|
2002 IBM Internal Email on Project Monterey - "No One Wants It" |
|
Thursday, August 25 2005 @ 12:09 PM EDT
|
This is a fascinating document, because it tells us about IBM's efforts regarding Project Monterey and how it died. It's an email, dated November 6, 2002, from Bill Bulko in IBM's Austin unit to another IBM guy in Austin.
It's Exhibit 4 of document #495-1, a collection of exhibits SCO calls Unsealed Exhibits to Memorandum in Support of SCO's Motion for Leave to File Third Amended Complaint Pursuant to Federal Rules of Civil Procedure (15(a) and 16(b) [PDF]. The Third Amended Complaint is Exhibit 1 in the collection. More on that document next. The email says that IBM decided that Project Monterey had no prospects, after six months of trying to market it. They did release it, but almost no one bought it, or as the email puts it, "we have been trying to distribute Monterey, but no
one wants it." The royalties they had to pay to SCO, as a result, are laughable. It's evidence that what they said in the email about no one wanting it is true. The email lists the internal position and what they planned to say externally. The "no one wants it" part is in the internal list. That matches what we found in old news clippings as well, that it was a product that died aborning, mainly because of issues with the Itanium chip, but also because the marketplace had started to move to Linux. One thought, in all fairness, that has to be pointed out is that the IBM external position may have ended up reflected in some of the news stories of the day. But the inescapable fact is that you can't make customers buy a product they don't want. IBM, the email informs us, put Project Monterey on the PRPQ list. What is that? According to this IBM Redbooks glossary for iSeries servers, it means Programming Request for Price Quotation (IBM's International Technical Support Center produces Redbooks --
how-to manuals on a wide range of technical subjects). As you will notice on this page, Communications Server for Linux v6.0.1.3 PRPQ and for Linux on zSeries v6.0.1.0 PRPQ are or were in that same category, for example, so it doesn't mean IBM doesn't want to sell something to you. Here is an explanation of the term in a comp.unix.aix FAQ: PRPQ - programming request for price quotation.
Essentialy [sic], a program that is not available off-the-shelf.
(e.g. HACMP started as a PRPQ, you could not order HACMP, but
you could ask IBM for a highly-available solution)
HACMP is IBM's High Availability Cluster Multi-Processing for AIX. If you go to Google or wherever you like to search and use keywords "HACMP IBM", you will see that PRPQ doesn't mean IBM is trying to kill a product. In this case, though, it does mean that they were aware that it had no future. There are more exhibits in this collection, all the worst emails SCO could find, and I'll write more on them and where this one fits into the timeline as we have them ready.
**********************************
Bill Bulko 11/06/2002 03:41 PM This document expires on
02/04/2003
To: Anthony Befi/Austin/IBM@IBMUS
cc:
From: Bill Bulko/Austin/IBM@IBMUS
Subject: Project Monterey update
Importance: Urgent
Tony: you asked me to collect some data on the current status of Project Monterey in order to update you before you meet with SCO(Caldera) again. I spoke with Ron Lauderdale, Sharon Dobbs, and Bill
Saulnier. Here is a capsule summary of our position with Project Monterey.
History:
* We made a deal with SCO (before it was bought by Caldera) that we would take the AIX and SCO
code and produce an IA-64 product.
* From IBM's side: we would jointly develop a 64-bit Intel version of Unix that would be the first
and best available on the market using AIX as a base, but with SCO code for portions of the
product.
* From SCO's side: since they were the 32-bit leader in the market, this would give them and their
customers an easy migration to 64-bit.
* The deal was to do joint development and then establish licensing back and forth between the two
companies. The license would be royalty-free everywhere else except in Monterey. Even though
SCO code is now embedded within AIX, we would only have to pay royalties to SCO when we
distributed Monterey. Likewise, SCO owes IBM royalties when it distributes the product.
* As you know, when we changed strategic direction, we made Monterey a PRPQ.
Status:
* We distributed 32 copies of the PRPQ in 2001, resulting in only $256 in royalties paid to SCO.
* We believe we owe SCO royalties in 2002 - but very little. We will be paying only $8 in royalties per
copy. Sharon Dobbs is trying to determine the actual number of copies distributed in 2002.
* The PRPQ is out of date and does not support the hardware currently being shipped by Intel. There
are no plans to update it. AIX 5L has moved considerably since development on this version was
stopped.
* The PRPQ is still available as of this date but is being withdrawn as we speak: the announcement
has already been made, and it will be effective sometime in December.
Implications:
* SCO(Caldera) is still entitled to royalties from any distribution or sales of Monterey. IBM will no
longer owe royalties once PRPQ is withdrawn. They may be checking or they may want to know
if there is any plan to resurrect the product.
* Our initial license to SCO code was contingent on our making an attempt to distribute an IA-64
product. Consequently, we need to be clear that we have been trying to distribute Monterey, but no
one wants it.
What our external position regarding Monterey should be:
* We have been distributing Monterey since May 2001, but there have been very few takers.
* We have no plans to make AIX available on the Itanium platform.
* The slow adoption of Itanium in the market place and the quick maturity of Linux over the past few
years makes Linux better suited for the Itanium market. Many of the qualities of service that
customers have come to expect from AIX will be made available on Linux.
* We are planning to EOL Monterey by the end of this year.
|
|
Authored by: Anonymous on Thursday, August 25 2005 @ 12:19 PM EDT |
But the inescapable fact is that you can't make customers buy a
product they don't want.
... unless you're a monopolist and
bundle the products that people don't want with the monopoly product. [ Reply to This | # ]
|
|
Authored by: Anonymous on Thursday, August 25 2005 @ 12:26 PM EDT |
"Even though SCO code is now embedded within AIX,"
What does IBM mean here exactly? Is this where sco is getting its basis for a
lawsuit? Can someone clear this up for me?[ Reply to This | # ]
|
|
Authored by: rsteinmetz70112 on Thursday, August 25 2005 @ 12:27 PM EDT |
That pretty much tells the story. Intel had everyone lined up for IA64 and then
failed to deliver on their promise. Everyone who believed Intel lost a lot of
money, including Intel.
---
Rsteinmetz - IANAL therefore my opinions are illegal.
"I could be wrong now, but I don't think so."
Randy Newman - The Title Theme from Monk
[ Reply to This | # ]
|
|
Authored by: CustomDesigned on Thursday, August 25 2005 @ 12:53 PM EDT |
My company looked at Monterey. Our conclusion was that 32-bits was more than
enough for our small-business servers for the forseeable future. We drooled
over the coolness of it, but couldn't cost justify it for any of the
applications we sell and maintain.
We still have no application for 64-bits.
I can imagine applications: scientific modeling, film rendering. But even a
large mainframe type business server will want a 64-bit physical bus and lots of
memory - but run lots of 32-bit virtual machines for the applications.
AMD
is changing the equation, however. A 64-bit AMD system is low cost, and still
runs 32-bit applications. It lets you play with 64-bits and find uses for it.
I see AMD 64 in our immediate future. [ Reply to This | # ]
|
|
Authored by: overshoot on Thursday, August 25 2005 @ 12:58 PM EDT |
You all know the drill: clicky, instructions at bottom, preview ... [ Reply to This | # ]
|
- King Tut's Curse :) - Authored by: Anonymous on Thursday, August 25 2005 @ 01:24 PM EDT
- V interesting - Authored by: Anonymous on Thursday, August 25 2005 @ 01:33 PM EDT
- Ruling on the DMCA and S 1179c) from the First Circuit - Authored by: Anonymous on Thursday, August 25 2005 @ 01:43 PM EDT
- Linux servers sales up 31.6% in Q2. - Authored by: Anonymous on Thursday, August 25 2005 @ 03:21 PM EDT
- SCOX - Authored by: tiger99 on Thursday, August 25 2005 @ 06:51 PM EDT
- "Windows XP configuration vulnerable to Zotob" - Authored by: Anonymous on Thursday, August 25 2005 @ 08:27 PM EDT
- Did M$ receive their answer from OSDL at Linux World? - Authored by: Anonymous on Thursday, August 25 2005 @ 08:50 PM EDT
- Firefox Builds at 1.6a1 - Authored by: Anonymous on Thursday, August 25 2005 @ 11:38 PM EDT
- OT Here, please - Authored by: Anonymous on Friday, August 26 2005 @ 08:08 AM EDT
- The network structure of US Supreme Court cases - Authored by: Anonymous on Friday, August 26 2005 @ 11:28 AM EDT
- Open source Skype clone - Authored by: vadim on Saturday, August 27 2005 @ 07:14 AM EDT
|
Authored by: Anonymous on Thursday, August 25 2005 @ 01:07 PM EDT |
A PRPQ is a request for a custom software program product. Doesn't really mean
it's a product that is going to die, in fact it could mean that it's a trial
ballon to see what type of interest there is for the product, or it could be
that there is a very limited market for the software, and instead of ramping up
a huge support/marketing base for the product, it's targeted for a more
individual sale.
Basically it's not quite "off the shelf" software.
If you think in terms of the automobile industry, it would be kind of like
buying a US car with some non-standard feature (like a hand throttle, and hand
brakes). Or the type of cars that driving schools use (with extra controls over
in the passenger side)[ Reply to This | # ]
|
- PRPQ's - Authored by: dkpatrick on Thursday, August 25 2005 @ 01:26 PM EDT
- PRPQ's - Authored by: Anonymous on Thursday, August 25 2005 @ 06:10 PM EDT
|
Authored by: lm on Thursday, August 25 2005 @ 01:12 PM EDT |
IBM managed to sell 32 copies of Monterey in 2001. Considering the number of
Itaniums that shipped during that period, that's actually pretty impressive. At
minimum, that's probably approaching 1 out of every 100.
It will be interesting if documentation surfaces on how many copies of Monterey
shipped in 2002. On the surface, though, it looks like Intel sank Monterey much
more than IBM. If shipments of Itanium had been what Intel had wanted, Monterey
may have been a good investment. But when shipments of your processor is what
Itanium's was in 01 and 02, there isn't a whole lot of room for a new contender
in the 64 bit space.[ Reply to This | # ]
|
|
Authored by: OmniGeek on Thursday, August 25 2005 @ 01:48 PM EDT |
I keep seeing these exhibits being unsealed, and every one lurches across the
page, hoarsely groaning that SCO has no "BRAINS... BRAINS...
BRAINS...". And, of course, they just keep coming, and they haunt SCO
mercilessly. Clearly, they are The Undead Exhibits.
So I suppose it shouldn't be surprising that I read "Unsealed
Exhibits" as Undead Exhibits" on the headline today.
Popcorn, anyone?
---
My strength is as the strength of ten men, for I am wired to the eyeballs on
espresso.[ Reply to This | # ]
|
|
Authored by: jig on Thursday, August 25 2005 @ 01:59 PM EDT |
i think this email has been referenced because IBM's guy makes no reference to
developing monterey for anything but ia-64.
or maybe i'm getting confused... wasn't there some contention between ibm and
sco on contract claims that involved developing for the power line or something?
was that for aix, or was that for monterey? i thought i had seen some press
releases that showed that ibm and sco expected monterey to extend to the 64 bit
mac line...[ Reply to This | # ]
|
|
Authored by: Anonymous on Thursday, August 25 2005 @ 05:55 PM EDT |
It doesn't seem that fair to me.
2) If this email is correct SCO signed a contract on the
assumption that IA64 was the future. To get the contract
across the line they let everthing else go. They were
wrong.
Intel kept going as they had a solid position in the
32bit place and contracts in place that allowed them to
pick up AMD's solution. For IBM it was R&D, R&D doesn't
always work out.
SCO's 32bit space was taken by linux, they had nothing
left.
You would have to say whoever signed the contract with IBM
sealed SCO's fate, but they were not to know IA64 was what
it was.
Crazy Engineer [ Reply to This | # ]
|
- 1) Why is SCO unsealing IBM's stuff. - Authored by: nofundsleft on Thursday, August 25 2005 @ 06:54 PM EDT
- 1) Why is SCO unsealing IBM's stuff. - Authored by: dbc on Thursday, August 25 2005 @ 07:34 PM EDT
- 1) Why is SCO unsealing IBM's stuff. - Authored by: Anonymous on Thursday, August 25 2005 @ 07:55 PM EDT
- 1) Why is SCO unsealing IBM's stuff. - Authored by: Anonymous on Thursday, August 25 2005 @ 08:12 PM EDT
- 1) Why is SCO unsealing IBM's stuff. - Authored by: Anonymous on Friday, August 26 2005 @ 04:18 AM EDT
- Meant to add - Authored by: Anonymous on Friday, August 26 2005 @ 04:31 AM EDT
- What did IBM have against Caldera? n/t - Authored by: Anonymous on Friday, August 26 2005 @ 07:52 AM EDT
- Nothing - Authored by: Anonymous on Friday, August 26 2005 @ 08:22 AM EDT
- Don't agree - Authored by: Anonymous on Friday, August 26 2005 @ 08:50 AM EDT
- Thank you - Authored by: Anonymous on Friday, August 26 2005 @ 09:15 AM EDT
- Don't agree - Authored by: Anonymous on Friday, August 26 2005 @ 10:22 AM EDT
- Don't agree - Authored by: Anonymous on Friday, August 26 2005 @ 12:57 PM EDT
- Very helpful - Authored by: Anonymous on Friday, August 26 2005 @ 01:04 PM EDT
- Very helpful - Authored by: Anonymous on Friday, August 26 2005 @ 01:24 PM EDT
- Very helpful - Authored by: Anonymous on Friday, August 26 2005 @ 01:32 PM EDT
- Very helpful - Authored by: PJ on Saturday, August 27 2005 @ 11:47 PM EDT
- Very helpful - Authored by: PJ on Saturday, August 27 2005 @ 11:49 PM EDT
- Very helpful - Authored by: tbogart on Saturday, August 27 2005 @ 02:48 PM EDT
- Don't agree - squared - Authored by: tbogart on Saturday, August 27 2005 @ 02:18 PM EDT
- Also - Authored by: Anonymous on Friday, August 26 2005 @ 06:18 AM EDT
- How is it IBM's fault? - Authored by: Anonymous on Thursday, August 25 2005 @ 08:17 PM EDT
- 1) Why is SCO unsealing IBM's stuff. - Authored by: jdg on Friday, August 26 2005 @ 12:20 AM EDT
- Nothing left? - Authored by: Anonymous on Friday, August 26 2005 @ 11:42 AM EDT
|
Authored by: Anonymous on Thursday, August 25 2005 @ 07:20 PM EDT |
Seems PJ was at LinuxWorld, and these picture are proof positive of the (red)
caped crusader!
PJ Picture [ Reply to This | # ]
|
|
Authored by: AntiFUD on Thursday, August 25 2005 @ 08:08 PM EDT |
whilst I have only been using the Netcraft Anti-Phishing Toolbar for Firefox for
a couple of weeks, I can't help but express my congratulations to PJ, for
providing such an extraordinary site.
When I first installed the Netcraft Toolbar, Groklaw's 'Rank:' was 984 and today
it has dropped (risen) to 960. Now, while I appreciate that Netcraft's 'Rank:'
is defined in their glossary as: "The popularity of a particular web site
amongst users of the Netcraft Toolbar.", this still very impressive since
we all know that only the 'wise' have moved to Firefox (even if some of us use
it on non-FOSS driven machines).
How many web-sites does Netcraft monitor?
---
IANAL - But IAAAMotFSF(not related to Daniel Wallace) - Free to Fight FUD
[ Reply to This | # ]
|
|
Authored by: Anonymous on Thursday, August 25 2005 @ 11:07 PM EDT |
hasn't worked in marketing :-).
[ Reply to This | # ]
|
|
Authored by: dwandre on Friday, August 26 2005 @ 02:07 PM EDT |
RPQs (request for price quotation) for hardware (and microcode) are a way for a
customer to order a custom feature and pay for it. The pricing is based on
estimated cost and committed number of customers. It's a way to deliver a
feature that some customers want (usually want very much) but for which most
customers don't care and won't pay. Programming RPQs are the same thing for
software. Some products start out as RPQs or PRPQs and the demand develops and
the planners get a better feel for the market so the feature or product become
an full-blown product. Offering Monterey as a PRPQ means that market estimate
(in terms of units sold) was so low that it didn't justify the cost of releasing
it as a product. PRPQs don't have to have full internationalization support,
for example (at least I don't think so), and the documentation requirements are
much less stringent. Defect support is usually very expensive since the
developers do the defect support. [ Reply to This | # ]
|
|
|
|
|