NetNews Usenet Archive 1993 #1/NN_1993_1.iso/spool/comp/unix/sysv386/17670

Internet Message Format | 1993-01-06 | 8.3 KB

Xref: sparky comp.unix.sysv386:17670 comp.unix.sys5.r4:1133
Newsgroups: comp.unix.sysv386,comp.unix.sys5.r4
Path: sparky!uunet!uunet.ca!canrem!telly!evan
From: [e-mail] (Evan Leibovitch)
Subject: It's Time for a Shakedown (was Re: A few answers to Dell Unix FAQ)
Organization: Somewhere just far enough out of Toronto
Date: Mon, 4 Jan 1993 01:45:53 GMT
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]> <[e-mail]>
Sender: [e-mail] (Evan Leibovitch)
Lines: 160

[e-mail] (bill davidsen) writes:
>[e-mail] (Shane Bouslough) writes:
>>Charlie Sauer ([e-mail]) wrote:

First things first. Thanks, Charlie, for doing the posting you did. Having
some factual information to work with, even if I'm not entirely happy with
what's said, is far better than any rumour or speculation. Especially in
light of how Dell people have been burned on the net before after being
up-front and honest, your approach is to be applauded and admired.

>>> Q.  Is Dell changing the source base for Dell Unix to V.4.2 or Solaris?

>>> Dell does not plan to change Dell Unix to a new source base, neither 4.2
>>> nor Solaris.

>> Gee, good thing I just plunked down $1300 for Dell UNIX... wouldn't
>> want to get stuck with a dead-end product, now would I?

>Did you think that you were never going to change again? You got the
>best product available today for what you want to do, and it is still
>being supported and enhanced.

This is *exactly* what SCO said when everyone else was going to 4.0 and
they were barely making their 3.2 ready for market. Come to think of it,
SCO is still saying this...

My problem with the policy stated by Charlie above -- as with SCO's -- is
that it's further splintering the R4 UNIX market at a time when the Intel
UNIX market least needs this extra confusion. It's not enough for people to
choose UNIX rather than OS/2 or NT or whatever, then they have the
non-trivial task of sorting through all the crap (not only marketing crap,
but the also the well-intentioned comparisons done here on the net and
elsewhere) to choose "whose UNIX"?

Not one company in the bunch has shown a true commitment to work with

- The most current USL release
*AND*
- As much third-party hardware and software as possible
*AND*
- Reasonable prices
*AND*
- Reasonable VAR/reseller support
*AND*
- Reasonable end-user support
*AND*
- Stability and certainty based on past performance and future direction.

Not one of the players so far supports more than two of the above
criteria. Each has their own little niche, while none has even bothered
to address the full job.

I am now absolutely convinced that the scattering of small, semi-committed
players in the Intel UNIX marketplace is causing far more harm then good.
The confusion, the endless re-invention of wheels, and the mindless petty
infighting amongst these small players more than makes up for the benefits
that the diversity of many competitors is supposed to offer.

Look at the big picture. The differences between Dell's R4 product and
Consensys' R4 product are absolutely nothing, compared to the difference
between either of them and MS-DOS, OS/2, Pick, Desqview, NT, or any other
PC operating environment.

The whole pile of R4 companies, together, have the marketing clout of a
speck of dust. Dell could have added some oomph, but it appears that just
the reverse is true, the company seems almost embarrassed about its UNIX
product (except, of course, for its UNIX people). Everex, too, put only a
half-hearted effort into ESIX, which explains why it hasn't performed to
expectations. Ditto for Kodak and Interactive.

UHC? Consensys? MST? Microport? Give me a break. The level of excellence
of their products notwithstanding, these companies are generally unknowns
in the UNIX marketplace, let alone the PC market in general.

In my opinion, these companies' continued existence in their current states
threatens the industry as a whole. Their unwillingness to work together
on such basic issues as bug fixes and third-party certification has ensured
that they have spent more time competing amongst themselves than with
the real competition. It has been only this petty, provincial bickering
that has prevented an Intel-equivalent to 88open, a concept which has
been desperately needed in this market, and without which the smaller
vendors have no hope for survival.

Blaming USL is no longer an excuse, the companies have had plenty of
time to collectively deal with the very real problems in USL's handling
of their UNIX. Individually they have no clout with USL. As a group they
could, but it's likely we'll never know how well this might have worked.

Intel, to its credit, has tried to achieve some level of consensus with
its ABI and BCS programmes. The ABI programme was sabotaged by SCO, and
the other vendors haven't have the backbone to put some teeth into it
anyway. (That would take co-operation, wouldn't it?) All the BCS programmes
have done is ensure a certain level of backwards (ie, SCO) compatability in
present and future USL releases. SCO has expressed no intention in making
its own systems forward-compatible with R4.2.

As many of you may remember, Intel even purchased Bell Technologies with
the intent of using its resources to produce a definitive R4. In
retrospect, I wish they had stuck to it a bit longer. They eventually sold
out to Interactive, which itself was sold to first Kodak and later Sun
(another company helping to fragment the market, but big enough not to
care).

Where is this leading?

In my opinion, this market will not realize its potential until there is
a massive shakedown. Almost all of the existing R4 vendors will have to
merge or fold. A single company doing all the business of the existing
R4 vendors *might* have the clout and installed base to provide the
resources necessary to being all of the above features I mentioned into
a single, stable company.

Whose?

I have high hopes for Univel, which has already taken some decent steps in
this direction. Part-ownership by USL would (I hope) mean that the product
would have a fast channel into USL's bug-fixing R&D. The other half-owner
(now, at least indirectly, the full owner), Novell, has experience in the
PC market and specifically taking on Microsoft at its own game. It also has
good reputations for end-user and reseller support.

Univel's startup has been rough, but then so has every other R4 vendor's.
Its product is pricey, but then high pricing and margins have done pretty
well to date for Novell, Apple and others. It's likely that a low-end leader
like Consensys or MST can make a go of the market by riding on the coattails
of a definitive R4, by making sure it's 100% compatible, and offering a low
enough cost to keep the high-priced spread honest.

Other low-cost options like Linux, if they're compatible and stable
enough, will also keep downwards pressure on the price of USL's stuff.

Then again, I have no great loyalty to Univel. If a different company can
come out of a shakedown as a leader in the industry, I'll work with it.
But Univel seems the only R4 vendor with the budget and resources behind
it enough to emerge unharmed.

I see the market as being able to realistically support only Solaris, SCO,
and no more than two R4 vendors (and even then only if the larger one is
overpriced). It will be interesting indeed to see the evolution of such
a market, for SCO to compete on a field of equals rather than as a
average-sized person in a land of dwarves.

Which R4 company will be *the* one? Who cares anymore? Let's just get on
with the issue, which is making UNIX better and more popular. The current
obstacle course of implementors working in self-imposed vacuums, only
detracts from this goal.

We've had our nice little experiment of all these nice little companies
selling UNIX. This hasn't worked. Time to let some big boys play.

--
 Evan Leibovitch, Sound Software Ltd., located in beautiful Brampton, Ontario
         [e-mail] / uunet!utzoo!telly!evan / [phone]
What's with all this multimedia stuff? Most vendors can't get *one* done right.


--
 Evan Leibovitch, Sound Software Ltd., located in beautiful Brampton, Ontario
         [e-mail] / uunet!utzoo!telly!evan / [phone]
What's with all this multimedia stuff? Most vendors can't get *one* done right.