bit that is inaccurate in the above is that "SCO get the
: >> Florham Park development team". Whatever the corporate types may have done,
: >> they DID have control over the expenditure that went into that site.
: >Martin, I don't quite understand what you're saying here. Can you
: >elaborate?
: From what I can gather, as the Keeper of the Holy Unix Source, SCO is
: going to need most of the Bell/USL/USG engineers. They will finally
: have the ability to do things in-house that until now had to sub-license.
: As Martin suggested, it's likely that while most technical UnixWare people
: will find work at HP and SCO, most of the Novell UW non-technical folk
: probably won't.
I suggested no such thing. From what I heard, FPK is going to remain a
Novell site. This would make sense, since a lot of money was spent on it.
Others may share facilities nearby. (Perhaps, those large empty buildings
within the confines of the Exxon Development Center in Florham Park, NJ,
don't need to worry about the forthcoming Thanksgiving Day after all :-).
Martin Sohnius
: In a previous article, Craig Nordin (cno...@charm.net) said:
: > Doesn't this mean that if you are on SCO right now, you'll have
: > one helluva move up to SVR4.2, no matter what SCO does to make
: > it more friendly?
: > --
: An even better question would be: Did our Unixware skills just become
: much more valuable??????
: This SCO deal makes me think about the unpublished Microsoft credo - If
: you can't write your own, slander someone who can until they depreciate
: enough to buy them........
Thanks, Arthur. You have finally provided me with a suitable posting to
tag on a notorious saying by the illustrious Lyndon B. Johnson: "I'd
rather have them inside the tent pissing out than outside pissing in."
Timothy Hu
>
>Also, as Kent has pointed out via the URL for SCO's web page, UnixWare
>as a product will continue at least until 1997. It stays with Novell
>through December 1995 so if you have any deals coming up around then
>you may want to push them through to get current pricing.
>
>The licensing issues are what really worry me. Will UW be the low priced
>entry model (even though clearly superior) for the Open Server 5 star or
>will it be the other way around?
I wonder what the upgrade policy will be? Are they gonna stick us
$MEGA for the next full version?
Cheers,
Tim
--
Timothy Y Hu (t...@nyx.cs.du.edu) AT&T Bell Labs
Foresight Enterprises, Inc. -- provider of Information Technologies solutions
Cheyenne, Wyoming 82001 (303)692-4169
--
Dave Parker
>In article <DFDx...@news.beach.net>, Dan Busarow <d...@dpcsys.com> wrote:
>>
>>Also, as Kent has pointed out via the URL for SCO's web page, UnixWare
>>as a product will continue at least until 1997. It stays with Novell
>>through December 1995 so if you have any deals coming up around then
>>you may want to push them through to get current pricing.
>>
>>The licensing issues are what really worry me. Will UW be the low priced
>>entry model (even though clearly superior) for the Open Server 5 star or
>>will it be the other way around?
Should we be looking for a very tempting Solaris-on-Intel competitive
upgrade offer right around that time?
Jim Balson
: Time to start reading the Solaris news group.
: In further support for solaris - I just read that Sunsoft has internally
: ported Solaris to a power PC -based system.
After thinking about everything that has unravelled in the last week,
I can't argue. With SCO not supporting CDE, X86 only, I truly think Solaris
is the future Unix that will be on my intel machine. Runs on SPARC, x86, PPC.
Supports CDE. tough to beat.
--
Jim Balson
bal...@world.std.com
Jim Balson
I am anticipating just such an offer, an am ready to jump when it comes.
--
Jim Balson
bal...@world.std.com
David Gurr
|> In article <811844...@sambusys.demon.co.uk>
|> p...@sambusys.demon.co.uk "Paul Beardsell" writes:
|>
|> >
|> > where it was brilliant.
|> I know someone who is in the know on this sort of thing.
Unfortunately not *that* much in the know ... see below:
|> he reckons that SCO support in the UK is naff. Only 2 distibutors can sell SCO
|> here. Frontline and Sphinx. both of whom (apparently) know little of what they
|> are on about!
The two UK distributors for SCO are Toplog and Sphinx/Level V. Frontline now
sell UnixWare rather than SCO.
Toplog and Sphinx/LevelV are well placed to provide support for the majority of
resellers issues. If however, you have specific needs and this support doesn't suit
you (for example if you're a developer) you can get support direct from SCO.
|> you can buy a support contract from SCO for loads of wonga though.
Depending on the level of support contract you need, a contract direct with
SCO can cost anything from "loads of wonga" down to a few hundred pounds.
If you want 24x7 support with a guaranteed response time, you can get it (but
expect to pay for it). If you want ad-hoc support, then you can get it at an
appropriate price.
--------------------------------------------------------------------
David Gurr dav...@sco.COM
Market Development Manager (+44) (0)1923 813548 (phone)
SCO Ltd. (+44) (0)1923 813816 (fax)
Watford, England
Evan Leibovitch
Martin Sohnius <msoh...@novell.co.uk> wrote:
>: It should be noted that the evening before the announcement (Tuesday),
>: the UnixWare Technology Group had held a small champagne celebration
>: at its booth to honour UnixWare's standings in the latest AIM Technology
>: report (Fall 1995).
>If any Novell employees were present, this would have been a quite
>astonishing display of what they call "stiff upper lip" over here. After
>all, the announcement of the sale of UnixWare to SCO was made to employees
>via conference call at 5:30 p.m. EST, Tuesday.
It was quite the eerie feeling at the time, I recall. There'd been a
special edition of the Unigram-X newsletter put out earlier that day
that had most of the basics down (including the surprise of the SCO
involvement), but its loopy commentary on the deal was a bit much for
Dan and me to stomach. To the extent we could, we tried to corroborate
whether the news was even possibly true; when the info seemed real enough,
we "went live" from the show floor with the original posting that started
this thread, clearly titled "rumour alert" since we *knew* the news we
had would be incomplete or partially wrong.
Making the scene even more nervous was a reporter from Information Week,
hounding Dan and I much of the day for reaction to an event that hadn't
yet happened. No doubt he'd caught us before anything had sunk in.
Thoughtful comment? Not a chance.
I'll consider myself lucky if my quotes haven't been twisted to made me
look stupid. If we're *really* lucky, he's left our names totally out of
the final article.
By the time the corks were popping at UTG, there was not a red Novell
golf shirt in sight; all had been sequestered away to hear The Call.
We all knew something big was up; some people wondered whether there
would even *be* a UnixWare the next day. Unigram editor Maureen O'Gara
turned up to let us know that her extraction of this hush-hush
stuff had been one of the most nighmarish assignments of her life.
In the best of times, I suffer small-talk very poorly; this night, the
atmosphere of the chit-chat was downright asphixiating, kind of a
dance-floor-of-the-Titanic mood that even the wine would not mask.
Everyone else seemed to be having a wonderful time; me, I'm not much
of an actor.
Dan and I did the only possible sane thing that evening; we caught Carol
Burnett in a Broadway show, and didn't talk about UnixWare any more that
day. After we split up I spent a little longer than usual pacing around
Times Square; I think HBO had on "Speed", a good-enough diversion of the
mind from anything analytical.
The next morning we had a breakfast meeting about moving ahead with
UUX; it's just about ready to roll with formal membership, but now
many things are on hold. We know something's about to happen, we're
just not sure what. We do what needs to be done, and by 10:00 we're
back at Javitz.
I trade my press badge for a plain-white button that's necessary for
admittance to The Announcement, and take my place in a tiny meeting
room with heavy security and no signage.
Unigram had some facts wrong, mostly the level of HP's involvement and
the amount of SCO shares that Novell was going to get. Much is in the
air; questions on NDS's role in DCE and the status of UnixWare OEMs are
not really addressed; some reporters spend more time making statements
than asking questions. I'd found out later that the official statement
had been done by a committee of all three participants, with every word
pored over in minute detail and tons of last-minute changes in evidence.
By the time I reached the show floor it was old news, except for the
trivia. There was relief that UnixWare still existed by Wednesday noon;
but uncertainty for the future of the product, and certainly the future
of some people who were still cheerfully doing demonstrations at the
Novell booth. What would stay? What would go? Who would go? Did anyone
give a damn anymore about those Explorer CDs?
The rest of the show ran kinda strange; I'd finished my reporter duties,
and went about the usual job of collecting toys for the kids; sandals from
Sun, sports bags from HP, balsa airplanes from OSF, and so on. Those few
of us who attended the UW BOF got a chance to put things in perspective
and mull the events over a bit in a quieter atmosphere than the show floor.
That's when it settled in for me that this deal was be a pretty good
idea if the right things fall into place. Still, a pretty big "if"...
The exhibitors' party Wednesday night seemed to provide just the right
amount of oblivion -- who would have thought that a band called The Nerds
could be that good? I guess you had to be there; one moment they're doing
the Diamond/Streisand duet "You Don't Bring Me Flowers", next they're
doing a damn good rip into Springsteen's "Rosalita".
What happened this morning? Shut up and have another Corona.
By Thursday I think Dan and I just wanted to get home. Just In Case, I put
in a lengthy appearance at the Caldera booth. They were quite happy about
the deal, figuring it would put a greater spotlight on Linux as *the*
alternative Unix to SCO now that UnixWare was no longer a competitive force.
I said my goodbyes to numerous people who had helped with the UUX booth;
some I wondered if I'd ever see again, some I had no doubt I would. After
all, this industry isn't *that* big; lots of people tend to re-appear in
different places.
Visits to the UUX booth had dropped off to nothing by Thursday, and
I caught an earlier-than-planned flight home. I'd never thought that a
forest-green Dodge Caravan with Ontario plates would ever be such a
welcome sight -- but by that time, a little stability in my life went
a very long way.
John Moyer
>In further support for solaris - I just read that Sunsoft has internally
>ported Solaris to a power PC -based system.
I saw Solaris running on a PowerPC last spring. It did not yet at that
time have the CDE desktop. I dont know if it is shipping yet. I do not
know whether Solaris is better or worse than AIX on the PowerPC. It
appears to me that each has some advantages over the other.
--
John Moyer http://www.questconsult.com/
Quest Consultants Inc. (tm) j...@questconsult.com KC5GSX
P.O. Box 721387 (405) 329-7475
Norman, Oklahoma 73070-8069, USA Fax: (405) 329-7734
W. Kent Rutan
> After thinking about everything that has unravelled in the last week,
> I can't argue. With SCO not supporting CDE, X86 only, I truly think Solaris
> is the future Unix that will be on my intel machine. Runs on SPARC, x86, PPC.
> Supports CDE. tough to beat.
Yep. Now that SCO admits that SVR4 has something to offer, doesn't it
make one think that Sun is in a much better position to forge ahead
while SCO is thrashing about for a couple of years trying to get
things together (literally).
You have to give Sun a lot of credit for painfully supporting
standards such as Motif and CDE for the better of the Unix industry.
I get the impression that SCO likes the Microsoft approach to
standards: "we have the market share, so we make the standards".
I'm just waiting for a competitive upgrade offer for Solaris 2.5 that
rivals the UnixWare AS+SDK bundle deal. Hopefully Sun will take this
opportunity to grab some market share.
--
========================================================================
ru...@cat.com Caterpillar Inc.
Tel (309) 578-2278 Kent Rutan Technical Center
Fax (309) 578-2354 Peoria, Illinois USA
========================================================================
The opinions expressed herein are mine, not Cat's [etc., ad infinitum].
Evan Leibovitch
W. Kent Rutan <ru...@cat.com> wrote:
>Yep. Now that SCO admits that SVR4 has something to offer,
Actually, it admits that SVR4.2 has something to offer, and there's
a world of difference between 4.0 and 4.2; MoOLIT has come and gone,
Novell created its own desktop manager for, SMP and NetWare support
has been added, hardware support extended to auto-configure PCI and
EISA boards, and the SDK dropped in price to $100.
>doesn't it
>make one think that Sun is in a much better position to forge ahead
>while SCO is thrashing about for a couple of years trying to get
>things together (literally).
No it doesn't, for Sun is one of the few companies capable of worse
marketing of a Unix-on-Intel product than Novell -- right down there
with Dell.
Sun has proven time and time again that it can't market its X86 product
out of a paper bag. Its user relations are worse than Novell's, and its
VAR/reseller programs are worse than SCO's. The worst of both worlds.
>You have to give Sun a lot of credit for painfully supporting
>standards such as Motif and CDE for the better of the Unix industry.
I give no such credit. What a joke. Sun fought Motif until the bitter
end, pushing its own GUI view of the world for so long that it may no
longer be relevant. Sun certainly resisted System V (of *any* kind)
longer than any other major vendor; its central role in all of the
Unix wars cannot be dismissed.
As for CDE, who cares?
>I get the impression that SCO likes the Microsoft approach to
>standards: "we have the market share, so we make the standards".
Oh, please. Sun's arrogant GUI battles with all the companies that
standardized on Motif set the Unix industry back by years. Sun, like any
other vendor, complies with standards for the same reason as any other
vendor: greed. They don't want to get shut out of contracts which
specify standards compliance. After a while Sun just got tired of its
Unix and its GUI being the Betamax of the Unix world.
>I'm just waiting for a competitive upgrade offer for Solaris 2.5 that
>rivals the UnixWare AS+SDK bundle deal. Hopefully Sun will take this
>opportunity to grab some market share.
Solaris X86 will go the way of Sun's other Intel-based efforts; into the
toilet. The company doesn't have its heart in the X86 product since it
can't make any hardware sales. And Sun certainly hasn't adapted to the
oddities of the commodity hardware market like SCO (and to a lesser extent,
Novell) did.
The half-hearted effort that Sun has put into Solaris X86 gives one no
more assurance about its future than anyone had about UnixWare's.
IMO Solaris is one of the collateral casualties of the Novell/SCO
deal. Caught between SCO/UnixWare at one end and Linux/NetBSD at the
other, it will langish until Sun puts it out of its misery.
Simon Ritter
<stuff deleted>
: The one consolation: almost no-one, in Europe at least, knew anything about
: this deal prior to Wednesday. I was at the Novell headquarters in Paris just
: that morning, chitchatting away when the French CEO busted in with the draft
: press release that came out later in the US. The guy from USG in London almost
: fell off his chair, _and_ he didn't know all the details yet (obviously Novell
: tried to put a happy face on it, but it's bloody treason all the same,
: whatever the beneficial long-term consequences to Unix itself...). I talked to
I was the unfortunate Novell person in that meeting. Due to the high level
of secrecy surrounding the announcement, the first I knew about the details
was when the customer translated the press release for me which had been
presented in French. Not a meeting that I want to repeat from that point of
view.
I did, however, enjoy lunch.
-----------------------------------------------------------------------------
Simon Ritter | Sell your house, sell your car,
Senior Consultant (The UNIX^WSCO bits) | sell everything you own, but don't
Novell Consulting Services (Europe) | sell out.
sim...@novell.co.uk | Life is Vertical!
-----------------------------------------------------------------------------
DISCLAIMER:
The views expressed above are my own and in no way reflect those of my
employer, whether they be past, present or future.
-----------------------------------------------------------------------------
nsch...@aardvark.ucs.uoknor.edu
wrote:
> *******************************************************************
> Martin F. Sohnius msoh...@novell.co.uk
> Novell IS & T, Bracknell, England +44-1344-724031
> *******************************************************************
> * if (status = UNDER_NUCLEAR_ATTACK) *
> * launch_full_counterstrike(); *
> *******************************************************************
Martin,
Do you mean
if (status == UNDER_NUCLEAR_ATTACK}
or are you really warlike?
Joe Doupnik
Sigh. I guess there is always a faction of the population without
a sense of humor. I do believe the signature is the way it is because of
precisely this syntax; it's great. One aspect of British humor is
everyone except the gullible knows there is a joke being played, and the
latter is allowed to proceed for a long time with not a hint being passed.
Stiff upper wit (or is that twit).
Joe D.
Scott Williams
>As for CDE, who cares?
Why don't you care about CDE? This was one way of providing a
moderately appealing interface (to end-users at least) on a
bunch of different platforms. To my mind the fact that HP-UX,
Solaris, and AIX all look the same to an end-user is a plus.
>Oh, please. Sun's arrogant GUI battles with all the companies that
>standardized on Motif set the Unix industry back by years.
No kidding. If you have to take sides, Sun's role in decelerating
the acceptance and development of Unix was probably more significant
than anyone elses.
>The company doesn't have its heart in the X86 product since it
>can't make any hardware sales. And Sun certainly hasn't adapted to the
>oddities of the commodity hardware market like SCO ...
Not to mention whatever happened to ISC (not that ISC was any
great raving hell at marketing either, mind you).
>IMO Solaris is one of the collateral casualties of the Novell/SCO
>deal. Caught between SCO/UnixWare at one end and Linux/NetBSD at the
>other, it will langish until Sun puts it out of its misery.
Sad but probably true.
Scott.
Martin Sohnius
You ask any Englishman to explain the rules of cricket, and you get exactly
that funny smile that says "hey, if you don't know yet that it is all a
con, *I* am not going to be the one to tell you!" He'll then proceed to
explain the most far-flung *exception* to the rule he can think of.....
--
*******************************************************************
Martin F. Sohnius msoh...@novell.co.uk
Novell IS & T, Bracknell, England +44-1344-724031
*******************************************************************
* if (status = UNDER_NUCLEAR_ATTACK) *
* launch_full_counterstrike(); *
*******************************************************************
Bryan Althaus
: Why don't you care about CDE? This was one way of providing a
: moderately appealing interface (to end-users at least) on a
: bunch of different platforms. To my mind the fact that HP-UX,
: Solaris, and AIX all look the same to an end-user is a plus.
I currently run CDE on Solaris 2.4 and I give NOVELL/HP/SUN/IBM alot
of credit.
: >Oh, please. Sun's arrogant GUI battles with all the companies that
: >standardized on Motif set the Unix industry back by years.
: No kidding. If you have to take sides, Sun's role in decelerating
: the acceptance and development of Unix was probably more significant
: than anyone elses.
Sun had OpenLook out for years before Motif was even a dream. The only
reason for Motif was to combat Sun's OpenLook along with OSF/1. So
is obvious to see why after being ganged up on, Sun wasn't exactly
thrilled with having to go to Motif.
I like Motif, never was a OpenLook person but it is funny to see that
OpenLook is free and open while Motif still requires a licen$e.
Just check out the X programs that come with linux, they are usually
OpenLook based.
As for setting the UNIX industry back, this didn't help but Sun has
done more for UNIX than any other company. And had these companies
improved OpenLook and worked on adopting SVR4 instead of wasting
years on OSF/1 everyone would have been better off.
: >The company doesn't have its heart in the X86 product since it
: >can't make any hardware sales. And Sun certainly hasn't adapted to the
: >oddities of the commodity hardware market like SCO ...
: Not to mention whatever happened to ISC (not that ISC was any
: great raving hell at marketing either, mind you).
SunSoft still markets it, but they used ISC for expertise in x86 UNIX
to help on Solaris x86.
: >IMO Solaris is one of the collateral casualties of the Novell/SCO
: >deal. Caught between SCO/UnixWare at one end and Linux/NetBSD at the
: >other, it will langish until Sun puts it out of its misery.
: Sad but probably true.
Sun has big plans for Solaris on both x86 and PowerPC. Solaris x86
will be used as a platform to compete with NT. Sun already is releasing
their Netra Server software for Solaris x86 to be used as an internet
server out of the box, Netscape now considers Solaris x86 a supported
product, and Oracle is supporting Solaris x86 as a good alternative to
NT/SQLServer.
I see only promise for Solaris x86, especially now that even Sun
realises not everyone can afford a Sparc or even those who can,
might want to mix and match Sparc's with cheap Pentiums.
As for the future of x86 UNIX, it will be a couple years before the combined
UnixWare/SCO/HP OS is shipping, but thankfully SCO and HP will then be SVR4
based, with real SMP & thread support. HP support for threads in HP-UX 10.0
is really sad.
So all in all I think everyone wins. SVR4 becomes even more standard,
SCO gets a real underlying OS as does HP.
Since Solaris Sparc/x86/PowerPC use a common code base,
--
Regards,
Bryan Althaus br...@krf.com b...@netcom.com
Mike Convey
Martin Sohnius <msoh...@novell.co.uk> wrote:
>You ask any Englishman to explain the rules of cricket, and you get exactly
>that funny smile that says "hey, if you don't know yet that it is all a
>con, *I* am not going to be the one to tell you!" He'll then proceed to
>explain the most far-flung *exception* to the rule he can think of.....
Sounds more like Mornington Crescent :-) (c.f. alt.games.mornington.cresent).
Mike
Jim Vlcek
> No kidding. If you have to take sides, Sun's role in decelerating
> the acceptance and development of Unix was probably more significant
> than anyone elses.
You have no idea of what you're talking about. The truth more closely
approaches the precise opposite of this statement.
--
Jim Vlcek I don't understand...
vl...@byteware.com ...it's all UNIX to me.
The Black Box of Lowertown
Beautiful downtown St. Paul
c...@interaccess.com
>SUN has always set the pace for UNIX because they are 100% UNIX from the
>beginning.
>Michael J Matthews mmat...@fast.net
Wrong... AT&T WAS the grand dad of Unix, which went to USL (Unix
Systems Labs), then to Novell, now to SCO. Sun, is nothing but a *nix
clone. They were not THE Unix... They may be the favorite, perhaps the
only one that stayed true to the AT&T Unix style, but to say they are
the true Unix from the beginning is just false...
----------------------------------------------------------------------------------------------------------
C.D. Silva, Esq. | And on the 8th day, God looked down on these
* Unix Powered. | United States and saw that it was good.
* NetWare Ready. | November 8, 1994
* OS/2 Tolerant. | *************************
* NT Pending. | http://www.interaccess.com/cds
----------------------------------------------------------------------------------------------------------
Michael J Matthews
>Scott Williams (sc...@willcam.com) wrote:
>: ev...@telly.telly.org (Evan Leibovitch) wrote:
>: >As for CDE, who cares?
>
>: Why don't you care about CDE? This was one way of providing a
>: moderately appealing interface (to end-users at least) on a
>: bunch of different platforms. To my mind the fact that HP-UX,
>: Solaris, and AIX all look the same to an end-user is a plus.
>
>I currently run CDE on Solaris 2.4 and I give NOVELL/HP/SUN/IBM alot
>of credit.
>
>: >Oh, please. Sun's arrogant GUI battles with all the companies that
>: >standardized on Motif set the Unix industry back by years.
>
>: the acceptance and development of Unix was probably more significant
>: than anyone elses.
And look what happened to them (billions in losses later). SUN bent over
backwards trying to appease these fools. Of course the BS is over now and
MS is on the move.
>
>Sun had OpenLook out for years before Motif was even a dream. The only
>reason for Motif was to combat Sun's OpenLook along with OSF/1. So
>is obvious to see why after being ganged up on, Sun wasn't exactly
>thrilled with having to go to Motif.
>
>I like Motif, never was a OpenLook person but it is funny to see that
>OpenLook is free and open while Motif still requires a licen$e.
>Just check out the X programs that come with linux, they are usually
>OpenLook based.
SUN gave Xview away once they knew no one was commercially interested.
Even before SUN gave up to Motif there was OLIT which IMHO was superior
to MOTIF at one time but not significantly different. There is not enough
pie for 2 toolkit standards.
>
>As for setting the UNIX industry back, this didn't help but Sun has
>done more for UNIX than any other company. And had these companies
>improved OpenLook and worked on adopting SVR4 instead of wasting
>years on OSF/1 everyone would have been better off.
Agreed!
SUN has always set the pace for UNIX because they are 100% UNIX from the
beginning. THis is not to say they don't need competition. But how about
some intelligent competition.
I will give an honorable mention to HP for commiting to UNIX at an early
stage even though it was a no-brainer even then. Weren't they OSF traitors
also? :)
--
/*---------------------------------------------------*/
Michael J Matthews mmat...@fast.net
Scott Williams
>Scott Williams <sc...@willcam.com> writes
>> No kidding. If you have to take sides, Sun's role in decelerating
>> the acceptance and development of Unix was probably more significant
>> than anyone elses.
>
>approaches the precise opposite of this statement.
I might rephrase my first statement to say "decelerating the acceptance"
of Unix rather than "acceptance and development". With that change, I'll
stand by my original statement.
On the "development of Unix" side, I, like many other Unix users, am a
big fan of most of the technology Sun has developed over the years.
I've been a happy user of various versions of SunOS and Solaris for a
long time, not to mention little things like NFS.
But IMO, the "accpetance" of Unix has a lot more to do with the
consistency of the technology - the degree of consensus amongst the
players - than on the absolute excellence of the technology. That, IMO,
was the essence of the Motif/OpenLook war. Lots of folks wanted to stick
with OpenLook because they felt it was simply superior. But when the
rest of the Unix world - not just the wrench-tossers like IBM and DEC -
had settled on Motif, it was couterproductive for Sun to dogmatically
stick to it's OpenLook guns.
The OpenLook/Motif issue was of enormous importance to anyone trying to
sell Unix back then. There might have been things that customers would
have been better to focus on, but the first thing the end-user would
complain about was that their workstation didn't "look" like the
workstation of the person down the hall.
My disillusionment with the Unix industry isn't isolated to Sun. I think
all the players - IBM, DEC and HP most certainly included - need to take
a *whole* *lot* of responsiblity for the problems of Unix acceptance.
But I have often felt - and would stand by this - that when other players
had finally given up on their proprietary axes-to-grind, Sun would hold
out for a while longer, decelerating a slow process further. To give
Sun credit where it's due in this "acceptance" thing though, I think
they give a whole lot more credibility to things like CDE than any of
the other players, including HP (which given the relationship between
CDE and VUE seems ironically self-interested).
I write and teach Unix courses for a living and I think generally I'm
getting tired of explaining to students why one version of Unix is so-
close-but-quite-the-same as every other version.
Scott.
Dave Parker
>mmat...@fast.net (Michael J Matthews) wrote:
>
>>beginning.
>
>
>Wrong... AT&T WAS the grand dad of Unix, which went to USL (Unix
>Systems Labs), then to Novell, now to SCO. Sun, is nothing but a *nix
>clone. They were not THE Unix... They may be the favorite, perhaps the
>only one that stayed true to the AT&T Unix style, but to say they are
>the true Unix from the beginning is just false...
Logic 101:
(SUN == 100% UNIX) != (UNIX == 100% SUN)
Martin Sohnius
: mmat...@fast.net (Michael J Matthews) wrote:
: >SUN has always set the pace for UNIX because they are 100% UNIX from the
: >beginning.
: >Michael J Matthews mmat...@fast.net
: Wrong... AT&T WAS the grand dad of Unix, which went to USL (Unix
: Systems Labs), then to Novell, now to SCO. Sun, is nothing but a *nix
: clone. They were not THE Unix... They may be the favorite, perhaps the
: only one that stayed true to the AT&T Unix style, but to say they are
: the true Unix from the beginning is just false...
You guys really are in the process of reliving every single battle in the
Great Unix Wars, aren't you?
Remember, there are, and always have been two major strands of UNIX, AT&T
and Berkeley. Distinguished, among other things whether !true is pronounced
"shriek true" or "bang true", and more importantly, whether the shell prompt is
$ or % :-). Whereas AT&T was, in fact, the "granddad of UNIX" (after all,
it was Dennis Ritchie who first brought Unix to the kids out in Calif, back
in '75), it was and is SUN who are most associated with UNIX in the academic
world even today (with SunOS). SunOs' ancestry in BSD is as impeccable as
is SCO's or ISC's in Bell Labs. It's actually SVR4 (and Solaris) which is
the true bastard child, and it took Sun some 6 years to adopt it.
Considering that it was their participation in SVR4 that brought about the
Second Great Unix War between the OSF and UI (the first was between BSD and
System V) it would have behooved them to get there a bit quicker. A
positive aspect, IMHO, is that the Regents of the University of California
(a far greedier lot than any "owner" of New Jersey UNIX has ever been) are
finally out of the main stream commercial picture (assuming the ultimate
demise of Ultrix).
Bryan Althaus
: In article <RUTAN.95S...@flea.cat.com>,
: W. Kent Rutan <ru...@cat.com> wrote:
: >doesn't it
: >make one think that Sun is in a much better position to forge ahead
: >while SCO is thrashing about for a couple of years trying to get
: >things together (literally).
: No it doesn't, for Sun is one of the few companies capable of worse
: marketing of a Unix-on-Intel product than Novell -- right down there
: with Dell.
: Sun has proven time and time again that it can't market its X86 product
: out of a paper bag. Its user relations are worse than Novell's, and its
: VAR/reseller programs are worse than SCO's. The worst of both worlds.
Agreed. SunSoft has said with NT sales taking off ( 99% of NT sales
are on x86 hardware ), Solaris x86 will become an important product.
: >You have to give Sun a lot of credit for painfully supporting
: >standards such as Motif and CDE for the better of the Unix industry.
: I give no such credit. What a joke. Sun fought Motif until the bitter
: end, pushing its own GUI view of the world for so long that it may no
: longer be relevant.
And why? Which came first. Why was Motif invented in the first place?
IBM/DEC/HP etc. ganged up on Sun and you expect the company just to
give up without being somewhat bitter.
But that was then, now Sun is shipping CDE with Solaris 2.5 and is willing
to support industry standards. It already supports POSIX threads. It
is compling with Spec-1170(sp).
: Sun certainly resisted System V (of *any* kind)
: longer than any other major vendor;
Sun's customer base at the time cared very little about System V. When
Sun went to Solaris (System V) the customers complained at first. It's
not an easy call.
By the way SunOS 4.1.x does have alot of System V in it, its just based
on a BSD kernel, so I'm not sure what you are going on about.
: its central role in all of the Unix wars cannot be dismissed.
The "WAR" was caused by IBM & DEC not wanting UNIX to eat away at their
proprietary systems. Why is it that no one but DEC ever adopted OSF/1?
Sun's Solaris is based on SVR4, so look who won.
: As for CDE, who cares?
never mind....
: >I get the impression that SCO likes the Microsoft approach to
: >standards: "we have the market share, so we make the standards".
: Oh, please. Sun's arrogant GUI battles with all the companies that
: standardized on Motif set the Unix industry back by years. Sun, like any
: other vendor, complies with standards for the same reason as any other
: vendor: greed. They don't want to get shut out of contracts which
: specify standards compliance. After a while Sun just got tired of its
: Unix and its GUI being the Betamax of the Unix world.
See above.
: >I'm just waiting for a competitive upgrade offer for Solaris 2.5 that
: >rivals the UnixWare AS+SDK bundle deal. Hopefully Sun will take this
: >opportunity to grab some market share.
: Solaris X86 will go the way of Sun's other Intel-based efforts; into the
: toilet. The company doesn't have its heart in the X86 product since it
: can't make any hardware sales. And Sun certainly hasn't adapted to the
: oddities of the commodity hardware market like SCO (and to a lesser extent,
: Novell) did.
: The half-hearted effort that Sun has put into Solaris X86 gives one no
: more assurance about its future than anyone had about UnixWare's.
You really hate Sun don't you!
Agreed. But Sun has said it will use Solaris x86 to compete with NT,
and will promote it. And it has gotten vendors to simple recompile
there Solaris Sparc code onto Solaris x86. No porting necessary!
Oracle uses Solaris x86 as the platform to compete with NT/SqlServer.
NetScape considers Solaris x86 a supported product.
: IMO Solaris is one of the collateral casualties of the Novell/SCO
: deal. Caught between SCO/UnixWare at one end and Linux/NetBSD at the
: other, it will langish until Sun puts it out of its misery.
As long as you realise that your opinion, and your opinion has been
wrong over the years.
I think the reason your so bitter with Solaris is that Sun still supports
it, and will long after the word UnixWare is a distant, fadding memory..
Bryan Althaus
: In article <bryDG3...@netcom.com>, Bryan Althaus <b...@netcom.com> wrote:
: >Evan Leibovitch (ev...@telly.telly.org) wrote:
: >: Sun has proven time and time again that it can't market its X86 product
: >: out of a paper bag. Its user relations are worse than Novell's, and its
: >: VAR/reseller programs are worse than SCO's. The worst of both worlds.
: >Agreed. SunSoft has said with NT sales taking off ( 99% of NT sales
: >are on x86 hardware ), Solaris x86 will become an important product.
: You agree on my assessment of Sun's marketing; yet you say that
: Sun's *belief* that Solaris X86 will be an important product has
: anything to do with whether or not that will come to pass?
Let me talk alittle slower for you. I agree that up until now Sun's
marketing of Solaris x86 has been poor. But with NT generating sales
SunSoft has come to realise that Solaris x86 is an important product,
and in the coming months they will make there x86 intentions clear.
If they didn't think they could market Solaris x86, why would they waste
time coming out with a PowerPC version?
: That's one for rec.humor.
That and your 10 page rantings about what? UnixWare. Talk about religious.
: Sun can have all the wishful thinking it wants. I can run all the glossy
: ads it wants in SCO World touting its new GUI-less edition of x86 Solaris.
: That doesn't mean it'll sell a single one.
It does mean that they are *trying* to do something. And now with SCO
spending the next few years rewriting their kernel, some might want
to jump ship.
: >: >You have to give Sun a lot of credit for painfully supporting
: >: >standards such as Motif and CDE for the better of the Unix industry.
: >: I give no such credit. What a joke. Sun fought Motif until the bitter
: >: end, pushing its own GUI view of the world for so long that it may no
: >: longer be relevant.
: >And why? Which came first. Why was Motif invented in the first place?
: >IBM/DEC/HP etc. ganged up on Sun and you expect the company just to
: >give up without being somewhat bitter.
: Bitter, shmitter. This is business.
If it was really business, then Motif would never had been created. They
would have worked on improving OpenLook.
: >: Sun certainly resisted System V (of *any* kind)
: >: longer than any other major vendor;
: >By the way SunOS 4.1.x does have alot of System V in it, its just based
: >on a BSD kernel, so I'm not sure what you are going on about.
: Real (licensed) System V stuff, or cloned facsimiles?
Can't admit when your wrong. You said Sun resised System V (of *any* kind),
and I'm telling you System V stuff has been in there for years.
It was put in there for the Wall St. people.
: >: its central role in all of the Unix wars cannot be dismissed.
: >The "WAR" was caused by IBM & DEC not wanting UNIX to eat away at their
: >proprietary systems. Why is it that no one but DEC ever adopted OSF/1?
: >Sun's Solaris is based on SVR4, so look who won.
: I said nothing about winners and losers. I fault Sun for dragging on the
: war, that may have helped its own fortunes at the expense of the whole
: Unix indistry.
Your mad at Sun for dragging on the war, but the fact that IBM & DEC
started the war to protect their proprietary doesn't seem to register.
: >: The half-hearted effort that Sun has put into Solaris X86 gives one no
: >: more assurance about its future than anyone had about UnixWare's.
: >You really hate Sun don't you!
: I don't hate *any* companies; they're not worth that strong an emotion.
: While I very much admire what Sun does, and the position it has established
: in the industry, I maintain that Intel platforms are its Achilles' heel.
: I have a number of first-hand experiences that Sun does not know this
: market, and likely never will.
: We have already seen one abortive Intel effort by Sun that hung many,
: many users out to dry. How long will the next one last?
Longer than the 2 years Novell had UnixWare. Sun deserves a chance
to see if they can do something with a great product.
: >But Sun has said it will use Solaris x86 to compete with NT,
: >and will promote it.
: ...using all the marketing expertise at its disposal? Using all
: the skills that have made Interactive the leader in *its* field?
: Yeah, right.
: >And it has gotten vendors to simple recompile
: >there Solaris Sparc code onto Solaris x86. No porting necessary!
: That it is possible does not mean it's happening. The number of Catalyst
: participants who actually support Intel versions of their stuff is not
: very high. The actual effort to port is only a small part of the
: expense of supporting a new hardware platform, especially one with
: such a broad range of commodity hardware as Intel boxes. And given the
: miniscule installed base Solaris X86 actually has, who can blame ISVs
: if they don't believe the sales potential is worth the effort?
: >Oracle uses Solaris x86 as the platform to compete with NT/SqlServer.
: It also uses SCO and UnixWare; the platform it chose to call OracleWare
: uses UnixWare as a base. Oracle works with all the OS vendors; what's
: your point?
They dropped the OracleWare because it wasn't selling. My point is
ORacle uses Solaris x86 as the OS to run on when competing for
contracts against NT.
: >NetScape considers Solaris x86 a supported product.
: Whop-de-doo. They support Linux and NetBSD too.
But not UnixWare or SCO. All the above are free UNIX's which they
support as a favor.
: Do you really want to start counting the amount of Intel-Unix apps and
: hardware supported by SCO compared to Solaris? Do you really think that
: Netscape's exception-to-the-rule can win you this particular pissing match?
That was not my point. I'm just pointing out that after years, companies
are taking Solaris x86 seriously.
: >: IMO Solaris is one of the collateral casualties of the Novell/SCO
: >: deal. Caught between SCO/UnixWare at one end and Linux/NetBSD at the
: >: other, it will langish until Sun puts it out of its misery.
: >As long as you realise that's your opinion, and your opinion has been
: >wrong over the years.
: The first part of your sentence doesn't grok; the second has been
: refuted by people within both SCO and Novell.
I fixed it, reread.
: It still stands as a fact that Sun has already abandoned its only other
: foray into the Intel world. It's also true that Solaris X86 has a miniscule
: share of the Intel Unix market, way behind UnixWare. And you agree with
: me that Sun's marketing efforts in this field suck big time.
: Judging by its market share, Solaris X86 is *already* languishing. What
: great moves is Sun capable of doing to change this?
: When Solaris X86 first came out I said it would not seriously challenge
: UnixWare or Solaris for Intel market share. While not every opinion I
: have expressed has been realized, this one most certainly has.
: >I think the reason your so bitter with Solaris is that Sun still supports
: >it, and will long after the word UnixWare is a distant, fadding memory..
: UnixWare will continue in much its present form, and name, into 1997. What
: possible signs outside of Sun's (and your) own wishful thinking give you
: the belief that Solaris X86 will last that long?
: Solaris on SPARC is great stuff; that will last for a long, long time.
: Sun's second try on Intel, however, will inevitably meet the same fate
: as its first.
For the industry, lets hope not. If SCO becomes the only x86 UNIX,
it will be a clear sign to go with NT, even for the religious.
Evan Leibovitch
>Evan Leibovitch (ev...@telly.telly.org) wrote:
>: Sun has proven time and time again that it can't market its X86 product
>: out of a paper bag. Its user relations are worse than Novell's, and its
>: VAR/reseller programs are worse than SCO's. The worst of both worlds.
>Agreed. SunSoft has said with NT sales taking off ( 99% of NT sales
>are on x86 hardware ), Solaris x86 will become an important product.
You agree on my assessment of Sun's marketing; yet you say that
Sun's *belief* that Solaris X86 will be an important product has
anything to do with whether or not that will come to pass?
That's one for rec.humor.
Sun can have all the wishful thinking it wants. I can run all the glossy
ads it wants in SCO World touting its new GUI-less edition of x86 Solaris.
That doesn't mean it'll sell a single one.
>: >You have to give Sun a lot of credit for painfully supporting
>: >standards such as Motif and CDE for the better of the Unix industry.
>: I give no such credit. What a joke. Sun fought Motif until the bitter
>: end, pushing its own GUI view of the world for so long that it may no
>: longer be relevant.
>And why? Which came first. Why was Motif invented in the first place?
>IBM/DEC/HP etc. ganged up on Sun and you expect the company just to
>give up without being somewhat bitter.
Bitter, shmitter. This is business.
>: Sun certainly resisted System V (of *any* kind)
>: longer than any other major vendor;
>By the way SunOS 4.1.x does have alot of System V in it, its just based
>on a BSD kernel, so I'm not sure what you are going on about.
Real (licensed) System V stuff, or cloned facsimiles?
>: its central role in all of the Unix wars cannot be dismissed.
>The "WAR" was caused by IBM & DEC not wanting UNIX to eat away at their
>proprietary systems. Why is it that no one but DEC ever adopted OSF/1?
>Sun's Solaris is based on SVR4, so look who won.
I said nothing about winners and losers. I fault Sun for dragging on the
war, that may have helped its own fortunes at the expense of the whole
Unix indistry.
>: The half-hearted effort that Sun has put into Solaris X86 gives one no
>: more assurance about its future than anyone had about UnixWare's.
>You really hate Sun don't you!
I don't hate *any* companies; they're not worth that strong an emotion.
While I very much admire what Sun does, and the position it has established
in the industry, I maintain that Intel platforms are its Achilles' heel.
I have a number of first-hand experiences that Sun does not know this
market, and likely never will.
We have already seen one abortive Intel effort by Sun that hung many,
many users out to dry. How long will the next one last?
>But Sun has said it will use Solaris x86 to compete with NT,
>and will promote it.
...using all the marketing expertise at its disposal? Using all
the skills that have made Interactive the leader in *its* field?
Yeah, right.
>And it has gotten vendors to simple recompile
>there Solaris Sparc code onto Solaris x86. No porting necessary!
That it is possible does not mean it's happening. The number of Catalyst
participants who actually support Intel versions of their stuff is not
very high. The actual effort to port is only a small part of the
expense of supporting a new hardware platform, especially one with
such a broad range of commodity hardware as Intel boxes. And given the
miniscule installed base Solaris X86 actually has, who can blame ISVs
if they don't believe the sales potential is worth the effort?
>Oracle uses Solaris x86 as the platform to compete with NT/SqlServer.
It also uses SCO and UnixWare; the platform it chose to call OracleWare
uses UnixWare as a base. Oracle works with all the OS vendors; what's
your point?
>NetScape considers Solaris x86 a supported product.
Whop-de-doo. They support Linux and NetBSD too.
Do you really want to start counting the amount of Intel-Unix apps and
hardware supported by SCO compared to Solaris? Do you really think that
Netscape's exception-to-the-rule can win you this particular pissing match?
>: IMO Solaris is one of the collateral casualties of the Novell/SCO
>: deal. Caught between SCO/UnixWare at one end and Linux/NetBSD at the
>: other, it will langish until Sun puts it out of its misery.
>As long as you realise that your opinion, and your opinion has been
>wrong over the years.
The first part of your sentence doesn't grok; the second has been
refuted by people within both SCO and Novell.
It still stands as a fact that Sun has already abandoned its only other
foray into the Intel world. It's also true that Solaris X86 has a miniscule
share of the Intel Unix market, way behind UnixWare. And you agree with
me that Sun's marketing efforts in this field suck big time.
Judging by its market share, Solaris X86 is *already* languishing. What
great moves is Sun capable of doing to change this?
When Solaris X86 first came out I said it would not seriously challenge
UnixWare or Solaris for Intel market share. While not every opinion I
have expressed has been realized, this one most certainly has.
>I think the reason your so bitter with Solaris is that Sun still supports
>it, and will long after the word UnixWare is a distant, fadding memory..
UnixWare will continue in much its present form, and name, into 1997. What
possible signs outside of Sun's (and your) own wishful thinking give you
the belief that Solaris X86 will last that long?
Solaris on SPARC is great stuff; that will last for a long, long time.
Sun's second try on Intel, however, will inevitably meet the same fate
as its first.
--
Bryan Althaus
: Bryan Althaus (b...@netcom.com) wrote:
: : And now with SCO
: : spending the next few years rewriting their kernel, ....
: Just for the record, how long did it take them to move from XENIX to the
: SVR3.2 based SCO UNIX? Not exactly a "few years", if I remember correctly.
: Considering that it took Sun about 6 years after SVR4 was ready to bring
: Solaris to market, SCO's track record is decidedly better.
If you believe SCO can take their lame SVRV3.2 kernel OS, at least that was
what the Novell people like yourself said about SCO before they bought
you out, and make it a true SVR4 kernel with kernel threading & SMP
the way its done on UnixWare 2.0, make it backward compatiable for SCO
& UnixWare customers and make it rock solid in less than 2 years
you are kidding yourself.
Then after that version comes out, they have the task of again merging
this kernel with HP's new 64bit OS.
The timeline for all this is Mid 1997 for the SCO/UnixWare merge and
*sometime* in the 1998! for the final 64bit Unix on HP/Intel P7 release.
This according to this weeks InformationWeek's article on the merger
in which a timeline is given on page 60.
: : Longer than the 2 years Novell had UnixWare. Sun deserves a chance
: : to see if they can do something with a great product.
: Except of course, that everyone except you shares the opinion that UnixWare
: is a far greater product than Solaris x86.
And everyone in the OS/2 group thinks OS/2 is a far greater product than
anything else. In the Solaris group you would find a differ view from
UnixWare converts like myself.
UnixWare is a fine OS but to think it is better than Solaris x86 is only
kidding one's self. Since you work at Novell you don't have the benefit
to have used in the last few years both UnixWare, Solaris, NT and HP-UX
and come up with an unbiased result.
Also, I only posted about Solaris because *UnixWare people* kept saying
they are not happy with going with SCO, you remember SCO the company
this group loves to bash, and so I added info on Solaris.
: : : It also uses SCO and UnixWare; the platform it chose to call OracleWare
: : : uses UnixWare as a base. Oracle works with all the OS vendors; what's
: : : your point?
: : They dropped the OracleWare because it wasn't selling. My point is
: : ORacle uses Solaris x86 as the OS to run on when competing for
: : contracts against NT.
: Bullshit. They use SCO, or UnixWare, or NetWare, or Solaris, or whatever
: the hardware vendor and/or client is most comfortable with.
Well they did drop OracleWare. The CEO of Oracle said OracleWare was
a mistake. As for contracts, I meant with the government.
Sun & Oracle have a joint marketing & sales dept. just to promote
Oracle on Solaris, with Solaris x86 being what is used to compete
with NT/SqlServer for price/performance.
Oracle does not have this relationship with SCO.
: : That was not my point. I'm just pointing out that after years, companies
: : are taking Solaris x86 seriously.
: You mean *you* are taking it seriously.
That's what I meant, I'm a company. Yeah...
Read comp.unix.solaris and to see who takes Solaris x86 seriously.
You can't tell me that before the merger you took SCO seriously!!
Martin Sohnius
: And now with SCO
: spending the next few years rewriting their kernel, ....
Just for the record, how long did it take them to move from XENIX to the
SVR3.2 based SCO UNIX? Not exactly a "few years", if I remember correctly.
Considering that it took Sun about 6 years after SVR4 was ready to bring
Solaris to market, SCO's track record is decidedly better.
: Longer than the 2 years Novell had UnixWare. Sun deserves a chance
: to see if they can do something with a great product.
Except of course, that everyone except you shares the opinion that UnixWare
is a far greater product than Solaris x86.
: : It also uses SCO and UnixWare; the platform it chose to call OracleWare
: : uses UnixWare as a base. Oracle works with all the OS vendors; what's
: : your point?
: They dropped the OracleWare because it wasn't selling. My point is
: ORacle uses Solaris x86 as the OS to run on when competing for
: contracts against NT.
Bullshit. They use SCO, or UnixWare, or NetWare, or Solaris, or whatever
the hardware vendor and/or client is most comfortable with.
: That was not my point. I'm just pointing out that after years, companies
: are taking Solaris x86 seriously.
You mean *you* are taking it seriously.
--
W. Kent Rutan
Sohnius) writes:
> : That was not my point. I'm just pointing out that after years, companies
> : are taking Solaris x86 seriously.
>
> You mean *you* are taking it seriously.
As am I, several other people that have posted to this group, and many
more lurkers.
Jim Vlcek
> Considering that it took Sun about 6 years after SVR4 was ready to bring
> Solaris to market, SCO's track record is decidedly better.
"After SVR4 was ready" is a key phrase here, Martin. Solaris became workable
with release 2.3, which became available in 1994. That's the same year as
UnixWare 1.1, the first workable version of UnixWare.
In both cases, the "late" to-market date was due to the fact that, while SVR4
incorporated a lot of excellent conceptual developments (like pkgadd, HINT,
HINT to SCO), it was a swarming bugfest as delivered by USL.
> Bryan Althaus (b...@netcom.com) wrote:
> : That was not my point. I'm just pointing out that after years, companies
> : are taking Solaris x86 seriously.
>
> You mean *you* are taking it seriously.
I take Solaris x86 seriously, if for no other reason than that it has the
large installed base of Solaris/SPARC to leverage against.
Evan Leibovitch
>Martin Sohnius writes
>> Considering that it took Sun about 6 years after SVR4 was ready to bring
>> Solaris to market, SCO's track record is decidedly better.
>"After SVR4 was ready" is a key phrase here, Martin. Solaris became workable
>with release 2.3, which became available in 1994. That's the same year as
>UnixWare 1.1, the first workable version of UnixWare.
Don't confuse SVR4.0 (on which Solaris is based) with SVR4.2 (on which
UnixWare is based). While, by my estimations, Martin may be a slight bit
off (six years rather than five), he is essentially right in that the
SVR4 which Solaris used as a base was ready for use quite some time ago.
According to my AT&T documentation, SVR4 was available in source in 1989
(though it might have been available even earlier); I was had usable SVR4
for Intel (Esix 4.0.1) in late 1990.
>In both cases, the "late" to-market date was due to the fact that, while SVR4
>incorporated a lot of excellent conceptual developments (like pkgadd, HINT,
>HINT to SCO), it was a swarming bugfest as delivered by USL.
Esix had no marketing, and its engineering staff was tiny, but it was
certainly a credible (and very usable) product. Others such as
Microport and UHC also had working implementations (Dell was a
latecomer).
>I take Solaris x86 seriously, if for no other reason than that it has the
>large installed base of Solaris/SPARC to leverage against.
And how has that leverage been exploited? Not much. It's easy to argue
that software for SPARC Solaris can be easily ported, but it's a fact
that not much has. Even the oft-used example of Netscape, which is
available on Solaris but not SCO/UnixWare, has apparently not been
ported to Solaris X86.
The SPARC insalled base of Solaris means *nothing* to the Intel world,
except to explain how Sun can afford to lose money on such a complex
operating system (X86) with such poor sales.
Siamak Nazari
>In article <DGA9D...@byteware.com>, Jim Vlcek <vl...@byteware.com> wrote:
>
>>Martin Sohnius writes
>>> Considering that it took Sun about 6 years after SVR4 was ready to bring
>>> Solaris to market, SCO's track record is decidedly better.
>
>>"After SVR4 was ready" is a key phrase here, Martin. Solaris became workable
>>with release 2.3, which became available in 1994. That's the same year as
>>UnixWare 1.1, the first workable version of UnixWare.
>
>Don't confuse SVR4.0 (on which Solaris is based) with SVR4.2 (on which
>UnixWare is based). While, by my estimations, Martin may be a slight bit
>off (six years rather than five), he is essentially right in that the
>SVR4 which Solaris used as a base was ready for use quite some time ago.
>
>According to my AT&T documentation, SVR4 was available in source in 1989
>(though it might have been available even earlier); I was had usable SVR4
>for Intel (Esix 4.0.1) in late 1990.
Neither should you confuse SVR4.0 with Solaris. Solaris in its current
incarnation looks nothing like the code delivered in SVR4. The MP and
Multithreaded support are an integral part of Solaris, so is dynamic
kernel module loading.
>
>>In both cases, the "late" to-market date was due to the fact that, while SVR4
>>incorporated a lot of excellent conceptual developments (like pkgadd, HINT,
>>HINT to SCO), it was a swarming bugfest as delivered by USL.
>
>Esix had no marketing, and its engineering staff was tiny, but it was
>certainly a credible (and very usable) product. Others such as
>Microport and UHC also had working implementations (Dell was a
>latecomer).
>
>>I take Solaris x86 seriously, if for no other reason than that it has the
>>large installed base of Solaris/SPARC to leverage against.
>
>And how has that leverage been exploited? Not much. It's easy to argue
>that software for SPARC Solaris can be easily ported, but it's a fact
>that not much has. Even the oft-used example of Netscape, which is
>available on Solaris but not SCO/UnixWare, has apparently not been
>ported to Solaris X86.
>
The leverage comes from the fact that a number of Sun tools are ported to
Solaris x86, Visual ProWorks, a very nice development environment and DiskSuite
which provides raid and disk journaling support. Your information about
Netscape
on Solaris x86 is outdated, I am posting this message using netscape Solaris
x86 version.
>The SPARC insalled base of Solaris means *nothing* to the Intel world,
>except to explain how Sun can afford to lose money on such a complex
>operating system (X86) with such poor sales.
Your statement is true for certain circles, admittedly that is a good part of
the Intel market. But there does exist a large user base for Sparc platforms
who would like to augment their hardware platform with Intel machines, and for
this (rather large group of people) Solaris x86 has advantages over both SCO
and UnixWare.
Martin Sohnius
: Martin Sohnius writes
: > Considering that it took Sun about 6 years after SVR4 was ready to bring
: > Solaris to market, SCO's track record is decidedly better.
: "After SVR4 was ready" is a key phrase here, Martin. Solaris became workable
: with release 2.3, which became available in 1994. That's the same year as
: UnixWare 1.1, the first workable version of UnixWare.
: In both cases, the "late" to-market date was due to the fact that, while SVR4
: incorporated a lot of excellent conceptual developments (like pkgadd, HINT,
: HINT to SCO), it was a swarming bugfest as delivered by USL.
ICL shipped an SVR4 based OS way back in the 80s, as did others. UnixWare
was the first reasonable implementation of 4.2 on PCs which is quite
different from implementing SVR4.0 on proprietary hardware which is what
Solaris on SPARC is.
Martin Sohnius
: According to my AT&T documentation, SVR4 was available in source in 1989
: (though it might have been available even earlier); I was had usable SVR4
: for Intel (Esix 4.0.1) in late 1990.
I wish I had known that at the time. I worked for a company who did AT&T
authorised training, and as of about early 1990, we had genuine AT&T SVR4
training material. There was some market for it (ICL's DRS6000 was SVR4,
and sold well in the UK), but my employer certainly was not prepared to
dish out for a DRS6000, so I had no hardware to run SVR4 on. I kept asking
the folks at USL Europe about a working Intel port, but drew blanks...
(Chris P., if you read this...) Eventually, we used SCO Unix to do at
least the new (ANSI) C Programming course, because our 3B2 only had the
venerable K&R compiler (literally *the* K&R, I guess!). Otherwise we stuck
to SVR3 training, and the UNIX training branch went out of business in late
1991.
J.J.Farrell
>ev...@telly.telly.org (Evan Leibovitch) wrote:
>>
>>The SPARC insalled base of Solaris means *nothing* to the Intel world,
>>except to explain how Sun can afford to lose money on such a complex
>>operating system (X86) with such poor sales.
>
>Your statement is true for certain circles, admittedly that is a good part of
>the Intel market. But there does exist a large user base for Sparc platforms
>who would like to augment their hardware platform with Intel machines, and for
>this (rather large group of people) Solaris x86 has advantages over both SCO
>and UnixWare.
If I'm running UnixWare 2 on my ICL SPARC boxes, how does Solaris x86 have
advantages over UnixWare 2 on my x86 boxes?
My opinions; I do not speak for my employer.
Bryan Althaus
: In article <DGA9D...@byteware.com>, Jim Vlcek <vl...@byteware.com> wrote:
: >Martin Sohnius writes
: >> Considering that it took Sun about 6 years after SVR4 was ready to bring
: >> Solaris to market, SCO's track record is decidedly better.
:>"After SVR4 was ready" is a key phrase here, Martin. Solaris became
: Don't confuse SVR4.0 (on which Solaris is based) with SVR4.2 (on which
: UnixWare is based). While, by my estimations, Martin may be a slight bit
: off (six years rather than five), he is essentially right in that the
: SVR4 which Solaris used as a base was ready for use quite some time ago.
Sun had to add into SVR4.0, SMP and kernel threads support in addition to
to making the kernel dynamic ( change a file and reboot ) and rewrite
key daemons to be multi-thread aware, etc. etc. So when SVR4.2 was
out with all these is a fair comparison time wise.
: According to my AT&T documentation, SVR4 was available in source in 1989
: (though it might have been available even earlier); I was had usable SVR4
: for Intel (Esix 4.0.1) in late 1990.
:>In both cases, the "late" to-market date was due to the fact that, while
:>SVR4 incorporated a lot of excellent conceptual developments (like pkgadd,
:> HINT, HINT to SCO), it was a swarming bugfest as delivered by USL.
: Esix had no marketing, and its engineering staff was tiny, but it was
: certainly a credible (and very usable) product. Others such as
: Microport and UHC also had working implementations (Dell was a
: latecomer).
They did not support SMP or threads, etc etc. They were where Linux is
now.
: >I take Solaris x86 seriously, if for no other reason than that it has the
: >large installed base of Solaris/SPARC to leverage against.
: And how has that leverage been exploited? Not much. It's easy to argue
: that software for SPARC Solaris can be easily ported, but it's a fact
: that not much has. Even the oft-used example of Netscape, which is
: available on Solaris but not SCO/UnixWare, has apparently not been
: ported to Solaris X86.
You have to remember one thing, there was not Solaris 2.2 or Solaris 2.3 x86!
Solaris 2.3 was the first stable version so x86 users such as myself had
to wait for Solaris 2.4 x86 which only was released in November of 1994,
less than one year ago. Had there been a Solaris 2.2 and 2.3 x86 version
I would be thinking what is going on.
Instead of Netscape which is delayed for x86 only because the 1.1 version
was shelved because 2.0 is coming out. Both the browser and server will
be out for Solaris x86.
Lotus Notes is another example of a major product supporting solaris x86,
and this is with the poor marketing of Solaris x86.
: The SPARC insalled base of Solaris means *nothing* to the Intel world,
: except to explain how Sun can afford to lose money on such a complex
: operating system (X86) with such poor sales.
How much money has SunSoft lost on Solaris x86? What are the current
sales figures/shipments of Solaris x86 and UnixWare?
Remember, both Solaris Sparc, x86, and PowerPC version are all based on the
same source tree, so its not like in the case of UnixWare that all development
effort goes towards on platform, x86.
And this sharing of source code is the key, Solaris is just getting better
and has gone over the porting/bug hill.
SCO/UnixWare merge is just starting out.
Also, is there some application that is currently on SCO/UnixWare that would
keep one from purchasing Solaris?
Last I remember Solaris runs SCO binaries.
Martin Sohnius
: Neither should you confuse SVR4.0 with Solaris. Solaris in its current
: incarnation looks nothing like the code delivered in SVR4. The MP and
: Multithreaded support are an integral part of Solaris, so is dynamic
: kernel module loading.
My original point was: where is that earlier release of *any* System V
technology by Sun, after their participation in SVR4 caused the "Unix Wars"
in the first place?
Bryan Althaus
: Jim Vlcek (vl...@byteware.com) wrote:
: : Martin Sohnius writes
: : > Considering that it took Sun about 6 years after SVR4 was ready to bring
: : > Solaris to market, SCO's track record is decidedly better.
: : "After SVR4 was ready" is a key phrase here, Martin. Solaris became workable
: : with release 2.3, which became available in 1994. That's the same year as
: : UnixWare 1.1, the first workable version of UnixWare.
: : In both cases, the "late" to-market date was due to the fact that, while SVR4
: : incorporated a lot of excellent conceptual developments (like pkgadd, HINT,
: : HINT to SCO), it was a swarming bugfest as delivered by USL.
: ICL shipped an SVR4 based OS way back in the 80s, as did others. UnixWare
^^^^^^^^^^^^^^^
That would be somewhere between 1980 and 1989, and ICL's SVR4 knew
nothing of SMP or threads. Not about to downsise the coporate
mainframe onto that.
: was the first reasonable implementation of 4.2 on PCs which is quite
: different from implementing SVR4.0 on proprietary hardware which is what
: Solaris on SPARC is.
Tell me what the difference between SVR4.2 and Solaris is? I'm sure the
kernel gurus's at SunSoft are staying up nights trying to be on par
with SVR4.2, seeing as SUN sells servers with multiple CPU's as does
Cray. You do remember that 1.x of UnixWare was neither SMP
or thread aware, a reason to go with Solaris x86 at the time.
Last I heard Solaris ran on Sparc, Intel, and PowerPC unlike UnixWare
which has never been ported off of Intel. Talk about proprietary!
Hey you guys have your hands full with just the bug fixes :) Also Could
you fix the MHS to SMTP Gateway bugs? UnixWare would be a great use
to many if this feature worked.
When are we going to see CDE, NFS version 3, NIS+, POSIX threads (P1003.1c),
ONC+ on UnixWare as in Solaris (all platforms)?
Plain and simple, UnixWare is a great OS, but SunSoft can kick SCO's and
Novell's butt technology wise.
Show me something impressive that Novell or SCO has every given back to
the UNIX community?
Is it any wonder SunSoft came out with Java and HotJava? That everyone
has licensed NFS, ONC+? That NIS(+) is used by everyone?
Someone has to come up with the technology that everyone years later
adopts. Distributed Objects Everywhere (DOE) is another SunSoft
attempt at pushing the state of the art that others will later just
adopt.
You can still be a fan of SCO/UnixWare without being anti-SunSoft. Someone
has to hold the technology torch for UNIX and thats always going to be
SunSoft.....
Go with state of the art, go with Solaris or at least go with NT.
Bob Palowoda
: If Sunsoft is given the free rein to market itself as it should, it
: will inevitably be directly competing with SMCC. The move to market
: Solaris X86 as a product in its own right, not as merely a way to
: keep Sun shops "in the fold" at the low end, may be a Pandora's box
: Sun may regret.
Not only that but Sunsoft rumored doing a port to IBM's and
Motorola PPC. What's next the Alpha? Somebody has got to
to stop Sunsoft from writing new software.
---Bob
--
+--------------------------------------------------------+
| palo...@fiver.sns.com http://fiver.sns.com/~palowoda/ |
| Solaris x86 Corner http://fiver.sns.com/ |
+--------------------------------------------------------+
Bryan Althaus
: In article <45k4fq$5...@abyss.West.Sun.COM> Siamak Nazari <siamak> writes:
: >ev...@telly.telly.org (Evan Leibovitch) wrote:
: >>The SPARC insalled base of Solaris means *nothing* to the Intel world,
: >>except to explain how Sun can afford to lose money on such a complex
: >>operating system (X86) with such poor sales.
: >Your statement is true for certain circles, admittedly that is a good part of
: >the Intel market. But there does exist a large user base for Sparc platforms
: >who would like to augment their hardware platform with Intel machines, and for
: >this (rather large group of people) Solaris x86 has advantages over both SCO
: >and UnixWare.
: If I'm running UnixWare 2 on my ICL SPARC boxes, how does Solaris x86 have
: advantages over UnixWare 2 on my x86 boxes?
Solaris on Sparc supports thousands of software packages, with Solaris
shipping on 80% of Sparc based machines currently ordered.
UnixWare 2 on Sparc is very low volume and so there are few packages that
UnixWare 2 x86 would benefit from. In fact UnixWare 2.0 x86 shipments are
probably 99% of all UnixWare shipments which is where the software base
is, so if anything UnixWare Sparc would benefit from Unixware x86's
installed base. On Solaris, its the other way around. Sparc helps out
the intel line.
Guy Harris
>>By the way SunOS 4.1.x does have alot of System V in it, its just based
>>on a BSD kernel, so I'm not sure what you are going on about.
>
>Real (licensed) System V stuff, or cloned facsimiles?
Real (licensed) System V stuff.
The SV IPC mechanism code in SunOS 4.x came from SV.
The SV commands and library routines in 4.x came from SV.
The STREAMS framework, and the TLI/TPI code, in SunOS 4.x came from SV.
The tty subsystem in 4.x, and the "/dev/tcp" driver, are only
significant SVisms that'd count as "cloned facimilies" (although the
STREAMS tty subsystem was, as far as I can tell, the basis of the
STREAMS tty subsystem in SVR4 - the "/dev/tcp" driver is pretty much a
"glue layer" connecting the BSD sockets-based TCP implementation and the
TPI code).
(The Bourne shell, "make", and "/usr/ucb/mail" in *3*.0 came from SV,
BTW ("/usr/ucb/mail" being "mailx"-based). And yes, I mean 3.0, not
3.2....)
J.J.Farrell
>
>ICL shipped an SVR4 based OS way back in the 80s, as did others.
Unless my memory has slipped (which is quite possible), this isn't
correct. We shipped two or three field trial systems with very early
versions of SVR4 on SPARC in 1989, but our first real customer shipments
were in 1990; we started shipping SVR4 on Intel some time later (mid to
late '90, I think). SVR4 on SPARC was ready to ship at the same time as
SVR4 on 3B2, which (by definition) was the first version of SVR4.
Evan Leibovitch
>You do remember that 1.x of UnixWare was neither SMP
>or thread aware, a reason to go with Solaris x86 at the time.
And you have already stated that any releases of Solaris earlier than
2.4 were unusable -- so what good was all that stuff if it didn't work.
Look at now -- when Solaris is stable, and UnixWare has SMP.
>Last I heard Solaris ran on Sparc, Intel, and PowerPC unlike UnixWare
>which has never been ported off of Intel. Talk about proprietary!
This has clearly degenerated into a religious debate with little
relationship to facts. Actually, less than that, it's become an
issue of little more than debunking the lies of the missionary.
Bryan has repeatedly tried to spread falsehoods in an attempt to push
Solaris as everyone's salvation. Despite being told publicly that there
are non-Intel ports of Solaris, statements like the above continue
to spew.
First it was "many magazine reports" that HP would not give 64-bit code
to SCO, in an attempt to increase the level of FUD regarding the
SCO/Novell deal; that was exposed as an untruth. When pressed, Bryan
couldn't produce *one* source of such inaccuracy. Now this.
What next?
>Plain and simple, UnixWare is a great OS, but SunSoft can kick SCO's and
>Novell's butt technology wise.
Sounds like holy war talk. Just so much evangelism. Yawn.
Come back when you have a cogent point to make, Bryan.
>Show me something impressive that Novell or SCO has every given back to
>the UNIX community?
Working, stable, usable Intel Unix for a decade. Oh, sorry, that's what
SCO has given to the *user* community. That's what matters to most of us.
A good set of policies and active support and communication to the reseller
channel. Oh, that's what Novell gave to Unix VARs for the first time. That's
what matters to others of us (and all users, indirectly).
Sun can be ever-so-proud of its contributions. So can Xerox PARC, for that
matter. Big fat hairy deal, you can't take pride to the bank. If Sun didn't
do it, someone else would, like has been done for such trivial non-Sun
inventions as X WIndows, Motif and DCE.
Does any *user* of an operating system give a damn where the OS vendor
licences technology from, as long as it works well?
>Someone has to come up with the technology that everyone years later
>adopts.
Which is just fine, thank you. Let Sun's existing users act as guinea
pigs for its invention, and give it to the rest of us when it's working
right. In the not-infrequent cases where these innovations stiff in the
marketplace (such as SunView or Open Look), the rest of us can sit back
and ignore them from a distance, without having to make a significant
investment in skills which must later be tossed. Sounds good to me.
>Someone
>has to hold the technology torch for UNIX and thats always going to be
>SunSoft.....
How noble. What a shame that nobility holds so little weight in this
market. All the nobility in the world won't get you on the bus without
exact change.
>Go with state of the art, go with Solaris or at least go with NT.
Boy, what a sorry choice *that* is. Better Linux than either.
Evan Leibovitch
>Tell me what the difference between SVR4.2 and Solaris is? I'm sure the
>kernel gurus's at SunSoft are staying up nights trying to be on par
>with SVR4.2, seeing as SUN sells servers with multiple CPU's as does
>Cray.
As do Novell and SCO. Your point?
>You do remember that 1.x of UnixWare was neither SMP
>or thread aware, a reason to go with Solaris x86 at the time.
And at the time that 1.X UnixWare was released, Solaris X86 was
unusable. So you had a choice of UW1, which didn't do SMP, or Solaris,
which claimed to do it but was so badly broken it didn't matter.
Some choice.
>Last I heard Solaris ran on Sparc, Intel, and PowerPC unlike UnixWare
>which has never been ported off of Intel. Talk about proprietary!
Talk about inaccuracy!
Either this is a deliberate lie, or your incoming information is so
incomplete as to render analyses based on it worthless.
(And it ain't the first time, either...)
>Hey you guys have your hands full with just the bug fixes :)
.... spoken with a straight face, I'm sure, considering how long
(and how many releases) Sunsoft took just to get Solaris usable.
Even a brief scan of the Solaris newsgroup indicates it's not
without headaches, either.
Dave Parker
>J.J.Farrell (j...@dsbc.icl.co.uk) wrote:
>: In article <45k4fq$5...@abyss.West.Sun.COM> Siamak Nazari <siamak> writes:
>: >ev...@telly.telly.org (Evan Leibovitch) wrote:
>: >>
>: >>The SPARC insalled base of Solaris means *nothing* to the Intel world,
>: >>except to explain how Sun can afford to lose money on such a complex
>: >>operating system (X86) with such poor sales.
>: >
>: >Your statement is true for certain circles, admittedly that is a good part of
>: >the Intel market. But there does exist a large user base for Sparc platforms
>: >who would like to augment their hardware platform with Intel machines, and for
>: >this (rather large group of people) Solaris x86 has advantages over both SCO
>: >and UnixWare.
>
>: If I'm running UnixWare 2 on my ICL SPARC boxes, how does Solaris x86 have
>: advantages over UnixWare 2 on my x86 boxes?
>
>Solaris on Sparc supports thousands of software packages, with Solaris
>shipping on 80% of Sparc based machines currently ordered.
>
>UnixWare 2 on Sparc is very low volume and so there are few packages that
VERY low volume. Less than one. Not very many less than one, mind you,
but less than one.
Raymond Shwake
>Evan Leibovitch (ev...@telly.telly.org) wrote:
>: If Sunsoft is given the free rein to market itself as it should, it
>: will inevitably be directly competing with SMCC. The move to market
>: Solaris X86 as a product in its own right, not as merely a way to
>: keep Sun shops "in the fold" at the low end, may be a Pandora's box
>: Sun may regret.
>Not only that but Sunsoft rumored doing a port to IBM's and
>Motorola PPC. What's next the Alpha? Somebody has got to
>to stop Sunsoft from writing new software.
It's hardly a rumor. The October 1995 issue of SunSoft Communicator
announced the pending release of Solaris (PowerPC Edition). Excerpting...
"The Solaris (PowerPC Edition) will run on IBM's new Power Series
and other manufacturers' PowerPC microprocessor-based systems that
adhere to the PowerPC Reference Platform" ...
"An early access version of the Solaris (PowerPC Edition) will be
released in the Fall of 199t. Product will ship in March, 1996, ..."
Now tell us why this is a bad thing, and why this comprises "writing new
software"? It's an operating environment, after all, and if it adds credi-
bility to Solaris as an enterprise solution across platforms (as Microsoft
is trying to do with NT), so much the better.
Martin Sohnius
: Martin Sohnius (msoh...@novell.co.uk) wrote:
: : ICL shipped an SVR4 based OS way back in the 80s, as did others. UnixWare
: That would be somewhere between 1980 and 1989, and ICL's SVR4 knew
: nothing of SMP or threads. Not about to downsise the coporate
: mainframe onto that.
That was not my point! My point was that it took Sun ages after the wreckage
they caused to the Unix market actually adopt the system which caused all that
havoc! I am not suggesting you downsize your corporate mainframe to what either
ICL or Sun shipped in 1989.
: You do remember that 1.x of UnixWare was neither SMP
: or thread aware, a reason to go with Solaris x86 at the time.
Gee, Bryan, for all of two months! The first workable version of Solaris x86
shipped in November, as you are telling us, and UnixWare 2 shipped on Jan. 20
(give or take a few days). It had been out in beta since about September.
: Last I heard Solaris ran on Sparc, Intel, and PowerPC unlike UnixWare
: which has never been ported off of Intel. Talk about proprietary!
I don't call an OS that runs on 98% of the world's CPUs proprietary.
(I know, I know, this is a dangerous argument, by which one could call
Windows an "open operating system". :)
: Hey you guys have your hands full with just the bug fixes :)
No more so than anybody else. The main difference appears to be that we
try to communicate with our customers about both the bugs and their fixes.
That's always a dangerous and time consuming undertaking, particularly if
people who haven't bought or used a single product of ours for years,
keep engaging us in mudslinging exercises.
: Show me something impressive that Novell or SCO has every given back to
: the UNIX community?
Since your "kicking butt" includes the former Unix Systems Group within
Novell as its targets, may I mention: SVR4 and all that came with it? :-)
: Is it any wonder SunSoft came out with Java and HotJava? That everyone
: has licensed NFS, ONC+? That NIS(+) is used by everyone?
Not by everyone, actually. Like: anyone who even halfways cares about
security won't touch it.
: You can still be a fan of SCO/UnixWare without being anti-SunSoft.
But, apparently, not the other way around :-)
: Go with state of the art, go with Solaris or at least go with NT.
Oh, gee, again! I am afraid, Bryan, you'll have to work hard to
re-establish your street-cred (other than, of course, Wall-Street-cred)
after this one!
Martin Sohnius
: In article <bryDGF...@netcom.com> b...@netcom.com (Bryan Althaus) writes:
: >J.J.Farrell (j...@dsbc.icl.co.uk) wrote:
: >
: >: If I'm running UnixWare 2 on my ICL SPARC boxes, how does Solaris x86 have
: >: advantages over UnixWare 2 on my x86 boxes?
: >
: >Solaris on Sparc supports thousands of software packages, with Solaris
: >shipping on 80% of Sparc based machines currently ordered.
: >
: >UnixWare 2 on Sparc is very low volume and so there are few packages that
: VERY low volume. Less than one. Not very many less than one, mind you,
: but less than one.
Nevertheless, it *is* ported to the SPARC architecture, and even if the
shipping date is something like the end of this year, the port exists, as I
am sure J.J. is not allowed to confirm :-).
Tim Wake
: >You do remember that 1.x of UnixWare was neither SMP
: >or thread aware, a reason to go with Solaris x86 at the time.
A comment on the Solaris vs UW talk. We have a system that runs on Interactive
About 2 months or so ago we formed a team to look at moving to another OS. The
choices were Solaris x86 or UnixWare. Now the other groups in the company that
use unix are Sparc shops so Solaris had already been declared the corp standard.
So that meant UW would have to be much better. The UW has blown away Solaris.
The install, administration, porting of ISC software, and using equinox.
Everything we do on Solaris is much more of a pain. It requires much more
intervention and the installation of the Dev sys is a nightmare ( using the
license manager) Running Solaris in 16 meg, while we waited for the rest of
the RAM to show up, caused the system to lockup twice a day. With Novell
selling UW to SCO has made the UW decision much harder ( while really need to
justify) but so far may be the decision. In conclusion UW just seems more
user friendly and polished. Solaris x86 seems much more oriented towards
Unix people and you feel like you are the first people that have tried it.
Tim Wake
tlw...@amoco.com
Bryan Althaus
: : Hey you guys have your hands full with just the bug fixes :)
: No more so than anybody else. The main difference appears to be that we
: try to communicate with our customers about both the bugs and their fixes.
: That's always a dangerous and time consuming undertaking, particularly if
: people who haven't bought or used a single product of ours for years,
: keep engaging us in mudslinging exercises.
I am the person who ordered UnixWare 2.01 for this company even after
the word UnixWare became a catch-word used by the NT people around here
to show how hard it is to work with UNIX. With NT 3.5 released ( now at
3.51 SP2 ), the argument about OS stability came up with the NT people
once again using UnixWare as "the" example of what UNIX is.
We as you know had a hard time with UnixWare 1.x, unlucky in that while all
OS's have bugs, the ones that were major just happened to be one's that
affected us greatly. If luck was on our side, maybe we would have been only
hit by one major bug.
In any case, when UnixWare 2.0x came out our company had already given
up on UnixWare. We are a big Novell Netware site and at the time Novell
had a great Rep and even the non-UNIX people were OK with UnixWare.
Anyway, UnixWare 2.0x was ordered by me because at the price that was
being offered, it would be perfect to be used just for the MHS Gateway.
Who better than Novell to handle this, also we all wanted to see what
UnixWare 2.01 would be like.
While I have said the OS it great, impressive whatever, the MHS Gateway
still has problems to the point that we still have not moved over for
it. Who gets to hear this but me, because I'm the one who ordered it.
Its not the money, but the time spent.
The bugs thing was a semi-joke meant more at, why can't this GateWay
be fixed once and for all. Its been like this since 1.x.
:: You can still be a fan of SCO/UnixWare without being anti-SunSoft.
: But, apparently, not the other way around :-)
I have said before UnixWare is a great OS. And have alot of respect
for Novell. Dropping UnixWare and getting down to their core business
was a smart move. UnixWare never made much market penetration and
now is being merged into SCO.
But as I have said, UnixWare is to Novell what OS/2 is to IBM. Both
OS/2 and UnixWare are great OS's but had no where to go. I
admit I do not like SCO. SCO reminds me of the SVR3.2 AT&T vanilla
UNIX machine we had in here for a few years.
:: Go with state of the art, go with Solaris or at least go with NT.
: Oh, gee, again! I am afraid, Bryan, you'll have to work hard to
: re-establish your street-cred (other than, of course, Wall-Street-cred)
: after this one!
You have to remember that I no longer consider UnixWare a product,
but part of SCO. You know how I feel about SCO so in a year and a half
when the merge is complete, if SCO is like UnixWare is now, and with
SCO x86 maketing savy, then yeah I think it will be one great OS. And
that above statement will be false.
But that's a wait and see thing. Remember that most people who bought
UnixWare did so becasue is was so far ahead of SCO when it came to
state of the art.
As for this post, I will appologise for the bug knock as I haven't had
any crashes with UnixWare 2.01. Yet as we speak I am trying to find
out why NT 3.51 just crashed with a bugcheck after being up a month :)
Martin Sohnius
[...]
: While I have said the OS it great, impressive whatever, the MHS Gateway
: still has problems to the point that we still have not moved over for
: it. Who gets to hear this but me, because I'm the one who ordered it.
: Its not the money, but the time spent.
You really were unlucky there: the whole email system took a serious
knock in the 2.01 release (it's actually slightly better in 2.0, and
much better in 1.1). I am not talking about the MHS gateway, but the
actual core email service. As a consequence, I could well imagine that
work on further improvements to the MHS gateway was set back as well.
Finally, it does not help that Novell internally now uses GroupWise mail,
rather than MHS, so one major incentive to get it going is gone...