#ACAD for Xenix
21 messages in this thread
Am I the only one in the free world using ACAD for XENIX??? There seems to
be a distinct lack of information about the future for this platform. Does
anyone have any clues about R11 for Xenix? We are actually using SCO Unix
and not having any significant problems, other than not being able to work
with R11 drawings.
Any information would be greatly appreciated!!
Bill,
Autodesk does not plan to produce any more releases of AutoCAD for SCO
Xenix. The good news is that AutoCAD Release 11 for SCO Unix has just been
released by our QA department for shipping. We also plan to support an SCO
Unix version of AutoCAD Release 12.
Dieter
Dieter – I wouldn't be so sure about SCO support for R12.
Huh? So, if not, will there be any support for PC based Unix in the future?
Solaris, perhaps, or is the market just not there for any variety?
I don't know, Peter. It does appear, however, that we haven't sold enough
copies for XENIX to make any profit on developing for it. Similar to the
OS/2 situation. We're always open to reconsideration… I'm not sure we've
made any final decision about R12 for SCO UNIX. As for Solaris, I know
we'll keep it in mind, but I haven't heard any commitments to support it on
386/486 machines yet. These are questions for Marketing, and I'm not in
Marketing.
You folks have had a marketing plan for the XENIX version of AutoCad
designed from the ground up to ensure failure.
If you SCO UNIX marketing plan is the same, you've got no chance to make
any money on it, and your “poor sales'' complaint will be a self
fulfilling prophesy.
Among other things, you brought out the XENIX version just when SCO UNIX
became a better platform. You are bringing out a UNIX version just when X
Windows has gotten enough support, performance, and device independence to
become a viable platform. In short, you are always developing for the
worst available version of the SCO operating system.
I'm not an Autocad dealer, but I support MANY UNIX customers who have
Autocad installed somewhere in the organization or who have it networked
using UNIX systems. The complaints are always the same. They want (NEED)
Autocad in a multi-user as well as a multi-tasking environment, and you
folks continually seem bound and determined to give clients what you THINK
they need, not what they KNOW they need.
Sorry for the flame here, but in this case I think the Powers That Be in
your company should read this.
Tom Podnar
Tom – I have to beg to differ with one of your points: Xwindows performance.
It sucks, it's a dog with fleas, and unless it has the power of a state of the
art RISC cpu behind it, it's not an option as far as I'm concerned.
-TonyT.
X Windows on a VGA card certainly fits your description. However, that is
1988 technology. X Windows on an X terminal or X Windows on an intelligent
graphics card provides a potential for higher performance than writing
directly to the hardware. Rejecting it out of hand would be tantanmount to
ignoring the realities of the marketplace, which seems to be what Autodesk
has done.
Clients are willing to suffer a performance penalty to gain the benefits of
standardization and hardware independence.
Tom Podnar
Tom – My experience with clients shows your observations to not be the case.
Clients are not willing to suffer a performance penalty for standardization,
because they can standardize without the performance penalty to begin with.,
Your comments are clearly beginning to reflect your own marketing agenda for
the extremely small market niche that you obviously have an interest in. I
personally see your postion on this matter, as conflicting with the interests
of all Autodesk customers collectively, because you obviously are not very
sensative to the economic realities surrounding the issue as a whole.
To suggest that Autodesk is ignoring the realities of the marketplace, is a
demonstration of the fact that you are conveniently ignoring the economic
realities of being in business in this day and age, where resources are very
limited.
Autodesk customers need the resouces of Autodesk concentrated more on making
improvements to the core, rather than porting to some relatively obscure and
commercially impotent platform. We need improvements to AutoCAD's core, not
more AutoCAD's without those needed improvements, running on more platforms.
I think you're point of view is entirely self-serving, and characterizes the
very same ignorance that you are suggesting of others.
-TonyT.
tonywhile the unix-on-the-desktop marketplace may not presently be very
large, and Autodesk's comments here seem to indicate that this is the case,
one could <hope> that they would not completely abandon us. The
out-of-the-mainstream areas just might be an important (not impotent)
platform for which to develop applications. Remember…it wasn't TOO many
years ago that nobody believed that CAD on the desktop would ever be
commercially viable. The built-in multitasking, multiuser environment
provided by Unix would seem to be natural for a CAD package. Let's face it,
we're <all> ties to a very narrow niche product as compared to, say word
processing or database management and the more platforms, especially
engineering-type platforms, that can be supported, the better for everyone.
Bill – I certainly hope that you and the rest of the Intel-based *NIX people
aren't abandoned. My comments were more with regards to an X-windows port of
AutoCAD, and X-performance on Intel systems in general. Personally, I think
Autodesk has much more important goals right now, and there are already many
(well, some) X-windows-based systems that AutoCAD runs on, for those with the
desire and money to go that route.
Keep in mind that AutoCAD is roughly ten years old, and its about time for a
RADICAL change in the product. The next-generation of AutoCAD will require
massive resources, and continuing to expend them on ports of a product that is
clearly not going to cut it in the ninties without some drastic changes to its
basic structure, can't be good for everyone.
-TonyT.
Well, Tony, as I mentioned, I don't sell Autocad and have no desire to
become an Autocad dealer. I just reflect the thoughts of my clients, and
they want multi-user CAD, cause they're smart enough to know that a 486/33
or 50 has lots of cycles left to spare when only one user is on it.
My perspective of Autocad for XENIX is that it was put into the field with
VERY little third party device support as compared to DOS. This is probably
the reason sales are so poor, as UNIX is a far more robust environment than
DOS.
The X Window System provides a common interface that can eliminate much of
the requirement for third party support. Now that X windows R4 is
available, most of the limitations of R3 (including performance) are
history. So I'm merely suggesting they take another look using 1991
realities and not 1989 realities as the basis of their decisions.
By the way, if they were to say they've done that and I'm all wet and that
Frame and Corel and Island an Arris and all the other people that support X
are fools, that's their prerogative. Tom
Tome – R10/DECwindows, I believe, was our first port to an Xwindoes
platform, followed by R11 for Sun's OpenWindows. We're not ignoring
Xwindows. As I understand it, however, SCO's Open Desktop doesn't have
backing store support (or at least it didn't the last time I heard), and
lack of that was a major stumbling block, since it would force AutoCAD to
perform a vector redraw (rather than a fast bitmap restore) if a portion of
its graphics window is obscured by another window and then brought to the
top again. Performance in such an environment would not be good, and our
customers crave performance. But even performance isn't enough; when the
SCO XENIX version of R10 first shipped, it offered far and away the best
performance of any platform, but relatively few licenses for it were sold.
I don't see how Autodesk can be "ignoring the realities of the
marketplace". We're participants in the marketplace and have received
clear messages from it. We've also got a business to run, and shareholders
who expect to see a good return on their investment.
Duff, Again, my suggestion was merely to ask SCO what it is currently
shipping in X Windows. It bears very little relation to what it was
shipping the last time you looked. I think you'll be (very pleasantly)
surprised. SCO has made a tremendous effor to upgrade server performance
and features, and servers with 50,000 and 60,000 Xstone performance are a
reality, as compared to 3,000 Xstones for their first VGA servers and
10,000 Xstones as a Sun3/50 benchmark standard. Backing store has been
around for a long time.
Tom
Regarding previous message regarding SCO shipping date: I have not
double-checked this – so cautious anticipation please.
Tom: It takes between six to 12 months or morew to port AutoCAD to a new
platform. When operating systems change as fast as SCO is changing, it is
difficult to stay in synch. Developers only get new OS releases,
sometimes, only a few days before dealers or customers. The only
consolation is that the same thing happens on other all other platforms.
Except maybe DOS. But Microsoft have upgraded DOS only about once every
two years – and then with, really, only minor modifications.
Regarding multi-user (ass opposed to mult-tasking). One major reason that
AutoCAD took off and blew away all the minicomputer systems was because it
was NOT multi-user. AutoCAD users do not have to share a CPU with other
users. Therefore they have predictable response times and a whole host of
other benefits.
Dinners on the table, got to go…
Elsewhere on this thread is a message I wrote to Bill Sako. It was meant
for Tom Podnar and his flame on Autodesk's SCO Unix efforts.
To continue the reply:
Tom, you say you support customers. So to a certain extent you are
selling, I am assuming. I am a salesman. My recent sales experience is in
Asia. I am used to responding to customer objections that go like this:
"When you buy a TV, the TV programs are free. When you buy a computer,
isn't it the same thing?" I am still getting used to your North American
ways.
Can you tell me about the CAD programs that are multi-user and
multi-tasking on SCO Unix that you *DO* like to support? What forums
should I be visting in order to find out about these applications?
I do not personally run SCO Unix – but have met many of their people –
including their "John Walker" <Autodesk's is one of the 13 founders>. I
admit to being impressed and potentially desirous of running this OS. Here
is a thought: What If SCO concentrated on getting out great debuggers and
compilers? What if an SCO compiler could read DEC or SUN files? What if
porting an application from an extended DOS compiler to SCO was a piece of
cake?
Unfortunately, this is not so. Look at Duff's message. I'm sure he's only
talking about the tip of the iceberg.
Autodesk is as attracted to SCO Unix as you are. This is obvious because
they spent the millions needed to support a platform. <I am not privy to
to the amount Autodesk spends on a platform port – but looking at the
people's rolling eyes – it's not cheap.>
Therefore, we've got to live with the fact that SCO UNIX releases will NOT
be timely. Because it's difficult and expensive.
Hm, do you have any easier suggestions?
I don't want to repeat myself, so I'd like to terminate the thread.
An answer of “perhaps we should go look at ODT again, but I'll make no
promises'', would have been served to prove that Autodesk cares about the
opoff
inions of those expressed on this Forum. It certainly would have been
enough fo me.
Tom
Bill,
I switched to Xenix shortly after Autodesk released a version ported for it
and I've become known since as the resident Xenix "evangelist" but that has
not had any influence at all on Autodesk' decisions. I've jumped on each
new release since version 2.5 just as soon as each one was announced. Maybe
if more Bills and Kevins and Michaels and Gales spoke up more often, the bean
counters at Autodesk would realize there are some loyal followers even if we
may following the beat of a different drum. No <g>'s here!!
Gale
There seems to be a feeling here of abandonment of the very *leading-edge*
types that Autodesk needs. Those of us who are willing to take a chance on
the highest technology available, whether it be hardware or software.
Granted, the Unix platform may -never- be mainstream, but maybe it is the
10 or 5 or 1 percent absolutely essential for future development. Like the
yeast in bread or something??? If this train of thought at Autodesk is open
for reconsideration or not yet decided, I hope someone will relay the
importance for some of us of continued support for ths platform!!!
Bill,
Thanks for YOUR support!! Always glad to meet another 'convert'.
Gale
If I understand a voice mail correctly, R11 SCO UNIX is shipping December
2, 1991! Ladies and Gentlemen, you can place your orders today…