OT: (slightly) DOS and OS/2 code names?
Bob Eager
versions of OS/2, and indeed DOS if there were any.
I know:
OS/2 Warp 4 - Merlin
Warp Server for e-Business - Aurora
OS/2 v3 - Warp
Anyone know any more?
--
Bob Eager
http://www.tavi.co.uk/ps2pages
http://change.to/ps2
Jim Shorney
Bob Eager wrote:
>
> I'm trying to make a list of the internal code names for the various
> versions of OS/2, and indeed DOS if there were any.
>
> I know:
>
> OS/2 Warp 4 - Merlin
> Warp Server for e-Business - Aurora
> OS/2 v3 - Warp
I once heard OS/2 2.1 for Windows ('Red Spine') referred to as
'Ferengi'.
David L. Beem
> I'm trying to make a list of the internal code names for the various
You know the U.S. government will deny "Aurora" exists.
> Anyone know any more?
Undoubtably Tony will.
David
Da...@gilanet.com
David L. Beem
> I once heard OS/2 2.1 for Windows ('Red Spine')
> referred to as 'Ferengi'.
David
Da...@gilanet.com
Bob Eager
wrote:
> I once heard OS/2 2.1 for Windows ('Red Spine') referred to as
> 'Ferengi'.
Thanks - that was one I wasn't sure of.
Bob Eager
> You know the U.S. government will deny "Aurora" exists.
And send Louis after me?
Michal Necasek
>
> Anyone know any more?
>
Yup. See http://pages.prodigy.net/michaln/history/timeline.html
Some other development/code names of OS/2 1.0 include London,
DOS 4, DOS 5, Advanced DOS, 286-DOS, BigDOS. There were far
too many.
Presentation Manager was codenamed Winthorn.
You will see CP/DOS and Winthorn referred to a lot if you look
at OS/2 DDK sources.
Michal
David L. Beem
> > I once heard OS/2 2.1 for Windows ('Red Spine')
> > referred to as 'Ferengi'.
economics and the technology of its Windows support. Code-
named Ferengi when it was under development at IBM's
Personal Software Products Division in Boca Raton, Florida...".
Harvard.edu: "The name of the product, OS/2 Warp, is derived
from the codename used internally by the IBM programmers to
designate the project. The codename "Warp" was chosen for two
reasons: (1) most IBMers are Trekkies (all OS/2 codenames have
been Star Trek related. Version 2.1 was codenamed "The Borg,"
while the OS/2 for Windows product was codenamed "Ferengi")..."
University of Florida (Ufl.edu): "IBM will release two versions of
OS/2, code-named Warp, next month [page dated 09/09/94]. An
OS/2 2.1 version and an OS/2 for Windows version. It's also
preparing another upgrade, code-named Klingon. I wonder if Lt.
[sic] Warf can sue for trademark infringement? "
David
Da...@gilanet.com
Bob Eager
wrote:
> Yup. See http://pages.prodigy.net/michaln/history/timeline.html
Great! Thanks!
I don't need as much detail as that - just a one liner. I'll extract a
bit...
Tony Ingenoso
1.1 - Trimaran
1.2 - Sloop
1.3 - Cutter
2.0 - Cruiser
2.1 (full) - Borg
2.1 f/Windows - Ferenghi
PC DOS 7 - Baseball (but nobody ever called it that, it was just "7")
"Bob Eager" <rd...@spamcop.net> wrote in message news:176uZD2KcidF-p...@rikki.tavi.co.uk...
Unal Z
176uZD2KcidF-p...@rikki.tavi.co.uk...
> I'm trying to make a list of the internal code names for the various
> versions of OS/2, and indeed DOS if there were any.
I am quoting from "OS/2 Programmer's Guide" by Ed Iacobucci,
(1988, McGraw-Hill) a leader of the IBM OS/2 team (OS/2 1.0):
"Industry rumors abounded, at different time referring to the system
as NEWDOS, ADOS, DOS 5.0, DOS/286 or CP/DOS".
"Microsoft Corp. also markets versions of OS/2 called MS OS/2".
And this one is the opening line of the foreword of the same book:
"I believe OS/2 is destined to be the most important operating
system, and possibly program, of all times."
Bill Gates, November 1987.
--
Unal
Tony Ingenoso
The plethora of externally available codenames for 1.0 was largely an attempt to find out who was leaking information. If you've
got a dozen ISV sites you're sending a new product out to, you send docs out with a dozen diferent codenames -- then when one of
them appears in the press with some product info, you know exactly who the leaker was ;->
"Unal Z" <un...@nonesuch.invalid> wrote in message news:3c952a3d$0$31266$6e36...@newsreader02.highway.telekom.at...
>
> ...CP/DOS
Bob Eager
<aingeno...@prodigy.net> wrote:
> 1.0 - had quite a few, London was one I remember
> 1.1 - Trimaran
> 1.2 - Sloop
> 1.3 - Cutter
> 2.0 - Cruiser
> 2.1 (full) - Borg
> 2.1 f/Windows - Ferenghi
>
> PC DOS 7 - Baseball (but nobody ever called it that, it was just "7")
Thanks, Tony - I was sure you'd have some input!
Michal Necasek
> Hazy memories are returning...this was the most commonly used internal codename within IBM and MS for 1.0.
>
chosen relatively late in the development cycle (to match PS/2).
> The plethora of externally available codenames for 1.0 was largely an attempt to find out who was leaking information. If you've
> got a dozen ISV sites you're sending a new product out to, you send docs out with a dozen diferent codenames -- then when one of
> them appears in the press with some product info, you know exactly who the leaker was ;->
>
Now wait... since all those names are known, does that mean everyone
was leaking<g>?
Michal
Tony Ingenoso
docs, but in the hallways it was always CP/DOS.
I suppose the OS/2 mane was being kicked around marketing circles long before that name made its way down to us lab rats ;->
Internally, few people ever refered to OS/2 1.0 as "OS/2" - it was usually just called "1.0" after it shipped. Even the 1.1
"Trimaran" release wasn't called Trimaran much - in the hallways it typically went by the PM codename "Winthorn".
1.2, 1.3, 2.0, and Warp were frequently referred to internally by their codenames.
FWIW historically, 1.3 was a defacto "renegade" release. Early on and most of the way through development is was a very small
development team operating mostly under the radar. Its proponents (I was a vocal one) believed 2.0 was going to schedule slip (it
did) and not meet all the claimed performance it was supposed to have (it didn't) and we would need an interim solution because 1.2
had a number of problems that needed fixing (no Adobe fonts, printing sucked, it was a pig, etc)
"Michal Necasek" <mic...@prodigy.net> wrote in message news:3C95435F...@prodigy.net...
Michal Necasek
> FWIW historically, 1.3 was a defacto "renegade" release. Early on and most of the way through development is was a very small
> development team operating mostly under the radar. Its proponents (I was a vocal one) believed 2.0 was going to schedule slip (it
> did)
>
When was 2.0 originally supposed to be delievered? I know I read
articles from 1990 that talked about 2.0 in detail... and supposedly the
first MS 2.0 betas appeared as early as 1989 (ie. before 1.3).
> and not meet all the claimed performance it was supposed to have (it didn't) and we would need an interim solution because 1.2
> had a number of problems that needed fixing (no Adobe fonts, printing sucked, it was a pig, etc)
>
And you guys did a damn fine job. Some people claim that OS/2 1.3 was
the leanest and meanest PC OS ever (DOS doesn't count as OS<g>). I
believe one of the other big 1.3 fans was Marc L Cohen...
Funny thing is that although Microsoft supposedly didn't do any work
on 1.3, they were still selling it as part of LAN Manager... because NT
release kept slipping and slipping.
Michal
Gfretwell
exactly what it was.
Tony Ingenoso
2.0 could have shipped 9 months to a year earlier if it went out with the slightly modified 1.3 shell it sported for so long during
development - IOW, plenty of time to beat Windows 3.1 to market. This was the path I and many others within IBM advocated - as well
as Microsoft. Lee Reiswig got a bug up his ass about the WPS though and that slipped the whole thing to where it lost the market
advantage it would have had shipping long before Win31. The WPS also blew the 4M memory target to shreds, which would have been met
by using the 1.3 shell. Sure, 2.0 would "run" in 4M, but it sucked bad, real bad.
You must remember - quite a ways into 2.0 development Microsoft was still wholeheartedly "in the boat" and viewed Win31 as a low end
toy, not a serious contender. It was only after IBM/Reiswig insisted on including that damn WPS in 2.0 that MS really (in
retrospect, wisely) decided to defect and ride the Win31/W95 horse. Its my considered opinion(as IBM's System test technical lead
for 1.3 and 2.0) that the inclusion of WPS in the 2.0 release is was the "root cause" of all OS/2's later problems - because it
caused MS to jump ship and caused a big loss of inertia. From that point on IBM was fighting an uphill battle.
> And you guys did a damn fine job. Some people claim that OS/2 1.3 was
> the leanest and meanest PC OS ever
I would be one of those people<g>. The 1.3 release got my official blessing as tech test lead. On a 25mhz 486, 1.3 rocked! It
would take a P166MMX to get similar responsiveness out of Windows 95. Linux or NT don't even come close. OS/2 1.3 was indeed a
thoroughbred the likes of which have never been seen since. Make it PROTECTONLY and 1.3 was near bullet proof too. A Mod 70 with
6M on my desk ran 1.3 for something like 3 years straight 7x24 (other than forced reboots when the site power was turned off over
long holidays) and NEVER crashed when running and that was doing all sorts of development work, LAN attached, running 3270
emulators, etc. It was the most reliable machine I've ever had. A few of my M95's running NT may come close eventually, but
they've only been powered up around 2 years now without crashing ;->
2.0 went out the door without my blessing - indeed I cautioned that it would likely be a support disaster and needed at least 3 more
months - which it was and it did ;-> My warnings were cast aside and the four horsemen of the apocalypse rode into Boca for a
lengthy stay after 2.0 shipped ;->
> Funny thing is that although Microsoft supposedly didn't do any work
> on 1.3, they were still selling it as part of LAN Manager...
To be fair, there were some OEM adaptation work MS did, as well as adding tha LADR device driver interface. I don't believe they
added anything substantial to the performance tuning and memory work IBM did though. MS had full rights to all of 1.3 as part of
the JDA (Joint Development Agreement), so they were free to do as they cared with their version.
"Michal Necasek" <mic...@prodigy.net> wrote in message news:3C95638F...@prodigy.net...
Michal Necasek
> 2.0 and 1.3 were under development concurrently - 2.0 was concurrent with 1.2 as well.
>
OS/2 1.2, never mentions 1.3.
> 2.0 could have shipped 9 months to a year earlier if it went out with the slightly modified 1.3 shell it sported for so long during
> development - IOW, plenty of time to beat Windows 3.1 to market. This was the path I and many others within IBM advocated - as well
> as Microsoft. Lee Reiswig got a bug up his ass about the WPS though and that slipped the whole thing to where it lost the market
> advantage it would have had shipping long before Win31. The WPS also blew the 4M memory target to shreds, which would have been met
> by using the 1.3 shell. Sure, 2.0 would "run" in 4M, but it sucked bad, real bad.
>
Yet IBM for some strange reason insisted on giving 4MB as a minimum
requirement on the box.
> You must remember - quite a ways into 2.0 development Microsoft was still wholeheartedly "in the boat" and viewed Win31 as a low end
> toy, not a serious contender.
>
It _was_ a toy. A very popular toy.
> It was only after IBM/Reiswig insisted on including that damn WPS in 2.0 that MS really (in
> retrospect, wisely) decided to defect and ride the Win31/W95 horse. Its my considered opinion(as IBM's System test technical lead
> for 1.3 and 2.0) that the inclusion of WPS in the 2.0 release is was the "root cause" of all OS/2's later problems - because it
> caused MS to jump ship and caused a big loss of inertia. From that point on IBM was fighting an uphill battle.
>
That's interesting. Are you saying that had IBM not pushed WPS so
hard, MS might never have split?
You're probably right that WPS shouldn't have been in 2.0. Maybe 2.1
or 2.2. And/or it should have been optional - it wouldn't be impossible.
I mean, it's not hard to use MSHELL and have no WPS.
In retrospect, WPS was one of OS/2's strongest weapons - but almost no
one knew how to use it.
>> And you guys did a damn fine job. Some people claim that OS/2 1.3 was
>>the leanest and meanest PC OS ever
>
> I would be one of those people<g>. The 1.3 release got my official blessing as tech test lead. On a 25mhz 486, 1.3 rocked! It
> would take a P166MMX to get similar responsiveness out of Windows 95. Linux or NT don't even come close. OS/2 1.3 was indeed a
> thoroughbred the likes of which have never been seen since. Make it PROTECTONLY and 1.3 was near bullet proof too. A Mod 70 with
> 6M on my desk ran 1.3 for something like 3 years straight 7x24 (other than forced reboots when the site power was turned off over
> long holidays) and NEVER crashed when running and that was doing all sorts of development work, LAN attached, running 3270
> emulators, etc. It was the most reliable machine I've ever had.
>
Wow! That says something for both the software and hardware.
> A few of my M95's running NT may come close eventually, but
> they've only been powered up around 2 years now without crashing ;->
>
> 2.0 went out the door without my blessing - indeed I cautioned that it would likely be a support disaster and needed at least 3 more
> months - which it was and it did ;-> My warnings were cast aside and the four horsemen of the apocalypse rode into Boca for a
> lengthy stay after 2.0 shipped ;->
>
Yeah, I've heard many many horror stories about 2.0...
>>Funny thing is that although Microsoft supposedly didn't do any work
>>on 1.3, they were still selling it as part of LAN Manager...
>
> To be fair, there were some OEM adaptation work MS did, as well as adding tha LADR device driver interface. I don't believe they
> added anything substantial to the performance tuning and memory work IBM did though. MS had full rights to all of 1.3 as part of
> the JDA (Joint Development Agreement), so they were free to do as they cared with their version.
>
I thought MS added LADDR to OS/2 1.2? Or you mean that they added it
to whatever they got from IBM?
Michal
Tony Ingenoso
mailboxes, priorities, etc as a DOS device driver back in 1984.
"Gfretwell" <gfre...@aol.com> wrote in message news:20020317233641...@mb-ci.aol.com...
Gfretwell
MYTE and still have my 4 3270 sessions. Bill never knew he didn't "own" the
system.
Unal Z
3C95913B...@prodigy.net...
> > Sure, 2.0 would "run" in 4M, but it sucked bad, real bad.
> >
> Yet IBM for some strange reason insisted on giving 4MB as a minimum
> requirement on the box.
4 MB was the typical amount of memory most machines had.
Mod. 70 for instance shipped in 1990 with 4 MB (had initially
only 2 MB) while the kloning society was delivering only 2 MB.
My dealer offered me about a $ 1000 (thousand!) discount
in return for the extra 2 MB on my mod. 70 because he could
make $ 250 on it the next day (had I known that memory prices
would go down I would have accepted his offer).
Memory was expensive and IBM memory was more than
expensive. As 2.0 was released, memory sales boosted
also because IBM advised 6 MB for a better performance.
Third party, e.g. OKI memory, 2 MB, was priced about the
same as the OS/2 2.0 package. Even though we upgraded,
it wasn't much fun at all with 6 MB on a 20 Mhz 386.
It was slow, very slow, although the machines we had were
near top of the line (ok, 486 came). DOS with Win 3.x was
however much faster and this was an argument.
IBM could not recommend more MBs than people were actually
having or willing to buy just to run 2.0, an OS intended for the
mass market. The greatest marketing trick could have been
offering a 2 MB as a bonus to everyone buying OS/2 2.0.
In retrospect, the 4 MB min was a necessity of the times.
--
Unal
Michal Necasek
>
> IBM could not recommend more MBs than people were actually
> having or willing to buy just to run 2.0, an OS intended for the
> mass market. The greatest marketing trick could have been
> offering a 2 MB as a bonus to everyone buying OS/2 2.0.
>
poor may have hurt OS/2 even more than high memory requirements.
There is something to say for being honest and upfront :-) This
way many people probably got the impression that OS/2 2.0 was
dog slow and didn't realize how much an extra memory module
would help.
Heck, with 256M RAM OS/2 2.0 (and not just 2.0) really flies ;-)
> In retrospect, the 4 MB min was a necessity of the times.
>
Then I guess the WPS wasn't, as Tony said...
Michal
Tim Clarke
> > 2.0 and 1.3 were under development concurrently - 2.0 was concurrent
with 1.2 as well.
> >
> I know. The 2.0 SDK (1990) I have only talks about compatibility with
> OS/2 1.2, never mentions 1.3.
>
> > 2.0 could have shipped 9 months to a year earlier if it went out
> > with the slightly modified 1.3 shell it sported for so long during
> > development - IOW, plenty of time to beat Windows 3.1 to market.
> > This was the path I and many others within IBM advocated - as well as
Microsoft.
> > Lee Reiswig got a bug up his ass about the WPS though and that
> > slipped the whole thing to where it lost the market advantage it would
have had
> > shipping long before Win31.
> > The WPS also blew the 4M memory target to shreds, which would have been
met
> > by using the 1.3 shell. Sure, 2.0 would "run" in 4M, but it sucked bad,
real bad.
> >
> Yet IBM for some strange reason insisted on giving 4MB as a minimum
> requirement on the box.
IIRC, historically speaking, this was the recommended minimum for running
Win 3.0 and, presumably, IBM was attempting to force a "no system upgrade
necessary" marketing criterion onto the development people?
--
Regards,
Tim Clarke (Guildford, UK)
Tony Ingenoso
want to compromise ship dates for it. Users were already familiar with the Win30-OS/2 1.3 style UI so there would not be the big
learning curve either.
MS was between a rock and a hard place at that point in time. Win30 was enough of a success to demonstrate that that sort of thing
was viable in the market, but had enough technical problems that it couldn't become a defacto standard with a long shelf life (IBM's
original projections<ha ha ha> were for OS/2 2.0 to ship unchanged for 3 years!). OS/2 was still to be the next generation at that
point(with NT the generation after that). When Brad Silverberg came onboard at MS and took over Win31 I suspect Microsoft started
to think Win31 might be a viable "plan-B". Brad was a very good manager and whipped that product into pretty good shape in short
order -- now they had something that would get'em out from under the middling reputation that Win30 had and it would ship before
2.0+WPS did and not be a big retraining effort for the users.
So basically, yes, I believe IBM executive boneheadedness drove MS to seek alternatives in the form of Win31.
The MS/IBM OS/2 versions were never quite the same, we did various things for our customers, they did various things for theirs.
The 1.2 product was a joint development venture - both companies took away what they wanted at the end.
"Michal Necasek" <mic...@prodigy.net> wrote in message news:3C95913B...@prodigy.net...
>
> That's interesting. Are you saying that had IBM not pushed WPS so
> hard, MS might never have split?
> .....
Tony Ingenoso
hardware guys were pissed because OS/2 sucked so bad on the machine, so PSP footed the bill for the extra ram to keep the customers
quiet.
All those 4M preloaded 55SX's came back llike boomerangs.
"Unal Z" <un...@nonesuch.invalid> wrote in message news:3c962db5$0$31266$6e36...@newsreader02.highway.telekom.at...
Unal Z
> Unal Z wrote:
> >
> > IBM could not recommend more MBs than people were actually
> > having or willing to buy just to run 2.0, an OS intended for the
> > mass market. The greatest marketing trick could have been
> > offering a 2 MB as a bonus to everyone buying OS/2 2.0.
> >
> I don't know. Recommending 4MB even though the performance was
> poor may have hurt OS/2 even more than high memory requirements.
Marketing is about selling. After you buy it, performance becomes
your machine's problem. But they said "minimum" and told the truth.
Many other companies do exactly the same, just check the Win reqs.
Apart from the memory, you needed a better graphics performance
which the plain VGA could not deliver. You needed a faster HD for
that large swap file. You needed a bigger monitor for those windows.
You needed finally a new, better machine. You were not crazy and
not rich to change your machine just for an OS (the rich had macs).
I bought mine with the desire to run the next OS/2, since the
current version back then, 1.3, was awfully expensive, $ 2000.
I've now become aware that I own four versions of OS/2 + one beta,
1.3 + 2.0 + 2.99 (beta) + 3.1 + 4.0. Performance counts to my oldies
and I would say 1.3 + 2.99 + 3.1 were my choices. Especially the
beta was quick, I used it for a very long time until I got 3.1 and stripped
off its networking function (3.1 is Warp Server = Warp + LAN Server).
4.0 calls for a higher Pent class machine, 3.1 does quite good with P133,
memory hunger is modest, 32 MB is a toast, 48 is a burger, all more than
that is a gourmet dinner. I have a quick HD servant.
I just checked the DB2 Server 2.1 for OS/2. They have a min memory
requirement of only additional 3 MB. Have much fun. Fall asleep.
> There is something to say for being honest and upfront :-) This
> way many people probably got the impression that OS/2 2.0 was
> dog slow and didn't realize how much an extra memory module
> would help.
People knew and learned it, all mags wrote about that. But you
couldn't so easily afford memory, at least not here in AT. It was
simply expensive, ask the dealers. In my experience, a faster HD
compensates for a few MBs, a faster HD makes much difference
on OS/2.
> > In retrospect, the 4 MB min was a necessity of the times.
> >
> Then I guess the WPS wasn't, as Tony said...
Depends on who sees what. I agree with Tony's arguments. We
failed lately on a quite different project because of similar reasons.
Very few people care about technical excellence. Very few people
understand it at all. Very few people live for that excellence. It is
a difficult life.
--
Unal
Michal Necasek
> Absolutely - Microsoft desperately wanted to get the 2.0 product out the door.
>
I can see why - the DOS compatibility of 1.x sucked while the 2.0 MVDM
was and probably still is the best in the industry. And of course 32-bit
programming is SO much nicer.
> So basically, yes, I believe IBM executive boneheadedness drove MS to seek alternatives
> in the form of Win31.
>
Most people say that IBM can be a very difficut company to work with :-)
> The MS/IBM OS/2 versions were never quite the same, we did various things for our customers,
> they did various things for theirs. The 1.2 product was a joint
development venture - both
> companies took away what they wanted at the end.
>
I didn't notice any really major differences between the MS and IBM
releases apart from device support. I believe the MS versions were sold
exclusively through OEMs and hence did not for instance support PS/2
hardware at all - while the IBM versions officially only supported IBM
boxes but in my experience ran on pretty much anything as long as the
hardware was sufficiently AT compatible.
Oh, one difference I remember is that MS had dual boot since 1.0
unlike IBM. And of course IBM had the EE versions while MS had LAN
Manager and SQL Server.
Michal
William Walsh
Really? I use Warp 4 on my Model 77 Bermuda with 32MB RAM. Runs beautifully,
even with a few applications open. The only thing I've found is a
requirement is a fairly large hard disk.
William
Tony Ingenoso
of opportunity striving for ideal. Corollary: MS has also been real good at knowing when to cut and run (ex. "Bob" and that Turbo
Pascal clone they had for a while)
This was one of the fundamental differences between the two companies at that time. IBM idealism versus Microsoft pragmatism. The
difference in this case was a critical one and had its origins decades earlier. Pre-PC, the world IBM played in had longer
time-to-market, wasn't as dynamic, etc -- IBM could plan 5 years ahead and have reasonable expectations the plan wouldn't need a lot
of revision over time. IOW, IBM could drive the market at their own pace, and did. Microsoft grew up and came of age in the
microprocessor era and was used to the dynamic markets and constantly shifting plans of the small machine world -- they were used to
being driven by market forces rather than driving it the way IBM had in the past.
So here we are 10 years down the road, and it feels like IBM and Microsoft have switched places. Today's nimble new "Gerstnerized"
IBM would never have made the planning blunders the 1990 IBM did, and today's Microsoft is acting like the heavy handed
quasi-dictator IBM from the 60's/70's and 80's ;->
"Michal Necasek" <mic...@prodigy.net> wrote in message news:3C968370...@prodigy.net...
Michal Necasek
> This was one of the fundamental differences between the two companies at that time. IBM idealism versus Microsoft pragmatism. The
> difference in this case was a critical one and had its origins decades earlier. Pre-PC, the world IBM played in had longer
> time-to-market, wasn't as dynamic, etc -- IBM could plan 5 years ahead and have reasonable expectations the plan wouldn't need a lot
> of revision over time. IOW, IBM could drive the market at their own pace, and did. Microsoft grew up and came of age in the
> microprocessor era and was used to the dynamic markets and constantly shifting plans of the small machine world -- they were used to
> being driven by market forces rather than driving it the way IBM had in the past.
>
> So here we are 10 years down the road, and it feels like IBM and Microsoft have switched places. Today's nimble new "Gerstnerized"
> IBM would never have made the planning blunders the 1990 IBM did, and today's Microsoft is acting like the heavy handed
> quasi-dictator IBM from the 60's/70's and 80's ;->
>
I read a couple of books written by ex-Microsofties (Barbarians Led by
Bill Gates is a good one). They talk about how Microsofties despised
lumbering IBM with many layers of bureaucracy... and now Microsoft is
just like that and probably even worse.
Maybe we're in for another cycle :-) If competition and DOJ can't
break Microsoft's power, we'll have to wait till MS does it itself.
It's just a question of time.
Michal
Tony Ingenoso
guarantee that you could pump plans/specs in on one end and get shippable, decent quality product coming out the other end and NOT
need a bunch of prima donna superstars with attitudes to pull it off. IBM had quite a few very good people of our own with
generally less attitude than many of the MS crew had, and they understood the value of the process for the most part -- particularly
the testing aspect.
IBM test people humbled some of the MS prima donnas pretty early on and earned a grudging level of respect that Microsoft never gave
testing previously. The IBM test manager for 1.0 had a simple rule - "the system shall not crash". It was non-negotiable, and to
the best of my knowledge, 1.0 was the only OS/2 release that shipped without a single known (at the time) crash/hang/data corruption
bug.
What Microsoft has learned over the last 15 years is that you can't "wing it" for very large software projects. That process the
bitched about is the only way to keep a leviathan on track and on some sort of schedule. Without it, you wind up with the situation
Netscape had with Gecko - "the never ending product development cycle". Netscape had a bunch of prima donnas with attitude, and
nobody wanted to do the dirty work of pulling it all together. Well...things like NLS enablment, DBCS support, BiDi, UniCode, and
all the other "scutt work" involved in a major internationalized product ARE dreadfully dull, painstaking, and don't offer a lot of
personal satisfaction -- BUT someone has to do them else you don't have a real product. The IBM process simply assigned someone to
do them, and that's that, they do get done<g>. The Linux community is experiencing these same issues today - nobody wants to do
the "scutt work" to turn the product into something that has the fit/finish and overall completeness that it so desperately needs.
"Michal Necasek" <mic...@prodigy.net> wrote in message news:3C96ABA3...@prodigy.net...
Gfretwell
events and might only happen once or twice a year. I couldn't believe how flaky
PC operating systems were. That is why I still like DOS. I still never seem to
get through a day with Windoze that I don't end up booting because of some kind
of software anomaly. I am using W/95c and that seems to be pretty stable, in a
MS sort of way but still nowhere as good as old DOS 6.3
I "play" with windoze and rice burners but I run my business on DOS (on PS/2s)
Michal Necasek
> I came into this from the IBM mainframe business where IPL's were scheduled
> events and might only happen once or twice a year. I couldn't believe how flaky
> PC operating systems were. That is why I still like DOS.
>
apps. That's its biggest weakness. And when some program is stomping
over someone else's memory, finding it can be pretty hard.
> I still never seem to
> get through a day with Windoze that I don't end up booting because of some kind
> of software anomaly. I am using W/95c and that seems to be pretty stable, in a
> MS sort of way but still nowhere as good as old DOS 6.3
> I "play" with windoze and rice burners but I run my business on DOS (on PS/2s)
>
In my experience if a machine is very unstable, it is good to suspect
the hardware first (unless it's a PS/2 <g>). The worst problem is that
it's often unclear whether a problem is caused by hardware or software,
although after years of experience I can usually take a fairly accurate
guess. Still not always though.
The nice thing about hardware problems is that once you figure out
which bit is bad (that's the hard part), fixing it is easy - just
throw out the failing part an replace by good one. Replacement
software parts are sadly not so easy to get.
Michal
Michal Necasek
> Yep, they thought the IBM "process" was cumbersome, which it was to a degree. However what that process did, was virtually
> guarantee that you could pump plans/specs in on one end and get shippable, decent quality product coming out the other end and NOT
> need a bunch of prima donna superstars with attitudes to pull it off. IBM had quite a few very good people of our own with
> generally less attitude than many of the MS crew had, and they understood the value of the process for the most part -- particularly
> the testing aspect.
>
QA department :-) I'm not happy when they find a bug in my code but
I know that it's much better than customers finding it.
I think the prima donna aspect is exemplified by the fact that
many of the old-time Microsofties are known by name: Aaron Reynolds
(remember his evil AARD code in Win3.1?), Neil Konzen, Mark Zbikowski
(of MZ fame - nearly every EXE still has his initials!), Gordon Letwin.
In contrast, virtually no one outside IBM knows the names of the IBM
programmers.
Michal
Tony Ingenoso
Ring lan stuff. Bermuda 77's are also in this same stability class running NT. My machines with these setups quite frankly never
crash, BSOD, or need reboot. They're so reliable its boring<g>
Win95 has been more problematic...
Most of the DOS releases are pretty good - even DOS 4 with its bad rep isn't bad if you run it on the hardware it claims to support
specifically.
"Gfretwell" <gfre...@aol.com> wrote in message news:20020318235650...@mb-ce.aol.com...
Tony Ingenoso
I think a few IBM execs got miffed when PC Week profiled me next to Lotus's Ray Ozzie in a section they were running on hardcore
techie industry notables. I got more high visibility press in one day than a lot of them had gotten in years ;->
Then again, most IBM execs of that era didn't have a clue how to relate to the industry press people either. I'd developed a
cordial "backchannel" relationship with Larry Seltzer (then at ZD Labs) and he was routinely calling me up asking my opinion about
various technical issues (usually unrelated to anything regarding IBM). He apparently considered me a straight shooter who knew
what he was talking about technically. ZD apparently wasn't willing to trust the tech propaganda from the various manufacturers and
was seeking independent opinion on things. I suspect it was Larry who prompted PC Week to do the profile on me.
"Michal Necasek" <mic...@prodigy.net> wrote in message news:3C96D06D...@prodigy.net...
Gfretwell
on them and halfway removing it. I think they call it DLL Hell.
With no real standards the hardware and software behaves like the crowd after a
soccer game that went against the home team.
I am an Ebay junkie and I am always trying to get trash back to treasure. I did
get my "DOS only" IBM Options scanner to work with W/95 Paint Shop Pro4. ...
But when I shoved the 4th card in a 2144 I broke it (If you are very selective
you can share the ISA/PCI slot. I just had to cut off the Game/Midi port of my
ISA sound card to make room for the RJ45 of an ethernet PCI card. Unfortunately
I ran out of IRQs.
Michal Necasek
> Neil wasn't too bad, Gordon and Zibbo OTOH... ;->
>
> I think a few IBM execs got miffed when PC Week profiled me next to Lotus's Ray Ozzie in a section they were running on hardcore
> techie industry notables. I got more high visibility press in one day than a lot of them had gotten in years ;->
>
Hehe. When was this (I don't really read PC Week so I missed that)?
> Then again, most IBM execs of that era didn't have a clue how to relate to the industry
> press people either.
>
And the press didn't know how to relate to IBM execs I suppose ;-)
The difference in public relations between IBM and MS was really
striking.
Michal
Bob Eager
wrote:
> > 4.0 calls for a higher Pent class machine
>
> Really? I use Warp 4 on my Model 77 Bermuda with 32MB RAM. Runs beautifully,
I ran WArp 4 for about 18 months on an 8580 with BL upgrade and 16MB.
Not fast, but acceptable!
--
Bob Eager
http://www.tavi.co.uk/ps2pages
http://change.to/ps2
Unal Z
a762l...@enews4.newsguy.com...
They all run, no doubt. I used it on a 25 Mhz 486SX with 21 MB
for a while but Warp 3 performed much better. You do some Java
and you see immediately the difference. It is not so much about how
the OS alone runs, rather how well apps with different demands will
run. The higher Pent gives Merlin/Warp 4 the push not only for what
I want to do but also for application mixes with substantial demands.
--
Unal
Unal Z
> Tony Ingenoso wrote:
> > IBM had quite a few very good people of our own with
> > generally less attitude than many of the MS crew had,
> > and they understood the value of the process for the most part
> > -- particularly the testing aspect.
> Heh, that reminds me of the love/hate relationship I have with our
> QA department :-) I'm not happy when they find a bug in my code but
> I know that it's much better than customers finding it.
I had a co-worker with an overly allergic response to bug reporting,
he just refused to accept the facts. After a while, I stopped telling
him of his sinister bugs.
I have the opposite attitude earned from the years long experience
running a one-man software company: I am *very* grateful to
everyone who reports a bug in my code since it saves me a lot of
pain and time. The only problem is that they do not seem to find
any bugs and that makes me very suspicious, since I know where
that bug has been hidden.
I'd rather spend a beer or two to that guy from the QA.
--
Unal
Tony Ingenoso
was bullshit - they couldn't even tell us which of the API's (pen emulating a mouse) their supposed "testing" covered (100% would
have been the right answer<g>)
So...I spent a weekend at the hotel writing/testing a mouse API snooper, in the form of a TSR, that would record everything that got
hit during any given test run. It turned out that only about 70% of the API's were ever being called by their tests.
Then I ran my DOS kernel unit test with their drivers loaded and it turned out their code was destroying the upper halfs of some of
the 32 bit registers. The developers denied the bug when I verbally told them. I gave'em the test log showing exactly what got
whacked, where it got whacked, and when it got whacked. The unit test I'd written preloaded registers with signature values prior
to an API and checked all non-result ones after to ensure they were the same, as well as making sure nothing like the DF flag got
altered. The evidence was air tight - their shit simply didn't work.
One of them stormed off after being presented with all this damning evidence -- still not admitting there was bug. An hour later
they came back, still denying there was a "bug", but saying "it was fixed now" ;->
I love deflating egos like that<g>. Apparently the "test group" in this company wasn't very technical and got little respect, nor
were they capable of earning any. This was reflected in the distain their developers had for testing and test people. That IBM of
all people, could walk in and "bitch slap" them around in such a definitive manner came as a rather rude suprise. I think they
expected they could get over on us easily and blow the whole thing off. Imagine their suprise when the IBM lab rats turned out not
to be gullible bozos like the marketing dweebs and executives they'd dealt with previously. "Ha ha ha, up yours bucko" was my
feeling ;->
"Unal Z" <un...@nonesuch.invalid> wrote in message news:3c9748b6$0$25550$6e36...@newsreader02.highway.telekom.at...
Tony Ingenoso
"Michal Necasek" <mic...@prodigy.net> wrote in message news:3C96E941...@prodigy.net...
Unal Z
Tony Ingenoso <aingeno...@prodigy.net> schrieb in im Newsbeitrag:
9vBl8.8040$tR2.155...@newssvr16.news.prodigy.com...
> I've found NT4 Server (SP6a) ..... to be exceptionally stable on a M95
I did many bad things to it, my programs crashed but it didn't.
Some IBM programs did worse things to it but they crashed.
Absolutely stable with ASUS, Matrox, 3Com 100 and a number
of experimental and industrial servers running on it. It scared me.
--
Unal
Unal Z
> I spent a couple of weeks in CA working with the company that provided the
handwriting recognition stuff for PCDOS.
Thanks for the story, Tony. I'd better not tell mine, it is worse.
You told it all.
> I love deflating egos like that<g>.
Warning: deflating egos like that may make enemies.
Another story?
> Imagine their suprise when the IBM lab rats turned out not
> to be gullible bozos like the marketing dweebs and executives
> they'd dealt with previously. "Ha ha ha, up yours bucko" was my
> feeling ;->
Somebody's got to show them. You did it. I tell them "look,
the computer is not going to understand that you are so
much more intelligent, so there is no use in proving that to it,
prove it better to me and fix that damn bug." It works but
not always.
--
Unal
Tony Ingenoso
been bothered by enemies -- they're predictable. Its your supposed friends you have to worry about ;->
All I ever cared about as a developer or tester was the quality of the end result. If bruising a few egos was needed to make a
product better, then thats what had to happen. Of course you don't go out of your way to make enemies, but if being nice doesn't
work, than being not nice is what you're left with.
One of the great American generals once said something to the effect that he'd rather have his men fear and respect him than like
him. I believe the same probably should hold true for test organizations. Being liked is nice, but being feared and respected
works very well indeed too ;-> This implies the test organization has the technical ability to earn that fear and respect<g> If
you don't have that, then you're probably left with trying to be nice (and getting walked all over in the process)
IBM, either by accident or design, always managed to sprinkle some of the very best technical people around in the test
organizations. This helped a lot to earn respect for testing. I always prefered testing to pure development work. The areas of
the product you could rampage over were always much broader - you could develop a good sense of how the whole fit together. This
was something a developer working on some little corner of the world rarely got to see. I got to write as much code as I wanted to
and do development work anyway, because I'd frequently submit a fix along with the defect reports.
"Unal Z" <un...@nonesuch.invalid> wrote in message news:3c97d66f$0$10316$5039...@newsreader01.highway.telekom.at...
David L. Beem
> One of the great American generals once said something to the effect
Is he on the same battlefield as his men? Maybe fear his abilities.
Respect doesn't have to have the fear-factor though, even for a general.
David
Da...@gilanet.com
deo
Tony Ingenoso wrote:
> ... apparently considered me a straight shooter who knew
> what he was talking about...
That's precisely why I tune in :-) (Superb thread, BTW.)
Michal Necasek
> One of the great American generals once said something to the effect that he'd rather
> have his men fear and respect him than like him.
>
say they (at least to certain extent) mutually exclusive. But that's
besides the point.
> hold true for test organizations. Being liked is nice, but being feared and
> respected works very well indeed too ;->
>
I'm sure it does :-)
> IBM, either by accident or design, always managed to sprinkle some of the
> very best technical people around in the test organizations.
>
And then doesn't listen to them ;-)
> This helped a lot to earn respect for testing. I always prefered testing
> to pure development work. The areas of the product you could rampage
> over were always much broader - you could develop a good sense of how
> the whole fit together. This was something a developer working on
> some little corner of the world rarely got to see. I got to write as
> much code as I wanted to and do development work anyway, because I'd
> frequently submit a fix along with the defect reports.
>
I think I understand your attidude very well. Come to think of it, I
too for some reason prefer to fix bugs to writing new code. These two
activities require slightly different skills - a tester (in your sense
of the word, that is a "fix developer") has to be able to zoom in on
a problem, forget that the whole project has X million lines of source
code and navigate straight to the source of the problem. More often
than not, the fix is very easy and localized to one particular place
in the source code. Sometimes I'm amazed at how much can be fixed by
adding (or removing!) one line of code :-)
And I have to admit that I like going on "bug hunts", pack up my
debugger and not stop until the bug is localized, classified and
exterminated. I suppose in many ways it's similar to real hunts
with a rifle. Software bugs can be pretty darn elusive too.
Michal
Unal Z
> I've never been bothered by enemies -- they're predictable.
> Its your supposed friends you have to worry about ;->
Wise words, Tony, many thanks. You can never watch them enough,
the so-called friends.
> All I ever cared about as a developer or tester was the quality of
> the end result. If bruising a few egos was needed to make a
> product better, then thats what had to happen. Of course you
> don't go out of your way to make enemies, but if being nice doesn't
> work, than being not nice is what you're left with.
It is a very competitive business and most people perceive it rather
as a competition for intellectual superiority. People seem to neglect
that, although the work has to do a lot with intellectual activity,
there are proven models, processes, methods and approaches one
can learn well with the necessary patience and dilligence, with
hard learning work. That is why I admire the hardware people, they
know how to learn (ain't it so, you chip hackers?)
Being nice is also telling people all about these things, but there will
always be some who will perceive it as an assault to their intellectual
abilities. Then, as you say, "being not nice is what you're left with."
> One of the great American generals once said something to the
> effect that he'd rather have his men fear and respect him than like
> him. I believe the same probably should hold true for test
> organizations. Being liked is nice, but being feared and respected
> works very well indeed too ;->
Being liked is a luxury in this business. It is a great feeling, a great
thing. Ironically, you earn it by being first hated and feared, and the
respect comes only when you prove your abilities. Or, you are a
human relations and communications genius and can skip the
painful phase. So, I would say, with the necessary backup of
competence and achievements, this is the way I learned to go.
It has lots to do with the culture, but in general, you've got to
be hard and fair, respect and encourage the smallest achievements.
I mean, how many "nice" trainers are there? Just look at the faces
of those "angry" football (soccer) trainers shouting from the side line.
> IBM, either by accident or design, always managed to sprinkle
> some of the very best technical people around in the test
> organizations. This helped a lot to earn respect for testing.
I read about the testing process in a lengthy article in the
IBM Systems Journal. It earned my unconditional respect.
Naively, I proposed to emulate it in my former organization.
Tony, please keep us telling of your rich experiences. I would be
not the only to say that it is very reassuring to see and very helpful
to learn how this damn business works.
--
Unal
Johnny Johnson
> Hi Tony,
>
>> One of the great American generals once said something to the effect
>> that he'd rather have his men fear and respect him than like him...
--General George S. Patton
> Is he on the same battlefield as his men?
Georgie was notorious for "leading from the front." <g>
David L. Beem
> --General George S. Patton
Same one as "make the poor S.O.B. die for *his* country". He was in my
neck of the woods years ago chasing down Pacho Villa. Like Sadam, the Army
missed catching Pancho.
> > Is he on the same battlefield as his men?
> Georgie was notorious for "leading from the front." <g>
I don't agree that his men may have feared him. If he "led from the
front", his men would have respected him.
David
Da...@gilanet.com