Dave Cutler, Prism, DEC, Microsoft, etc.
Neil Rieck
can't be true. So imagine my surprise when I bumped into this:
http://radsoft.net/rants/20040831,00.shtml
<quote>Dave always wanted to rewrite VMS in C. He hated Unix but loved
C. As soon as he'd finished VMS he suggested the rewrite. He was
turned down flat. Several years later he found himself in Seattle and
essentially was doing the rewrite in C when word came DEC were tired
of him.</quote>
This article contains lots of other (possibly) questionable facts, but
think about the statement above "rewrite VMS in C". If this had
happened, we would have VMS or OpenVMS on any platform including
x86-64
NSR
Richard B. Gilbert
Not ANY platform! There are architectural requirements; things like
number and sizes of registers, processor modes, etc.
Many years ago, somebody wrote PCVMS. I bought a copy. It had a DCL
like command language, and VMS-like system services. It was NOT VMS.
Michael Kraemer
Neil Rieck <n.r...@sympatico.ca> writes:
> Since any idiot can put up a web page these days, you know it all
> can't be true. So imagine my surprise when I bumped into this:
>
> http://radsoft.net/rants/20040831,00.shtml
>
And:
"Dave took his engineers and his new operating system called 'Prism' cross town
to the Redmond campus."
How that?
Presumably everything he did at M$ would be IP of DEC,
so he couldn't just "take it with him", unless DEC permitted it.
AFAIK he left as a "disgruntled employee" so DEC would have had
even less reason to give it away.
Bill Gunshannon
Neil Rieck <n.r...@sympatico.ca> writes:
That's funny. I have always wanted to see a project started to take
FreeBSD and rewrite it in Pascal or even better, Ada. :-)
Oh well, I suppose that could be another of those retirement projects
I keep piling up.
bill
--
Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves
bill...@cs.scranton.edu | and a sheep voting on what's for dinner.
University of Scranton |
Scranton, Pennsylvania | #include <std.disclaimer.h>
Bob Gezelter
Neil,
"Written in C" and "architecture independent" are two very different
things. With all due respect, I will take exception to the "on any
platform" observation in the original post.
The are hardware dependencies for each of the architectures, and there
are hardware presumptions that all three current architectures have in
common. Some of the most important were noted in my July 2001
"Thoughts on the Alpha->Itanium Announcement" (see http://www.rlgsc.com/alphaitanium.html).
In any event, the use of MACRO-32 is not the impediment. It has been
proven twice (e.g., ALPHA, Itanium) that a compiler can be
straightforwardly constructed to convert MACRO-32 in to the native
binary needed. In fact, having worked on portable code generators
during my university research time, I could observe that one could
write a MACRO-32 -> C translator, which would then "solve" the initial
problem of getting onto a new architecture. Admittedly, such a
bootstrapping approach has efficiency issues, but they can be dealt
with.
The more significant problems are the details of the hardware
presumptions I mentioned in my July 2001 comments, and the
architecture specific code for any new architecture. If it is
technically doable, it remains a significant financial item.
Admittedly, I have not done an indepth review of X86-64 in these
respects, my comments are meant in the general sense.
- Bob Gezelter, http://www.rlgsc.com
VAXman-
Wendin Associate. I had a copy too when I was working at Lakehurst
NAEC. It was considered a toy and little more.
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
http://www.quirkfactory.com/popart/asskey/eqn2.png
"Well my son, life is like a beanstalk, isn't it?"
Allen, Daniel P.
>And:
Speaking as an old personal friend of one of his top team members I think I can say that the description is reasonably accurate.
Dan
Neil Rieck
presented H1N1 symptoms and have been banished to my house until
Monday (Nov-9). I was totally out of commission for 2-days and it took
another 3-days before I stopped presenting symptoms. I'm still
sleeping way too long each night.
During this time, some UNIX heads were sending me questions regarding
"Dave Cutler, Prism, Emerald, Mica" on my OpenVMS page so I decided to
continue my research to find out what really happened. That's when I
stumbled onto these two articles:
http://www.nationmaster.com/encyclopedia/DEC-PRISM (read to last
paragraph talking about Olsen and Alpha)
-AND-
http://windowsitpro.com/Articles/Print.cfm?ArticleID=7153 (read whole
article as well as readers' comments posted to the end)
Neil Rieck
Kitchener / Waterloo / Cambridge,
Ontario, Canada.
http://www3.sympatico.ca/n.rieck/
Richard B. Gilbert
> In article <iNqdnfhzKIfVi2nX...@giganews.com>, "Richard B. Gilbert" <rgilb...@comcast.net> writes:
>> Neil Rieck wrote:
>>> Since any idiot can put up a web page these days, you know it all
>>> can't be true. So imagine my surprise when I bumped into this:
>>>
>>> http://radsoft.net/rants/20040831,00.shtml
>>>
>>> <quote>Dave always wanted to rewrite VMS in C. He hated Unix but loved
>>> C. As soon as he'd finished VMS he suggested the rewrite. He was
>>> turned down flat. Several years later he found himself in Seattle and
>>> essentially was doing the rewrite in C when word came DEC were tired
>>> of him.</quote>
>>>
>>> This article contains lots of other (possibly) questionable facts, but
>>> think about the statement above "rewrite VMS in C". If this had
>>> happened, we would have VMS or OpenVMS on any platform including
>>> x86-64
>>>
>>> NSR
>> Not ANY platform! There are architectural requirements; things like
>> number and sizes of registers, processor modes, etc.
>>
>> Many years ago, somebody wrote PCVMS. I bought a copy. It had a DCL
>> like command language, and VMS-like system services. It was NOT VMS.
>
> Wendin Associate. I had a copy too when I was working at Lakehurst
> NAEC. It was considered a toy and little more.
>
It might have become more than a toy but it would not build on the then
current version of Microsoft C.
Neil Rieck
> On Nov 6, 6:53 am, Neil Rieck <n.ri...@sympatico.ca> wrote:
>
>
> Neil,
>
> "Written in C" and "architecture independent" are two very different
> things. With all due respect, I will take exception to the "on any
> platform" observation in the original post.
>
I couldn't agree more. But whenever a new hardware platform appears,
it is always *nix/c that gets there first because, as we all know, C
is not much more than a portable assembler. Is the first port any
good? Nope, it is usually just a very first step. Regarding your
comments on MACRO32, I guess I was thinking more of other weirdness
like BLISS.
I wonder what Cutler was think about.
NSR
Neil Rieck
lots of original PDFs published by DECwest Engineering in the late
1980's
http://bitsavers.org/pdf/dec/prism/
Tim E. Sneddon
I'm sure I'll get someone flaming me for this, but there is nothing
wrong with portability in BLISS (BLISS-10 and BLISS-11, yes). Common
BLISS was used to write a number of software packages including
DECnet, RUNOFF and others (including compilers) that were portable
(or certain portions anyway) across PDP-10, VAX and PDP-11. Not to
mention the fact that BLISS was the original implementation language
for the GEM compiler backend. That has been ported natively to IA-32,
IA-64, Alpha and MIPS as well as cross-compiling off of the VAX.
As for the statement that C is little more than a portable
assembler...well, that like a whole bunch of other stuff gets
flung around on a regular basis. BLISS is much more like a portable
assembler than C. Especially when it comes to type (mis-)matching.
Almost all languages have their place. It is the coders and their
standards that make the big difference in portable code.
>
> I wonder what Cutler was think about.
>
Who knows? Although you can point your browser to the link below
for a whole bunch of information on MICA as well as some internal
mail messages regarding the cancellation of PRISM, Mica and Ozix.
http://www.bitsavers.org/pdf/dec/prism/
Let's also not forget Titan, SAFE, H-32 and all the other designs
that, from what I understand, were eventually consolidated into
the PRISM project.
Tim.
Tim E. Sneddon
> In any event, the use of MACRO-32 is not the impediment. It has been
> proven twice (e.g., ALPHA, Itanium) that a compiler can be
> straightforwardly constructed to convert MACRO-32 in to the native
I think that might be a bit of a stretch of the meaning of
straightforward.
> binary needed. In fact, having worked on portable code generators
> during my university research time, I could observe that one could
Portable code generators almost always go:
High-Level-Language --> Intermediate-Language --> Machine-Code
They rarely go back the other way.
> write a MACRO-32 -> C translator, which would then "solve" the initial
> problem of getting onto a new architecture. Admittedly, such a
> bootstrapping approach has efficiency issues, but they can be dealt
> with.
I think that might be a pretty *huge* generalisation there. Source
to source translating is not generally a big task (depending on the
languages), but certainly going from machine level to a high level
language is. I think there would be much more than some "efficiency
issues". There especially needs to be careful consideration of the
target architecture's calling standard and how it compares to the
source. As I am sure you are aware writing code in assembly allows
the coder to do all sorts of neat "tricks" that just cannot be
replicated by a high level language.
Actually, a recent post by VAXman is the perfect example. It sites
a piece of code that passes information through registers R1-R3.
How do you translate that to a HLL like C using a compiler and
without significant hand-made changes? You couldn't even move that
to BLISS without a special linkage definition that would need to
be hand constructed. How can you pass out-of-band information on
an architecture that doesn't have enough registers?
Tim.
Phillip Helbig---undress to reply
<78797018-2cfb-402d...@p8g2000yqb.googlegroups.com>, Neil
Rieck <n.r...@sympatico.ca> writes:
> This article contains lots of other (possibly) questionable facts, but
> think about the statement above "rewrite VMS in C". If this had
> happened, we would have VMS or OpenVMS on any platform including
> x86-64
??????
Michael Moroney
>Bob Gezelter wrote:
>> In any event, the use of MACRO-32 is not the impediment. It has been
>> proven twice (e.g., ALPHA, Itanium) that a compiler can be
>> straightforwardly constructed to convert MACRO-32 in to the native
>I think that might be a bit of a stretch of the meaning of
>straightforward.
It certainly is. There are lots and lots of "tricks" used on VAX
assembly code that the Macro-32 compiler on Alpha/Itanic won't tolerate.
(Believe me, I know from experience)
>> write a MACRO-32 -> C translator, which would then "solve" the initial
>> problem of getting onto a new architecture. Admittedly, such a
>> bootstrapping approach has efficiency issues, but they can be dealt
>> with.
>I think that might be a pretty *huge* generalisation there. Source
>to source translating is not generally a big task (depending on the
>languages), but certainly going from machine level to a high level
>language is. I think there would be much more than some "efficiency
>issues". There especially needs to be careful consideration of the
>target architecture's calling standard and how it compares to the
>source. As I am sure you are aware writing code in assembly allows
>the coder to do all sorts of neat "tricks" that just cannot be
>replicated by a high level language.
One of the biggies is accessing portions of the stack not written by
the current routine. About the simplest example is messing with the
return address such as this:
JSB FOO
...
FOO:
...
ADDL #4,SP ; throw away return value
RSB ; return to the caller's caller
This returns not to the instruction after the JSB, but to the caller of
the routine that called FOO. Not uncommon in some code. Other code often
has "parameters" embedded in the code stream after the JSB instruction,
the called routine fetches the parameters and manipulates the return
address to return to the instruction after the parameters, not to the
first parameter itself.
>Actually, a recent post by VAXman is the perfect example. It sites
>a piece of code that passes information through registers R1-R3.
>How do you translate that to a HLL like C using a compiler and
>without significant hand-made changes? You couldn't even move that
>to BLISS without a special linkage definition that would need to
>be hand constructed. How can you pass out-of-band information on
>an architecture that doesn't have enough registers?
I won't bore everyone with dozens of examples of code snippets that
won't reverse translate into C or something, and aren't even accepted by
the Itanic Macro-32 compiler.
Michael Kraemer
So this means he stole DECs IP and got away with it?
(where are the IP zealots when they're needed?)
Bob Eager
> In article <iNqdnfhzKIfVi2nX...@giganews.com>, "Richard B.
> Gilbert" <rgilb...@comcast.net> writes:
>>Neil Rieck wrote:
>>> Since any idiot can put up a web page these days, you know it all
>>> can't be true. So imagine my surprise when I bumped into this:
>>>
>>> http://radsoft.net/rants/20040831,00.shtml
>>>
>>> <quote>Dave always wanted to rewrite VMS in C. He hated Unix but loved
>>> C. As soon as he'd finished VMS he suggested the rewrite. He was
>>> turned down flat. Several years later he found himself in Seattle and
>>> essentially was doing the rewrite in C when word came DEC were tired
>>> of him.</quote>
>>>
>>> This article contains lots of other (possibly) questionable facts, but
>>> think about the statement above "rewrite VMS in C". If this had
>>> happened, we would have VMS or OpenVMS on any platform including
>>> x86-64
>>>
>>> NSR
>>
>>Not ANY platform! There are architectural requirements; things like
>>number and sizes of registers, processor modes, etc.
>>
>>Many years ago, somebody wrote PCVMS. I bought a copy. It had a DCL
>>like command language, and VMS-like system services. It was NOT VMS.
>
> Wendin Associate. I had a copy too when I was working at Lakehurst
> NAEC. It was considered a toy and little more.
It's wasn't a serious attempt. The real product was a small OS kernel,
with multiple 'personalities' laid on top. There was a PC-UNIX which was
equally unlike UNIX.
I have disks and manuals for PC-VMS somewhere...
--
Use the BIG mirror service in the UK:
http://www.mirrorservice.org
JF Mezei
> You will need to forgive me this week. Last Thursday (Oct 29) I
> presented H1N1 symptoms and have been banished to my house until
> Monday (Nov-9).
Is it safe for you to send emails ? Someone running a microsoft PC might
get infected reading your messages :-)
Were you confirmed H1N1, or did you just get a a bad flue ?
Rob Brown
> So this means he stole DECs IP and got away with it?
> (where are the IP zealots when they're needed?)
According to some of the links that NSR gave us, DEC sued Microsoft
over this, and won. Although as also seemed to happen in DEC's suit
against Intel, DEC came out on the short end of the settlement.
--
Rob Brown b r o w n a t g m c l d o t c o m
G. Michaels Consulting Ltd. (780)438-9343 (voice)
Edmonton (780)437-3367 (FAX)
http://gmcl.com/
VAXman-
I have the disks (or images of) but no manuals. Can you scan them and
make them available? I'll put them on-line with the disk images if you
would.
Bob Eager
Yes, but since Vista has broken my scanner driver, it may take a while...
Michael Kraemer
> On Fri, 6 Nov 2009 at 20:38 +0100, Michael Kraemer wrote:
>
>> So this means he stole DECs IP and got away with it?
>> (where are the IP zealots when they're needed?)
>
>
> According to some of the links that NSR gave us, DEC sued Microsoft over
> this, and won. Although as also seemed to happen in DEC's suit against
> Intel, DEC came out on the short end of the settlement.
Only if one presumes that DEC actually wanted to win the lawsuit.
It is more likely that DEC simply wanted to put some
pressure on intel so they could get rid of Alpha.
From this point of view DEC actually succeeded.
JF Mezei
> Only if one presumes that DEC actually wanted to win the lawsuit.
> It is more likely that DEC simply wanted to put some
> pressure on intel so they could get rid of Alpha.
> From this point of view DEC actually succeeded.
Palmer was already in talks with Pfeiffer when the Intel stuff came up.
So it was really Compaq telling DEC that it wasn't interested in buying
the chip business, so Palmer found a way to get rid of the money losing
stuff.
This is quite different from Cutler since at the time Cutler left,
Digital had not yet begun to self cannabalise so it would be small
enough to be bought by someone. Although DEC had stopped growing, it
wasn't clear at that time that its had begun what would end up being a
terminal decline a decade later. DEC's desire to be Microsoft's poodle
had not yet begun, so when Cutler left, there would have been no reason
for DEC to "give away" its valuable IP to a competitor.
JF Mezei
Microvax I on the west coast. This is circa 1982-1984.
Neil Rieck
> In article <78797018-2cfb-402d-9eb9-b1f109c76...@p8g2000yqb.googlegroups.com>,
>
> Neil Rieck <n.ri...@sympatico.ca> writes:
> > Since any idiot can put up a web page these days, you know it all
> > can't be true. So imagine my surprise when I bumped into this:
>
> >http://radsoft.net/rants/20040831,00.shtml
>
> And:
>
> "Dave took his engineers and his new operating system called 'Prism' cross town
> to the Redmond campus."
>
> How that?
> Presumably everything he did at M$ would be IP of DEC,
> so he couldn't just "take it with him", unless DEC permitted it.
> AFAIK he left as a "disgruntled employee" so DEC would have had
> even less reason to give it away.
Here is one page of several I recently found (I am still looking)
http://www.roughlydrafted.com/2009/07/30/readers-write-how-microsoft-got-windows-nt/
<quote>
“So, Cutler walked down the street to Microsoft and offered them Mica
which became NT. Later DEC sued MS and, in an out of court settlement,
got royalties for the filched technology. Part of the deal included
targeting NT (back) onto the Alpha platform.
”BTW, this was not an usual procedure at DEC. Many employees left the
company with intellectual property from a cancelled project under
their arm, with the understanding that if they made it a commercial
success then DEC would come back knocking on the door for for
royalties.“
</quote>
This page, as well as others, claim the back-porting of WindowsNT to
Alpha was part of the settlement. Microsoft was the true beneficiary
of this deal when Compaq decided to walk away from the WindowsNT on
Alpha business.
Neil Rieck
Kitchener / Waterloo / Cambridge,
Ontario, Canada.
http://www3.sympatico.ca/n.rieck/
http://www3.sympatico.ca/n.rieck/docs/dave_cutler-prism-mica-emerald-etc.html
VAXman-
Ah, Vista (Visual Interface Similar To Apple's) Just buy a Mac! ;)
I have a Canon LiDE. It works like a charm with Ubuntu Linux and
on my Mac.
If you'd snail mail them to me, I'd make the copies and smail mail
the originals back to you.
VAXman-
>In the VMS 20th anniversary PDF, it is revealed that Cutler designed the
>Microvax I on the west coast. This is circa 1982-1984.
I didn't know that Cutler was a hardware guy.
Neil Rieck
Not sure here if you are interchanging the names "DEC" and "Compaq" on
purpose or not. I never for a moment thought DEC (pre Palmer) wanted
to get rid of Alpha. I had lots of personal contact with DEC between
1978 through to 1993 and can tell you that this company exuded "a
passion for computing" which was infectious. However, I was working in
Maynard Massachusetts in April-1993 and can tell you that "DEC felt
different" (BTW, Palmer replaced Olsen in 1992).
Now if you were referring to DEC after Olsen, it is anyone's guess
what was going on at DEC. I'll tell you one thing, Robert Palmer did
not have a passion for computing that was anywhere close to that of
Ken Olsen.
Neil Rieck
I was advised not to leave my house unless I had problems breathing,
or was coughing up blood. However, if adults presented either diarrhea
or vomiting then H1N1 was considered a certainty (I had both). They
told me it would take 5-days to clear up and it did.
Oh well. Still alive and kicking.
NSR
Neil Rieck
stuff including: PRISM Specs, emails including one confirming the
resignation of Dave Cutler, etc.
http://bitsavers.org/pdf/dec/prism/
http://computer-refuge.org/bitsavers/pdf/dec/prism/
http://ftp7.freebsd.org/sites/www.bitsavers.org/pdf/dec/prism/
Neil Rieck
Kitchener / Waterloo / Cambridge,
Ontario, Canada
Bill Gunshannon
Or send them to bitsavers. I am sure Dave would be glad to hos them.
bill
--
Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves
bill...@cs.scranton.edu | and a sheep voting on what's for dinner.
University of Scranton |
Scranton, Pennsylvania | #include <std.disclaimer.h>
Bill Gunshannon
VAXman- @SendSpamHere.ORG writes:
> In article <00ae7e22$0$6694$c3e...@news.astraweb.com>, JF Mezei <jfmezei...@vaxination.ca> writes:
>>In the VMS 20th anniversary PDF, it is revealed that Cutler designed the
>>Microvax I on the west coast. This is circa 1982-1984.
>
> I didn't know that Cutler was a hardware guy.
I heard he was God. At least that's what is usually sounds like here.
Bob Eager
> If you'd snail mail them to me, I'd make the copies and smail mail the
> originals back to you.
A bit too expensive given their size, and the distance.
I may just set up a spare machine with XP...
Michael Kraemer
> <quote>
> �BTW, this was not an usual procedure at DEC. Many employees left the
> company with intellectual property from a cancelled project under
> their arm, with the understanding that if they made it a commercial
> success then DEC would come back knocking on the door for for
> royalties.�
> </quote>
And still people wonder why DEC went down the drain.
John Reagan
"Neil Rieck" <n.r...@sympatico.ca> wrote in message
news:78797018-2cfb-402d...@p8g2000yqb.googlegroups.com...
> Since any idiot can put up a web page these days, you know it all
> can't be true. So imagine my surprise when I bumped into this:
>
> http://radsoft.net/rants/20040831,00.shtml
>
> <quote>Dave always wanted to rewrite VMS in C. He hated Unix but loved
> C. As soon as he'd finished VMS he suggested the rewrite. He was
> turned down flat. Several years later he found himself in Seattle and
> essentially was doing the rewrite in C when word came DEC were tired
> of him.</quote>
>
> This article contains lots of other (possibly) questionable facts, but
> think about the statement above "rewrite VMS in C". If this had
> happened, we would have VMS or OpenVMS on any platform including
> x86-64
>
> NSR
I don't believe a word of it. VAXELN wasn't written in C, but EPASCAL.
PRISM was going to have its own implementation language (a Don MacClaren
invention) that was going to be a mixture of Pascal, a dash of PL/I, and
some new language features. I have one of the numbered preliminary language
manuals stuffed in a box.
John
Michael Kraemer
> On Nov 7, 3:58 am, Michael Kraemer <M.Krae...@gsi.de> wrote:
>>
>>Only if one presumes that DEC actually wanted to win the lawsuit.
>>It is more likely that DEC simply wanted to put some
>>pressure on intel so they could get rid of Alpha.
>> From this point of view DEC actually succeeded.
>
>
> Not sure here if you are interchanging the names "DEC" and "Compaq" on
> purpose or not.
Alpha was sold to intel in 1997, so still DEC/Palmer were in charge.
> I never for a moment thought DEC (pre Palmer) wanted
> to get rid of Alpha.
How could they, Alpha was announced the same year
Palmer took over.
> I had lots of personal contact with DEC between
> 1978 through to 1993 and can tell you that this company exuded "a
> passion for computing" which was infectious. However, I was working in
> Maynard Massachusetts in April-1993 and can tell you that "DEC felt
> different" (BTW, Palmer replaced Olsen in 1992).
Because they were already in deep sh**?
> Now if you were referring to DEC after Olsen, it is anyone's guess
> what was going on at DEC. I'll tell you one thing, Robert Palmer did
> not have a passion for computing that was anywhere close to that of
> Ken Olsen.
The same could be said for most other IT companies in the 1990s
when compared to the decades before.
SGI and old HP being sad examples.
Allen, Daniel P.
-----Original Message-----
>From: info-vax...@rbnsn.com [mailto:info-vax...@rbnsn.com] On Behalf Of John Reagan
>Sent: Saturday, November 07, 2009 11:53 AM
>To: info...@rbnsn.com
>Subject: Re: [Info-vax] Dave Cutler, Prism, DEC, Microsoft, etc.
>>"Neil Rieck" <n.r...@sympatico.ca> wrote in message
>>news:78797018-2cfb-402d...@p8g2000yqb.googlegroups.com...
>> Since any idiot can put up a web page these days, you know it all
>> can't be true. So imagine my surprise when I bumped into this:
>>
>> http://radsoft.net/rants/20040831,00.shtml
>>
>> <quote>Dave always wanted to rewrite VMS in C. He hated Unix but loved
>
>I don't believe a word of it. VAXELN wasn't written in C, but EPASCAL.
>PRISM was going to have its own implementation language (a Don MacClaren
>invention) that was going to be a mixture of Pascal, a dash of PL/I, and
>some new language features. I have one of the numbered preliminary language
>manuals stuffed in a box.
>John
PILLAR - the manual is in the PRISM/MICA directories on bitsaver.
Dan
_______________________________________________
Info-vax mailing list
Info...@rbnsn.com
http://rbnsn.com/mailman/listinfo/info-vax_rbnsn.com
Richard B. Gilbert
It was never a mystery to me! DEC was totally out of touch with the
market place. The the RAM for the Rainbow that DEC sold for ~$700, I
bought at the Trenton Computer Festival for $30 US! The 20 MB disk
drive that DEC sold for $2200, I bought for $300. The 5-1/4" floppy
disks that DEC sold for $5 US each, I bought for $0.50 each and
formatted them myself in spite of the fact that my Rainbow "could not
properly format a floppy disk!"
There were about 20 different models of a 20MB disk from DEC! The
differences between the models? It was the mounting hardware! I could
go on but I hope you get the idea.
Storage-Works was nothing short of brilliant but it was rather late and
the marketing was nothing short of idiocy!
I could go on but most of you know the litany as well as I do!
George Cook
> Neil Rieck schrieb:
>> On Nov 7, 3:58 am, Michael Kraemer <M.Krae...@gsi.de> wrote:
>>>
>>>Only if one presumes that DEC actually wanted to win the lawsuit.
>>>It is more likely that DEC simply wanted to put some
>>>pressure on intel so they could get rid of Alpha.
>>> From this point of view DEC actually succeeded.
>>
>>
>> Not sure here if you are interchanging the names "DEC" and "Compaq" on
>> purpose or not.
>
> Alpha was sold to intel in 1997, so still DEC/Palmer were in charge.
No. It was not. One more time, from Wikipedia:
"The Alpha architecture was sold, along with most parts of DEC,
to Compaq in 1998. Compaq, already an Intel customer, decided to
phase out Alpha in favor of the forthcoming Hewlett-Packard/Intel
Itanium architecture, and sold all Alpha intellectual property to
Intel in 2001"
I suspect that you are confusing the sale of the Hudson fab with
the sale of the IP.
George Cook
WVNET
Neil Rieck
[...snip...]
> > I never for a moment thought DEC (pre Palmer) wanted
> > to get rid of Alpha.
>
> How could they, Alpha was announced the same year
> Palmer took over.
>
Not wanting to split hairs here, but I was in attendance at the Alpha
21064 announcement in Toronto, and it was at an industry trade show
Feb-1992. All the handouts were cheap photocopies and most people
wondered if this announcement was for real. I was a real a-hole in
those days (more so than now, anyway) and so jokingly asked if DEC had
any free chip samples. This caused the DEC rep to reach into his
briefcase to hand me a bag of potato chips with a taped-on label of
the space shuttle with the moniker "Alpha Chips". I still have the
unopened bag of chips on top of my book case.
Anyway, the 21064 appeared later that September and Palmer took over
in October. Speaking of a-holes, Palmer was employed by DEC since 1985
so he must have known something of Alpha. The way he decided to sell
off pieces of DEC (like RDB) makes me think he was just a hatchet man.
No company could have survived that many cuts in so short a time
without bleeding to death.
JF Mezei
> I didn't know that Cutler was a hardware guy.
OK, I had opened the document in quicklook and couldn'to select text
from it. (quicklook is apple's implementation of
dir/show=formatted_content filename)
Now, I have it on a PDF reader:
##
Evolution of the architecture
For almost a year, the VAX development
and review teams worked
back and forth on the VAX architecture.
After four versions, the
proposed architecture was found
to be too complex, too expensive,
and too complicated to execute.
The company formed a group that
became known as �The Blue Ribbon
Committee� that included three
hardware engineers: Bill Strecker,
Richie Lary, and Steve Rothman,
and three software engineers: Dave
Cutler, Dick Hustvedt, and Peter
Lipman
##
##
With the VAX hardware development underway, the software
development�code named Starlet�began a few months later in June of
1975. Roger Gourd led the project and software engineers Dave Cutler, Dick
Hustvedt, and Peter Lipman were technical project leaders, each responsible
for a different part of the operating system.
##
##
After the 750, DIGITAL
designed the MicroVAX I�one of the first DIGITAL projects to include
silicon compilers�with the consulting help of Carver Meade, a pioneer in
integrated circuit design. Building on the experience of the MicroVAX I,
the company soon followed with the more powerful MicroVAX II.
While the V-11 was designed as a full VAX implementation, the MicroVAX I
was designed as a VAX subset. The MicroVAX I system was developed
in the company�s Seattle facility, headed by Dave Cutler. Because the
MicroVAX I was a much simpler design than the V-11, and because of the
use of the silicon compiler tools, it was completed before the V-11.
##
##
�The sense I always had was that there were four key technical
visionaries at
the beginning of MicroVAX: Dave Cutler, with his creation of the MicroVAX
I system for early software development; Bob Supnik, who headed up
MicroVAX chip development and also wrote the microcode; Jesse Lipcon
who headed up MicroVAX II Server Development; and Dick Hustvedt, who
drove the MicroVMS Software Strategy.�
�Jay Nichols
Computer Special Systems, Manager of Engineering
##
##
Prism: VMS on RISC technology
DIGITAL began working on RISC technology in 1986 when Jack Smith,
VP of Operations, tapped Dave Cutler on the shoulder and said, �You will be
RISC Czar for DIGITAL. Organize a program.� The program, code named
Prism, was to develop the company�s RISC machine. Its operating system
would embody the next generation of design principles and have a
compatibility layer for UNIX and VMS.
##
##
Alpha was very much the �son of Prism.� The primary changes made to
produce Alpha were for VMS compatibility. The original Prism design had
serious compatibility problems with the VAX and VMS in two areas�
numerical data types and privileged architecture.
##
##
�When our customers had beta
copies of Windows NT, they told us
that it felt like they were revisiting
an old friend. That�s not surprising
because the chief architect of both
operating systems was Dave Cutler.
So there is a natural affinity from a
technical perspective between the two
environments. Wes Melling is often
quoted calling it the �Cutler effect.��
�Mary Ellen Fortier
Director, OpenVMS Marketing
##
JF Mezei
> Now if you were referring to DEC after Olsen, it is anyone's guess
> what was going on at DEC. I'll tell you one thing, Robert Palmer did
> not have a passion for computing that was anywhere close to that of
> Ken Olsen.
I think there maybe stuff we don't know about Palmer's mandate.
About the same time Olsen was ousted, IBM had begun plans to dismantle
itself to pay off debts and be sold and/or avoid bankrupcy.
It is possible that the board realised that Digital was going in the
same direction and gave Palmer a mandate to "rationlize" instead of
"fix" Digital and then, seeing how rationalisation was making things
worse, told him to sell off parts that could be sold and seek a buyer
for the rest to try to maximize shareholder value.
Meanwhile, IBM hired Gerstner and he decided to put an end to the
"dismantle IBM" right away and fix the company, which he succeeded.
JF Mezei
>> Not sure here if you are interchanging the names "DEC" and "Compaq" on
>> purpose or not.
>
> Alpha was sold to intel in 1997, so still DEC/Palmer were in charge.
Palmer began to discuss merger with Compaq's Pfeiffer in 1996. Talks
lasted 3 years. (The 3 year counter stopped on the day of the merger
announcement, not when it was signed/sealed/delivered, so perhaps the
talks began in 1995).
JF Mezei
> I suspect that you are confusing the sale of the Hudson fab with
> the sale of the IP.
In exchange for Digital apologizing to Intel, Intel agreed to relieve
Digital of its money losing Hudson Plant.
Basically, Digital found a way to get rid of a money losing plant which
Compaq was not interested in by leveraging the fact that Intel had
stolen Alpha technologies years before.
DEC should have sued Intel well before that time. It is only because
Compaq had told Palmer to shed the chip making business (Compaq did't
want it) that Palmer suddently decided to pursue this matter.
Arne Vajhøj
The projects were cancelled at DEC, so I doubt that
preventing other from using the ideas would have made
much of a difference.
Arne
Arne Vajhøj
> Michael Kraemer wrote:
>> Neil Rieck schrieb:
>>
>>> <quote>
>>> �BTW, this was not an usual procedure at DEC. Many employees left the
>>> company with intellectual property from a cancelled project under
>>> their arm, with the understanding that if they made it a commercial
>>> success then DEC would come back knocking on the door for for
>>> royalties.�
>>> </quote>
>>
>> And still people wonder why DEC went down the drain.
>
> It was never a mystery to me! DEC was totally out of touch with the
> market place. The the RAM for the Rainbow that DEC sold for ~$700, I
> bought at the Trenton Computer Festival for $30 US! The 20 MB disk
> drive that DEC sold for $2200, I bought for $300. The 5-1/4" floppy
> disks that DEC sold for $5 US each, I bought for $0.50 each and
> formatted them myself in spite of the fact that my Rainbow "could not
> properly format a floppy disk!"
>
> There were about 20 different models of a 20MB disk from DEC! The
> differences between the models? It was the mounting hardware! I could
> go on but I hope you get the idea.
That explanation sounds a lot more convincing than IP going
out the door.
Arne
Arne Vajhøj
> Allen, Daniel P. schrieb:
>>> On Behalf Of Michael Kraemer
>>> Neil Rieck <n.r...@sympatico.ca> writes:
>>> "Dave took his engineers and his new operating system called 'Prism'
>>> cross town
>>> to the Redmond campus."
>>
>>> How that?
>>> Presumably everything he did at M$ would be IP of DEC,
>>> so he couldn't just "take it with him", unless DEC permitted it.
>>> AFAIK he left as a "disgruntled employee" so DEC would have had
>>> even less reason to give it away.
>>
>> think I can say that the description is reasonably accurate.
>
> So this means he stole DECs IP and got away with it?
> (where are the IP zealots when they're needed?)
Would it have made a big difference?
VMS was not the only multiuser OS.
They could have used ideas from Unix, Multics etc. instead.
No NT user, sysadm or user mode code developer could tell
the difference anyway.
It would have been different from driver writers and
other kernel mode code developers, but even though they may
be the guys that know most about the OS, then I don't think
they have much influence on decisions on what OS to use.
Arne
Michael Kraemer
No, I'm well aware of the difference,
but I think it does not matter really.
The 1997 deal with intel (and a similar one with Samsung)
effectively marked alpha's sell out,
even if DEC kept IP rights on paper.
Moreover,
intel agreed to manufacture Alpha, a possible competitor to their
upcoming IA64.
At about the same time it was planned to port DEC Unix to IA64.
All that makes only sense if alpha's death was already sealed,
maybe secretly, with the 1997 deal.
Compaq just had to do the cleanup when
the time was right, in 2001, with the HP takeover in sight.
Arne Vajhøj
> In article <78797018-2cfb-402d...@p8g2000yqb.googlegroups.com>,
> Neil Rieck <n.r...@sympatico.ca> writes:
>> Since any idiot can put up a web page these days, you know it all
>> can't be true. So imagine my surprise when I bumped into this:
>>
>> http://radsoft.net/rants/20040831,00.shtml
>>
>> <quote>Dave always wanted to rewrite VMS in C. He hated Unix but loved
>> turned down flat. Several years later he found himself in Seattle and
>> essentially was doing the rewrite in C when word came DEC were tired
>> of him.</quote>
>>
>> This article contains lots of other (possibly) questionable facts, but
>> think about the statement above "rewrite VMS in C". If this had
>> happened, we would have VMS or OpenVMS on any platform including
>> x86-64
>
> FreeBSD and rewrite it in Pascal or even better, Ada. :-)
> Oh well, I suppose that could be another of those retirement projects
> I keep piling up.
C is undoubtedly easy to write OS code in (no surprise since
it was designed for this type of purpose), but it has proven to
relative hard to avoid buffer overruns, memory leaks etc.
(buffer runs are probably more relevant than memory leaks for OS).
It would be rather interesting with an OS in Ada or Modula-2/Oberon
(standard Pascal lacks too much - DEC Pascal may do).
Arne
Arne Vajhøj
> Since any idiot can put up a web page these days, you know it all
> can't be true. So imagine my surprise when I bumped into this:
>
> http://radsoft.net/rants/20040831,00.shtml
>
> <quote>Dave always wanted to rewrite VMS in C. He hated Unix but loved
> C. As soon as he'd finished VMS he suggested the rewrite. He was
> turned down flat. Several years later he found himself in Seattle and
> essentially was doing the rewrite in C when word came DEC were tired
> of him.</quote>
>
> This article contains lots of other (possibly) questionable facts, but
> think about the statement above "rewrite VMS in C". If this had
> happened, we would have VMS or OpenVMS on any platform including
> x86-64
If the various owners of VMS had believed in the market for
VMS/x86-64, then I don't think getting Bliss and Macro-32 for
x86-64 would have been the showstopper. My guess is that it
would have been a small part of the overall project. Porting
an OS to a new platform is more than just building the
source code.
Arne
Arne Vajhøj
> Neil Rieck wrote:
>> Since any idiot can put up a web page these days, you know it all
>> can't be true. So imagine my surprise when I bumped into this:
>>
>> http://radsoft.net/rants/20040831,00.shtml
>>
>> <quote>Dave always wanted to rewrite VMS in C. He hated Unix but loved
>> C. As soon as he'd finished VMS he suggested the rewrite. He was
>> turned down flat. Several years later he found himself in Seattle and
>> essentially was doing the rewrite in C when word came DEC were tired
>> of him.</quote>
>>
>> This article contains lots of other (possibly) questionable facts, but
>> think about the statement above "rewrite VMS in C". If this had
>> happened, we would have VMS or OpenVMS on any platform including
>> x86-64
>
> number and sizes of registers, processor modes, etc.
Not any platform.
But if there is let us call it flexibility to redesign to match
the new HW, then I would assume that all HW capable of running
OS/400, Unix, Linux, Windows NT could be made to run VMS.
Arne
Arne Vajhøj
> "Tim E. Sneddon" <tim.s...@bigpond.com> writes:
>> Bob Gezelter wrote:
>>> In any event, the use of MACRO-32 is not the impediment. It has been
>>> proven twice (e.g., ALPHA, Itanium) that a compiler can be
>>> straightforwardly constructed to convert MACRO-32 in to the native
>
>> I think that might be a bit of a stretch of the meaning of
>> straightforward.
>
> It certainly is. There are lots and lots of "tricks" used on VAX
> assembly code that the Macro-32 compiler on Alpha/Itanic won't tolerate.
> (Believe me, I know from experience)
Sure there would be problems.
But I don't think the problems would be great enough to stop
a porting projects with funding and support from the top.
Arne
Michael Kraemer
>
> Not wanting to split hairs here, but I was in attendance at the Alpha
> 21064 announcement in Toronto, and it was at an industry trade show
> Feb-1992. All the handouts were cheap photocopies and most people
> wondered if this announcement was for real. I was a real a-hole in
> those days (more so than now, anyway) and so jokingly asked if DEC had
> any free chip samples. This caused the DEC rep to reach into his
> briefcase to hand me a bag of potato chips with a taped-on label of
> the space shuttle with the moniker "Alpha Chips". I still have the
> unopened bag of chips on top of my book case.
So who was the a-hole then? :-)
> Anyway, the 21064 appeared later that September and Palmer took over
> in October. Speaking of a-holes, Palmer was employed by DEC since 1985
> so he must have known something of Alpha.
Sure, he was a semiconductor guy, iirc.
Of course pre-Palmer DEC did not want to get rid
of alpha, it was barely on the market then
(that's what I wanted to say).
One can't even say that Palmer-DEC wanted to get rid
of it in the early years, only later, when it turned
out to be an economic failure.
> The way he decided to sell
> off pieces of DEC (like RDB) makes me think he was just a hatchet man.
well, DEC was in such deep sh*t, maybe such a guy was needed?
> No company could have survived that many cuts in so short a time
> without bleeding to death.
Without selling parts of DEC, the company may have been
finished even earlier.
JF Mezei
> One can't even say that Palmer-DEC wanted to get rid
> of it in the early years, only later, when it turned
> out to be an economic failure.
The Hudson FAB wasn't underused because of lack of demand. From what I
was told, there were plenty of customers wanting DEC to build their
chips, but Palmer wanted to reserve FAB capacity in case Alpha became
popular. But by failing to address pricing, Alpha never became pppular
and Hudson FAB remain underused for years until it became obsolete at
which point DEC wanted to get rid of it without losing face. Hence the
Intel deal.
Michael Kraemer
> ##
> �When our customers had beta
> copies of Windows NT, they told us
> that it felt like they were revisiting
> an old friend. That�s not surprising
> because the chief architect of both
> operating systems was Dave Cutler.
> So there is a natural affinity from a
> technical perspective between the two
> environments. Wes Melling is often
> quoted calling it the �Cutler effect.��
> �Mary Ellen Fortier
> Director, OpenVMS Marketing
> ##
>
But why is it that,
whenever I'm forced to sit in front of a Windoze box,
I never feel "hey, this is just like VMS, only better"?
Michael Moroney
>This page, as well as others, claim the back-porting of WindowsNT to
>Alpha was part of the settlement. Microsoft was the true beneficiary
>of this deal when Compaq decided to walk away from the WindowsNT on
>Alpha business.
I remember it as Microsoft cancelling NT on Alpha after Compaq did
something (I don't remember what) to piss Microsoft off.
There were persistant rumors when I was in Compaq/late Digital that
the sources to VMS were part of the IP that M$ had access to, with the
idea that they could use the clustering technology, but M$ could never
figure it out.
Michael Kraemer
> The projects were cancelled at DEC, so I doubt that
> preventing other from using the ideas would have made
> much of a difference.
It still was DECs IP, and it was used to create
one of VMS' fiercest competitors.
Michael Kraemer
>
> That explanation sounds a lot more convincing than IP going
> out the door.
>
And yet another take on the whole DEC thing:
http://www.sigcis.org/files/Goodwin_paper.pdf
(gotta love armchair history :-)
Arne Vajhøj
But would its non-usage have had much impact?
Unix, OS/400, Linux seems to have worked out fine
without it.
Arne
Richard B. Gilbert
Well, it's clearly better in some ways! When $700 US buys the computer
and a licensed copy of Windows? ISTR that a VMS license sold for almost
$2000 per seat.
Michael Kraemer
> Michael Kraemer wrote:
>
>>
>> So this means he stole DECs IP and got away with it?
>> (where are the IP zealots when they're needed?)
>
>
> Would it have made a big difference?
>
I was wondering that an (apparently disgruntled) employee
can walk out of his previous employer's door,
with a tape (real or virtual) under his arm,
essentially offering that stuff to a competitor.
Without being sued.
Iirc SCO has sued IBM for lesser evidence.
Michael Kraemer
> Michael Kraemer wrote:
>
>>One can't even say that Palmer-DEC wanted to get rid
>>of it in the early years, only later, when it turned
>>out to be an economic failure.
>
> The Hudson FAB wasn't underused because of lack of demand. From what I
> was told, there were plenty of customers wanting DEC to build their
> chips, but Palmer wanted to reserve FAB capacity in case Alpha became
> popular.
Is there any first hand information on this one,
or do we have to rely, once again, to "heard it through the grapevine"?
> But by failing to address pricing, Alpha never became pppular
there are more reasons that Alpha did not succeed
(late market entry, ISVs not interested, development costs).
> and Hudson FAB remain underused for years until it became obsolete at
> which point DEC wanted to get rid of it without losing face.
I think they lost quite a bit of their face in that deal.
It signalled the end of Alpha. At least for those
who were able to read the signs on the wall.
> Hence the
> Intel deal.
Michael Kraemer
> Neil Rieck <n.r...@sympatico.ca> writes:
>
>
>>This page, as well as others, claim the back-porting of WindowsNT to
>>Alpha was part of the settlement. Microsoft was the true beneficiary
>>of this deal when Compaq decided to walk away from the WindowsNT on
>>Alpha business.
>
>
> I remember it as Microsoft cancelling NT on Alpha after Compaq did
> something (I don't remember what) to piss Microsoft off.
In 1999 Compaq announced discontinuance of
"support for NT" (whatever that means) on their Alpha machines
and withdrew their personnel from joint projects with M$.
Apparently Compaq found NT/x86 to be good enough
and counted on a future NT/itanic.
In turn M$ discontinued their apps for NT/alpha.
Compaq not only pissed off M$, but also loyal Alpha customers
who were allowed by their management to buy Alphas (with VMS)
because "they can also run NT".
Now this excuse had evaporated, another blow to Alpha (and VMS).
Michael Kraemer
> Michael Kraemer wrote:
>>
>> It still was DECs IP, and it was used to create
>> one of VMS' fiercest competitors.
>
> But would its non-usage have had much impact?
>
> Unix, OS/400, Linux seems to have worked out fine
> without it.
Inspecting other OSs from outside is one thing,
but having a competitors' chief developer
on board, together with source code and
implementation details, gives a much better
head start.
Arne Vajhøj
Things were different in 1988 than in 2009 when
it comes to IP.
Ken Olsen was an engineer not a lawyer.
And besides: how much was really copied?
Code was not copied - C instead of Bliss & Macro.
Shell was not copied - CMD instead of DCL (lousy change BTW).
GUI was not copied - MS Windows instead of DECWindows.
File system was not copied - NTFS instead of ODS-2.
Memory management and IO subsystem were apparently extremely
identical.
But if it ended up in court - how much of it was originall DEC and
not known from Multics/Unix/various IBM/standard computer science ?
(think MS - Apple - Xerox)
Probably still some, but probably not more than could easily be changed.
So DEC could only get as much much money as it would have cost MS
to rewrite N number of kernel modules.
Arne
Arne Vajhøj
Sure.
But that was not my point.
Those that I mention did not have a DEC chief developer
on board but managed to come through anyway.
Arne
Neil Rieck
True, but if Cutler wanted to immediately write VMS in C then I'm
thinking he must had a good reason for it. We all know that UNIX
enjoyed an immediate debugging when it was first rewritten from
assembler to "C". I wonder (in hindsight) if all these secondary and
tertiary components (like BLISS and MACRO) might be an indication of
too many chefs in the VMS kitchen :-)
NSR
Neil Rieck
> VAXman- @SendSpamHere.ORG wrote:
> > I didn't know that Cutler was a hardware guy.
>
> OK, I had opened the document in quicklook and couldn'to select text
> from it. (quicklook is apple's implementation of
> dir/show=formatted_content filename)
>
> Now, I have it on a PDF reader:
>
> ##
> Evolution of the architecture
> For almost a year, the VAX development
> and review teams worked
> back and forth on the VAX architecture.
> After four versions, the
> proposed architecture was found
> to be too complex, too expensive,
> and too complicated to execute.
>
> The company formed a group that
> became known as “The Blue Ribbon
> Committee” that included three
> hardware engineers: Bill Strecker,
> Richie Lary, and Steve Rothman,
> and three software engineers: Dave
> Cutler, Dick Hustvedt, and Peter
> Lipman
> ##
>
> ##
> With the VAX hardware development underway, the software
> development—code named Starlet—began a few months later in June of
> 1975. Roger Gourd led the project and software engineers Dave Cutler, Dick
> Hustvedt, and Peter Lipman were technical project leaders, each responsible
> for a different part of the operating system.
> ##
>
> ##
> After the 750, DIGITAL
> designed the MicroVAX I—one of the first DIGITAL projects to include
> silicon compilers—with the consulting help of Carver Meade, a pioneer in
> integrated circuit design. Building on the experience of the MicroVAX I,
> the company soon followed with the more powerful MicroVAX II.
> While the V-11 was designed as a full VAX implementation, the MicroVAX I
> was designed as a VAX subset. The MicroVAX I system was developed
> in the company’s Seattle facility, headed by Dave Cutler. Because the
> MicroVAX I was a much simpler design than the V-11, and because of the
> use of the silicon compiler tools, it was completed before the V-11.
> ##
>
> ##
> “The sense I always had was that there were four key technical
> visionaries at
> the beginning of MicroVAX: Dave Cutler, with his creation of the MicroVAX
> I system for early software development; Bob Supnik, who headed up
> MicroVAX chip development and also wrote the microcode; Jesse Lipcon
> who headed up MicroVAX II Server Development; and Dick Hustvedt, who
> drove the MicroVMS Software Strategy.”
> —Jay Nichols
> Computer Special Systems, Manager of Engineering
> ##
>
> ##
> Prism: VMS on RISC technology
> DIGITAL began working on RISC technology in 1986 when Jack Smith,
> VP of Operations, tapped Dave Cutler on the shoulder and said, “You will be
> RISC Czar for DIGITAL. Organize a program.” The program, code named
> Prism, was to develop the company’s RISC machine. Its operating system
> would embody the next generation of design principles and have a
> compatibility layer for UNIX and VMS.
> ##
>
> ##
> Alpha was very much the “son of Prism.” The primary changes made to
> produce Alpha were for VMS compatibility. The original Prism design had
> serious compatibility problems with the VAX and VMS in two areas—
> numerical data types and privileged architecture.
> ##
>
> ##
> “When our customers had beta
> copies of Windows NT, they told us
> that it felt like they were revisiting
> because the chief architect of both
> operating systems was Dave Cutler.
> So there is a natural affinity from a
> technical perspective between the two
> environments. Wes Melling is often
> —Mary Ellen Fortier
> Director, OpenVMS Marketing
> ##
Here is a quote from DEC Engineer Gordon Bell. (you can find the full
text online here: http://americanhistory.si.edu/collections/comphist/bell.htm).
According to this, DEC was drifting from project-to-project and really
didn't know what they had under Cutler
Transition from VAX to Prism, MIPS, and Alpha: Losing momentum
Next, I think what really got DEC into the most significant trouble
was the way it dealt with the transition from to RISC and to a 64-bit
address. Dave Cutler had an architecture called Prism that he had
designed at the Seattle lab. That was all done, the manuals were done,
people were working on chips, and the program was going along well.
Meanwhile, MIPS came to DEC and said: “Gee, you’re not there with RISC
or your one chip VAXen, you need a RISC machine for your workstations.
Why don’t you build a workstation on RISC?” And DEC did, introduced
it, and said: “Oh well, we’ll stay with MIPS.” Then they killed the
Prism project and Mr. Cutler left. They killed it, but Ken didn’t know
that it wasn’t dead. It was still alive in the semiconductor group and
it sprung up as Alpha. And so that came back several years later.
Meanwhile other people within the company were looking at building a
fast MIPS architecture machine including a group in Palo Alto which
built something called BIPS – a billion instructions per second
processor. In fact they have one. They had one about three years ago.
So all of those projects never came to market. And that’s why I said
when I left the company that you’ve got to get rid of VAX, you’ve got
to go open. The companies that I then started and worked with were
open systems companies. They were all UNIX. But it was deciding to go
to Alpha or deciding to do Prism, then killing Prism and going to
MIPS, and then coming back to Alpha and killing MIPS again. DEC could
have survived any of those decisions. It could have stayed with Prism,
got it out there a year earlier, and been significant in the
marketplace. It could have switched to MIPS, and I think that would
have probably been the best strategy. But coming in late, having to
build these very fancy FAB facilities to get the performance was
really costly. And today, there is no way I see that DEC can afford to
be a semiconductor supplier or microprocessor supplier when they have
to build their own, use their own, FAB facilities. So that was a
significant error in judgment and decision making. On the other hand,
the world is better off because Dave Cutler went to Microsoft and
built NT for a much larger market.
NSR
Michael Kraemer
> True, but if Cutler wanted to immediately write VMS in C then I'm
> thinking he must had a good reason for it. We all know that UNIX
> enjoyed an immediate debugging when it was first rewritten from
> assembler to "C". I wonder (in hindsight) if all these secondary and
> tertiary components (like BLISS and MACRO) might be an indication of
> too many chefs in the VMS kitchen :-)
The dish was different.
Unix was designed to run on just about any hardware,
whereas VMS' sole purpose was to help sell VAXen.
In the latter case one doesn't care much about
portability.
George Cook
> George Cook schrieb:
>> In article <hd49hq$m91$01$1...@news.t-online.com>, Michael Kraemer <M.Kr...@gsi.de> writes:
>
>>
>> No. It was not. One more time, from Wikipedia:
>>
>> "The Alpha architecture was sold, along with most parts of DEC,
>> to Compaq in 1998. Compaq, already an Intel customer, decided to
>> phase out Alpha in favor of the forthcoming Hewlett-Packard/Intel
>> Itanium architecture, and sold all Alpha intellectual property to
>> Intel in 2001"
>>
>> I suspect that you are confusing the sale of the Hudson fab with
>> the sale of the IP.
>>
>
> No, I'm well aware of the difference,
> but I think it does not matter really.
I guess it may not "really" matter in the grand scheme of the
universe, but in that case nothing at all matters "really". In
1997, DEC still claimed (and appeared) to believe in Alpha. I
went to all the same non-disclosure presentations as I suspect
others here did.
> The 1997 deal with intel (and a similar one with Samsung)
> effectively marked alpha's sell out,
> even if DEC kept IP rights on paper.
> Moreover,
> intel agreed to manufacture Alpha, a possible competitor to their
> upcoming IA64.
> At about the same time it was planned to port DEC Unix to IA64.
> All that makes only sense if alpha's death was already sealed,
> maybe secretly, with the 1997 deal.
Sounds a lot like "double secret probation" to me. IBM also agreed
to manufacture Alpha. If nothing else IBM is not stupid, and would not
have gotten involved in some super secret conspiracy to kill Alpha.
George Cook
WVNET
Simon Clubley
>
> It was never a mystery to me! DEC was totally out of touch with the
> market place. The the RAM for the Rainbow that DEC sold for ~$700, I
> bought at the Trenton Computer Festival for $30 US! The 20 MB disk
> drive that DEC sold for $2200, I bought for $300. The 5-1/4" floppy
> disks that DEC sold for $5 US each, I bought for $0.50 each and
> formatted them myself in spite of the fact that my Rainbow "could not
> properly format a floppy disk!"
>
It's still going on.
I currently have a quote on my desk for a Alpha server. The pricing
for 512MB of memory is over 900 UKP (which is around 1500 USD at current
exchange rates according to Google). (I don't have the exact price as I'm
not at work at the moment).
I don't expect commodity pricing in a server, but that's more than a bit
excessive.
Simon.
--
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980's technology to a 21st century world
Neil Rieck
Earth.UFP> wrote:
I agree. A few years back, I purchased a 1GB kit for AS-DS20e and we
paid $700 Canadian. The system ran so fast that we went back to them
for a second kit only to discover that the kit was now $600. We
purchased the kit from www.systemresale.ca who may have partners in
the UK (maybe they would ship to the UK?)
NSR
Neil Rieck
Show Stopper!: The Breakneck Race to Create Windows NT and the Next
Generation at Microsoft by G. Pascal Zachary
http://www.amazon.com/Show-Stopper-Breakneck-Generation-Microsoft/dp/0029356717
Showstopper! is a vivid account of the creation of Microsoft Windows
NT, perhaps the most complex software project ever undertaken. It is
also a portrait of David Cutler, NT's brilliant and, at times,
brutally aggressive chief architect.
Cutler surely ranks as one of the most impressive software engineers
the field has ever produced. After leading the team that created the
VMS operating system for Digital's VAX computer line--an
accomplishment that most would regard as a lifetime achievement--he
went on to conceive and lead the grueling multi-year project that
ultimately produced Windows NT. Both admired and feared by his team,
Cutler would let nothing stand in the way of realizing his design and
often clashed with his programmers, senior Microsoft management, and
even Gates himself. Yet no matter how involved he became in managing
his 100-programmer team, he continued to immerse himself in every
technical detail of the project and write critical portions of the
code himself.
Showstopper! is also a fascinating look at programmer and managerial
culture behind the Microsoft facade. The portraits of the men and
women who created NT not only reveal the brilliance of their work but
the crushing stress and the dislocating effects that new wealth had on
their lives. For some team members, the NT project ultimately
destroyed their marriages, friendships, and virtually every human
relationship outside of work. Showstopper! also reveals the
uncertainties, false starts, and blind alleys that dogged the project
as Microsoft repositioned NT from an improved OS/2 to something that
would ultimately challenge both OS/2 and Unix for the title of the
world's most powerful operating system.
Neil Rieck
Kitchener / Waterloo / Cambridge,
Ontario, Canada.
http://www3.sympatico.ca/n.rieck/
Arne Vajhøj
> On Nov 7, 9:27 pm, Arne Vajh�j <a...@vajhoej.dk> wrote:
>> Neil Rieck wrote:
>>> Since any idiot can put up a web page these days, you know it all
>>> can't be true. So imagine my surprise when I bumped into this:
>>> http://radsoft.net/rants/20040831,00.shtml
>>> <quote>Dave always wanted to rewrite VMS in C. He hated Unix but loved
>>> C. As soon as he'd finished VMS he suggested the rewrite. He was
>>> turned down flat. Several years later he found himself in Seattle and
>>> essentially was doing the rewrite in C when word came DEC were tired
>>> of him.</quote>
>>> This article contains lots of other (possibly) questionable facts, but
>>> think about the statement above "rewrite VMS in C". If this had
>>> happened, we would have VMS or OpenVMS on any platform including
>>> x86-64
>> If the various owners of VMS had believed in the market for
>> VMS/x86-64, then I don't think getting Bliss and Macro-32 for
>> x86-64 would have been the showstopper. My guess is that it
>> would have been a small part of the overall project. Porting
>> an OS to a new platform is more than just building the
>> source code.
>
> thinking he must had a good reason for it. We all know that UNIX
> enjoyed an immediate debugging when it was first rewritten from
> assembler to "C". I wonder (in hindsight) if all these secondary and
> tertiary components (like BLISS and MACRO) might be an indication of
> too many chefs in the VMS kitchen :-)
Today OS in C may be the standard.
But back in the mid 70's where the VMS decision were made, then
it was not the standard.
And my understanding is that VMS inherited this from
the PDP-11 OS's.
And when it was decided for them, then C was not available
(at least according to many C was invented to port Unix
to PDP-11).
Arne
Michael Kraemer
>
> I guess it may not "really" matter in the grand scheme of the
> universe, but in that case nothing at all matters "really". In
> 1997, DEC still claimed (and appeared) to believe in Alpha. I
> went to all the same non-disclosure presentations as I suspect
> others here did.
I think we all know the value of committments on paper.
Actions speak louder than words.
Handing over the flagship product to a direct competitor
in 1997, just 5 years after Alpha's introduction,
sent a strong negative signal.
It really didn't matter whether DEC still kept IP rights
or not.
Customers' comments in the trade press (at least in German),
including DECUS board members, were mainly negative,
certainly not enthusiastic.
> Sounds a lot like "double secret probation" to me.
What I hear is, that the guys from Microprocessor Report
came up with that "secret amendments" assumption.
But it makes a lot more sense to me than the rose-coloured pictures,
IMHO.
> IBM also agreed
> to manufacture Alpha. If nothing else IBM is not stupid, and would not
> have gotten involved in some super secret conspiracy to kill Alpha.
No conspiracy here. From 2000 onwards IBM manufactured Alpha's,
just as they produced HP's PA-RISC chips.
By that time such things weren't taboo anymore,
and no problem either, since both were a dying breed anyway.
Neil Rieck
As I have mentioned to others privately, if DEC would have listened to
Cutler and rewritten VMS in C, then porting VMS to Alpha, Itanium and
anything else would already have saved them tons of money.
Now if it is HP's intention to never port OpenVMS ever again, then
they should just sit back and milk it for all it's worth. However,
they might wish to consider tasking their Indian code-warriors to
quietly start a conversion to C. Even if the project was never fully
completed, future code maintenance and porting efforts would benefit
from this action.
NSR
Paul Raulerson
On Nov 7, 2009, at 8:18 PM, Arne Vajhøj wrote:
> Michael Kraemer wrote:
>> Allen, Daniel P. schrieb:
>>>> On Behalf Of Michael Kraemer
>>>> Neil Rieck <n.r...@sympatico.ca> writes:
>>>> "Dave took his engineers and his new operating system called
>>>> 'Prism' cross town
>>>> to the Redmond campus."
>>>
>>>> How that?
>>>> Presumably everything he did at M$ would be IP of DEC,
>>>> so he couldn't just "take it with him", unless DEC permitted it.
>>>> AFAIK he left as a "disgruntled employee" so DEC would have had
>>>> even less reason to give it away.
>>>
>>> Speaking as an old personal friend of one of his top team members
>>> I think I can say that the description is reasonably accurate.
>> So this means he stole DECs IP and got away with it?
>> (where are the IP zealots when they're needed?)
>
> Would it have made a big difference?
>
> VMS was not the only multiuser OS.
>
> They could have used ideas from Unix, Multics etc. instead.
>
> No NT user, sysadm or user mode code developer could tell
> the difference anyway.
>
> It would have been different from driver writers and
> other kernel mode code developers, but even though they may
> be the guys that know most about the OS, then I don't think
> they have much influence on decisions on what OS to use.
>
> Arne
Not to mention that there is at least as much "borrowed" technology
from UNIX in NT there is from VMS.
And when you hit Windows XP, there is far more UNIX based technology
cloned into NT than from VMS.
I have never fully understood this, since in some ways, VMS is clearly
superior to the contemporary UNIX
implementations of that time.
I suppose it was that DEC was still shuttered tightly with the idea
that computers would never be ubiquitious,
and therefor they should squeeze every nickle out of a limited market.
A shame that Olsen never realized
it was an essentially unlimited market. (*sigh*)
HP is faced with that legacy. If they were to release VMS as a free
OS, it would invade the data center's of the
world like wildfire. Especially if they went ahead and gave in to
porting it to Intel.
-Paul
Paul Raulerson
On Nov 7, 2009, at 8:25 PM, Arne Vajhøj wrote:
> Bill Gunshannon wrote:
>> In article <78797018-2cfb-402d...@p8g2000yqb.googlegroups.com
>> >,
>> Neil Rieck <n.r...@sympatico.ca> writes:
>>> Since any idiot can put up a web page these days, you know it all
>>> can't be true. So imagine my surprise when I bumped into this:
>>>
>>> http://radsoft.net/rants/20040831,00.shtml
>>>
>>> <quote>Dave always wanted to rewrite VMS in C. He hated Unix but
>>> loved
>>> C. As soon as he'd finished VMS he suggested the rewrite. He was
>>> turned down flat. Several years later he found himself in Seattle
>>> and
>>> essentially was doing the rewrite in C when word came DEC were tired
>>> of him.</quote>
>>>
>>> This article contains lots of other (possibly) questionable facts,
>>> but
>>> think about the statement above "rewrite VMS in C". If this had
>>> happened, we would have VMS or OpenVMS on any platform including
>>> x86-64
>> That's funny. I have always wanted to see a project started to take
>> FreeBSD and rewrite it in Pascal or even better, Ada. :-)
>> Oh well, I suppose that could be another of those retirement projects
>> I keep piling up.
>
> C is undoubtedly easy to write OS code in (no surprise since
> it was designed for this type of purpose), but it has proven to
> relative hard to avoid buffer overruns, memory leaks etc.
> (buffer runs are probably more relevant than memory leaks for OS).
>
> It would be rather interesting with an OS in Ada or Modula-2/Oberon
> (standard Pascal lacks too much - DEC Pascal may do).
>
> Arne
> _______________________________________________
Mmm- it is not difficult to avoid buffer overruns or memory leaks or
any other
problems that a lot of people think are endemic to C. You just have to
know
what you are doing, and *that* is true no matter what language you use.
Indeed, it is true of *anything* you do, but especially so in an
programming
exercise.
-Paul
VAXman-
>> >> Neil Rieck wrote:
>> >>> Since any idiot can put up a web page these days, you know it all
>> >>> can't be true. So imagine my surprise when I bumped into this:
>> >>>http://radsoft.net/rants/20040831,00.shtml
>d
>> >>> C. As soon as he'd finished VMS he suggested the rewrite. He was
>> >>> turned down flat. Several years later he found himself in Seattle and
>> >>> essentially was doing the rewrite in C when word came DEC were tired
>> >>> of him.</quote>
Keep it in Bliss and write a compiler for the target architecture,
if you believe it is that simple, would be the better solution.
--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG
http://www.quirkfactory.com/popart/asskey/eqn2.png
"Well my son, life is like a beanstalk, isn't it?"
JF Mezei
> I think we all know the value of committments on paper.
> Actions speak louder than words.
> Handing over the flagship product to a direct competitor
> in 1997, just 5 years after Alpha's introduction,
> sent a strong negative signal.
It wasn't so much the gift of the chip manufacturing that was the issue
but rather the fine print at the end of the deal where Digital committed
to porting Digital Unix (Tru64) to that then imaginary IA64 thing, but
did not commit to porting VMS to it. It is ironic that under HP, the
opposite happened.
The theft of Alpha IP happened before Pentium 3 was released, years
before. DEC let it go unchallenged for years before finally using that
card to get something from Intel (releive DEC of its chip business which
Compaq didn't want).
> Customers' comments in the trade press (at least in German),
> including DECUS board members, were mainly negative,
> certainly not enthusiastic.
At that point, it wasn't so much Alpha but Digital. People noticed
Palmer breaking up the company piece by piece.
> No conspiracy here. From 2000 onwards IBM manufactured Alpha's,
> just as they produced HP's PA-RISC chips.
From the time when Digital sold its chip manufacturing business, it was
perfectly normal that Digital would outsource manufacturing of its own
chips to the remaining manufacturers, whether Intel, IBM or any other.
JF Mezei
> As I have mentioned to others privately, if DEC would have listened to
> Cutler and rewritten VMS in C, then porting VMS to Alpha, Itanium and
> anything else would already have saved them tons of money.
I disagree with this. When DEC ported VMS to Alpha, it wasn't the
compilers that cost it money, it was the fact that they went with a
split code base.
When they ported from Alpha to that IA64 thing, they decided to do it
right and make the code more portable with a single code base. Once that
work is done, porting to another platform becomes much easier.
It is my understanding that all they need is a native assembler for the
really deep down stuff, MACRO compiler, C, and BLISS. Anything else ? I
think that they ported the last bits written in PL1 to something else,
right ?
John Wallace
Earth.UFP> wrote:
I don't know what memory you're buying but the fact it's for an Alpha
(Server) means its a long way from current generation memory and
therefore whether it's for an AlphaServer or a similar era Proliant or
whatever, you get to pay a premium for legacy memory. You're also
almost certainly paying the premium for ECC memory. ECC memory is more
expensive than non-ECC, far more than you'd justify just by the extra
bits, whether it's for DEC or anybody else. And you're payint the
premium for someone holding stock long after the volume market has
moved on.
Find out what kind of memory it is, find an equivalent memory for a
Proliant and compare the prices. If you want cheaper and unwarranted,
buy the Proliant (at your own risk :)). If you want really cheap
desktop-style prices, you have to go non-ECC, and that probably just
doesn't work. In between, there are various shades of grey; not all
1GB DDR400 36bit DIMMs are equal, for example. If you want someone to
warrant that it's going to work, someone's going to have to carry the
risk. You can carry that risk yourself by buying cheap, or you can pay
to have someone worry about it for you.
A similar story used to apply with DEC vs non-DEC SCSI drives. I used
to have one particular system builder that bought assorted non-DEC
drives and then complained when odd storage-related things happened.
"Reproduce the problem with any drive listed on the SPD and we can
talk about it further". His problems never were reproducible. RAM's
simpler than SCSI (e.g. no firmware) but not all RAM of a given size
and technology is the same.
One last thing re RAM prices: the cheap RAM traditionally came from
people buying it on the spot market whereas DEC in general bought
their RAM on contract at committed prices (and in committed
quantities) many months ahead. This mean that DEC prices were often
late following the market down, but it also meant that in times of RAM
shortage, DEC had a chance of getting supplies, whereas the folks
relying on the spot market could advertise whatever prices they want,
it didn't matter because the spot market folks had no chance of
actually delivering anything, because the contracted RAM customers had
all the supplies. You may not have known this, it wasn't the only
reason why DEC prices were seen as OTT, but it was a factor in the RAM
prices.
I don't know if there was a spot market as well as a contract market
in SCSI storage but it wouldn't be a surprise if there was and if the
same price-related logic applied to storage as well as RAM.
Michael Kraemer
> It wasn't so much the gift of the chip manufacturing that was the issue
> but rather the fine print at the end of the deal where Digital committed
> to porting Digital Unix (Tru64) to that then imaginary IA64 thing,
Sure. It doesn't require a lot of conspiracy theory
to guess what was negotiated behind the scenes.
Sell off Alpha to intel, who will manufacture it
as a stop-gap until Itanics arrive.
> but
> did not commit to porting VMS to it. It is ironic that under HP, the
> opposite happened.
well, collateral damage. HP found itself with two Unices
(three if one includes Linux) in the portfolio,
so the stepchild with least marketshare had to go.
This time it was an advantage that there is no other flavor of VMS.
> The theft of Alpha IP happened before Pentium 3 was released, years
> before. DEC let it go unchallenged for years before finally using that
> card to get something from Intel (releive DEC of its chip business which
> Compaq didn't want).
Maybe the "theft" wasn't that evident.
IIRC it was never proven that intel "stole" anything.
> At that point, it wasn't so much Alpha but Digital. People noticed
> Palmer breaking up the company piece by piece.
1997 it was quite a bit late to recognize that.
Alpha was about the last part to sell before the rest
went to Compaq.
> From the time when Digital sold its chip manufacturing business, it was
> perfectly normal that Digital would outsource manufacturing of its own
> chips to the remaining manufacturers, whether Intel, IBM or any other.
If they had such a fabless Alpha strategy right from the start,
just as Sun and SGI had with their CPUs,
they might have fared much better economically.
Neil Rieck
[...snip...]
>
> well, collateral damage. HP found itself with two Unices
> (three if one includes Linux) in the portfolio,
> so the stepchild with least marketshare had to go.
> This time it was an advantage that there is no other flavor of VMS.
>
> > The theft of Alpha IP happened before Pentium 3 was released, years
> > before. DEC let it go unchallenged for years before finally using that
> > card to get something from Intel (releive DEC of its chip business which
> > Compaq didn't want).
>
> Maybe the "theft" wasn't that evident.
> IIRC it was never proven that intel "stole" anything.
>
> > At that point, it wasn't so much Alpha but Digital. People noticed
> > Palmer breaking up the company piece by piece.
>
> 1997 it was quite a bit late to recognize that.
> Alpha was about the last part to sell before the rest
> went to Compaq.
>
> > From the time when Digital sold its chip manufacturing business, it was
> > perfectly normal that Digital would outsource manufacturing of its own
> > chips to the remaining manufacturers, whether Intel, IBM or any other.
>
> If they had such a fabless Alpha strategy right from the start,
> just as Sun and SGI had with their CPUs,
> they might have fared much better economically.
When you go back and read documents like this:
http://americanhistory.si.edu/collections/comphist/bell.htm#Transition%20from%20VAX%20to%20Prism
...you get the feeling that various groups of technical people were
pulling DEC in various directions while management was not
coordinating the activities of these very talented people. Statements
like this:
<quote>
Then they (DEC management) killed the Prism project and Mr. Cutler
left. They killed it, but Ken (Olsen) didn’t know that it wasn’t dead.
It was still alive in the semiconductor group and it sprung up as
Alpha.
</quote>
...indicates to me that "either the semiconductor group was not
notified that PRISM was dead" or "the semiconductor group ignored the
order to stop working on it" or "the semiconductor group was so far
along that DEC management told them to finish their current
activities".
Any one of these scenarios means that Alpha was as welcome as an
unplanned birth (at least to some people). This is strange because DEC
manufactured and sold way more Alpha-based equipment than VAX.
NSR
JF Mezei
> ...indicates to me that "either the semiconductor group was not
> notified that PRISM was dead" or "the semiconductor group ignored the
> order to stop working on it" or "the semiconductor group was so far
> along that DEC management told them to finish their current
> activities".
Dead is dead, long live dec is a book that analyses much of what was
wrong at DEC. Fairly depressing read.
From what I remember in that book, it wasn't quite like that.
At the same time, DEC has multiple projects going on, including the VAX
9000, prism and the nvax project (CMOS). Olsen rationalised it and
focused on the 9000. When the 9000 turned out to be an expensive dud
they took thge bits and pieces they had thrown in the trash can and
started the formal nvax and the alpha projects.
The VAX 9000 was a good personification of DEC's problem: Trying to
build mainframes to compete against IBM while ignoring the fact that the
world was moving towards smaller, faster and cheaper machines.
Bob Eager
> Dead is dead, long live dec is a book that analyses much of what was
> wrong at DEC. Fairly depressing read.
Just about to start reading my copy...!
--
Use the BIG mirror service in the UK:
http://www.mirrorservice.org
John Wallace
>
> ...you get the feeling that various groups of technical people were
> pulling DEC in various directions while management was not
> coordinating the activities of these very talented people. Statements
> like this:
>
> <quote>
> Then they (DEC management) killed the Prism project and Mr. Cutler
> left. They killed it, but Ken (Olsen) didn’t know that it wasn’t dead.
> It was still alive in the semiconductor group and it sprung up as
> Alpha.
> </quote>
>
> ...indicates to me that "either the semiconductor group was not
> notified that PRISM was dead" or "the semiconductor group ignored the
> order to stop working on it" or "the semiconductor group was so far
> along that DEC management told them to finish their current
> activities".
>
> Any one of these scenarios means that Alpha was as welcome as an
> unplanned birth (at least to some people). This is strange because DEC
> manufactured and sold way more Alpha-based equipment than VAX.
>
> NSR
With the greatest possible respect to the author, and even bearing in
mind the interview is in 1995 (a couple of years after NT first
emerged) what on earth is this quote about:
"Until Microsoft, I though DEC had the greatest engineering
organization, but Microsoft is substantially better."
VAXman-
{...snip...}
>"Until Microsoft, I though DEC had the greatest engineering
>organization, but Microsoft is substantially better."
What blaspheming fuckwit authored that statement? Albeit non-actionable,
it is absolutely reprehensible to make libelous and defamatory statements
about the dead.
JF Mezei
>>organization, but Microsoft is substantially better."
If it weren't for the first 2 words, the statement would be tolerable
since there is no longer any DEC engineering, so even Microsoft's dismal
engineering would be better.
However, when you consider Microsoft's engineering when Microsoft
started, it was closer to "how to steal other people's ideas/software"
than writing your own. Microsfoot lived for at least a decade without
any true engineering of its own.
I may be biased (not not completely biased, I don't yet have apple
branded socks and underwear :-), I'd say Apple has some great
engineering these days. Extremey innovative industrial designers, and
ability to integrate technologies (and open source software) into a
slick/sleek product.
Unlike DEC, Apple knows how to advertise. And Apple makes it easy to buy
their products. And their web site is far better organised than HP's.
Customer service ? Server got delayed by about a week. They upgraded the
delivery to "express" at no charge AND included a free iPOD nano with
it. Did Digital ever do that ? Nop because Digital didn't speak to small
customers.
Bob Koehler
>
> I couldn't agree more. But whenever a new hardware platform appears,
> it is always *nix/c that gets there first because, as we all know, C
> is not much more than a portable assembler. Is the first port any
> good? Nope, it is usually just a very first step. Regarding your
> comments on MACRO32, I guess I was thinking more of other weirdness
> like BLISS.
In every UNIX port, there's some small architecture dependent code
that has to be worked. No language can change that. If that
weren't so, we'd probably say "UNIX recompile" instead of "UNIX
port".
Bob Koehler
>
> Yes, but since Vista has broken my scanner driver, it may take a while...
Get a Macintosh.
Bob Koehler
>
> C is undoubtedly easy to write OS code in (no surprise since
> it was designed for this type of purpose), but it has proven to
> relative hard to avoid buffer overruns, memory leaks etc.
> (buffer runs are probably more relevant than memory leaks for OS).
BLISS was designed to write OS code in, but it does not have a
reputation for making it easy to write OS code.
Bob Koehler
> Not any platform.
>
> But if there is let us call it flexibility to redesign to match
> the new HW, then I would assume that all HW capable of running
> OS/400, Unix, Linux, Windows NT could be made to run VMS.
VMS has needs that UNIX and Linux do not. I'm not sure of OS/400,
or Windows.
Bob Koehler
>
> Would it have made a big difference?
>
> VMS was not the only multiuser OS.
>
> They could have used ideas from Unix, Multics etc. instead.
Does it make a difference? Compare MacOS X, which has a Mac GUI on
top of a well known UNIX kernel, with Windows, which has it's own
kernel with some small familiarities between IO subsystems with VMS.
Why is it so many of us VMS enthusiasts love those Mac commercials?
Perhaps we find quality and reliability attractive in both the
software we use as well as the software we write?
Bob Eager
No thanks. I actually thought of it recently, in terms of getting a
machine for my son. But 100% mark-up over a non-Mac machine changed my
mind.
Anyway, fixed the scanner now!
I'm very happy with FreeBSD for my everyday work, anyway...having started
with UNIX over 33 years ago...
--
Use the BIG mirror service in the UK:
http://www.mirrorservice.org
Bob Koehler
>
> And my understanding is that VMS inherited this from
> the PDP-11 OS's.
DEC OS's for PDP-11 tended to be written in Macro-11. A little bit
of BLISS-11 alter on was used for portable stuff like EDT. VMS
was initially mostly Macro-32 and BLISS; much of the Macro-32
was eventually replaced. Lots of other languages were also used.
> And when it was decided for them, then C was not available
> (at least according to many C was invented to port Unix
> to PDP-11).
UNIX was ported to PDP-11, using C, about a decade before VMS
was started. So DEC could have written a C compiler for VAX and
large parts of VMS in C if they'd wanted to. DEC didn't write a
C compiler for VAX early on because no one outside of a few UNIX
users were using C. UNIX was the broken down OS that AT&T couldn't
find customers for. The handwriting on the wall showed that there
was clearly no future for C and/or UNIX.
Then AT&T let some kids at Berkley have a copy of the UNIX source.
They ran it on PDP-11, ported it to VAX, added virtual memory, added
TCP/IP, and somehow got others interested in it. Start up vendors
like Sun and Apollo found they could throw together some commodity
hardware, toss BSD UNIX on it much faster than they could write their
own OS, and sell workstations. When the vendors switched to RISC
they blew away the performance of VAXen and people grudgingly learned
to survive using an OS with a late 1960's human interface, writing
code in a Frankenstein language that escaped from the lab, on hardware
that could grind numbers fast and cheap.
Bob Koehler
> Mmm- it is not difficult to avoid buffer overruns or memory leaks or =20
> any other
> problems that a lot of people think are endemic to C. You just have to =20=
>
> know
> what you are doing, and *that* is true no matter what language you use.
Been there, argued that. Why does it keep coming around? Are the
ivory towers still teaching that mantra to the kiddies?
Richard B. Gilbert
>> It was still alive in the semiconductor group and it sprung up as
>> Alpha.
>> </quote>
>>
>> ...indicates to me that "either the semiconductor group was not
>> notified that PRISM was dead" or "the semiconductor group ignored the
>> order to stop working on it" or "the semiconductor group was so far
>> along that DEC management told them to finish their current
>> activities".
>>
>> Any one of these scenarios means that Alpha was as welcome as an
>> unplanned birth (at least to some people). This is strange because DEC
>> manufactured and sold way more Alpha-based equipment than VAX.
>>
>> NSR
>
> With the greatest possible respect to the author, and even bearing in
> mind the interview is in 1995 (a couple of years after NT first
> emerged) what on earth is this quote about:
> "Until Microsoft, I though DEC had the greatest engineering
> organization, but Microsoft is substantially better."
DEC is gone! Microsoft is still here. That says it all!
Bob Eager
>> And when it was decided for them, then C was not available (at least
>> according to many C was invented to port Unix to PDP-11).
>
> UNIX was ported to PDP-11, using C, about a decade before VMS was
> started. So DEC could have written a C compiler for VAX and large
> parts of VMS in C if they'd wanted to. DEC didn't write a C compiler
> for VAX early on because no one outside of a few UNIX users were
> using C. UNIX was the broken down OS that AT&T couldn't find
> customers for. The handwriting on the wall showed that there was
> clearly no future for C and/or UNIX.
>
> Then AT&T let some kids at Berkley have a copy of the UNIX source.
> They ran it on PDP-11, ported it to VAX, added virtual memory, added
> TCP/IP, and somehow got others interested in it. Start up vendors
> like Sun and Apollo found they could throw together some commodity
> hardware, toss BSD UNIX on it much faster than they could write their
> own OS, and sell workstations. When the vendors switched to RISC
> they blew away the performance of VAXen and people grudgingly learned
> to survive using an OS with a late 1960's human interface, writing
> code in a Frankenstein language that escaped from the lab, on
> hardware that could grind numbers fast and cheap.
There are so many inaccuracies there I don't even know where to start...
Bob Eager
Here's a couple, for starters...
> UNIX was ported to PDP-11, using C, about a decade before VMS
> was started.
In 1973. Are you saying that VMS was started in 1983? I think not...
> So DEC could have written a C compiler for VAX and
> large parts of VMS in C if they'd wanted to. DEC didn't write a C
> compiler for VAX early on because no one outside of a few UNIX users
> were using C. UNIX was the broken down OS that AT&T couldn't find
> customers for.
It was a research project for many years. It had to be pried from their
hands to be used outside Bell Labs.
> The handwriting on the wall showed that there was
> clearly no future for C and/or UNIX.
It survives, though.
> Then AT&T let some kids at Berkley have a copy of the UNIX source.
> They ran it on PDP-11, ported it to VAX
AT&T ported it to the VAX before that.
> added virtual memory
AT&T did that first.
> added TCP/IP
There was an ARPA grant for that, I believe.
> and somehow got others interested in it. Start up vendors
> like Sun and Apollo found they could throw together some commodity
> hardware, toss BSD UNIX on it much faster than they could write their
> own OS, and sell workstations.
But they weren't selling BSD at all....it wasn't a commercial product
until much later. They were selling the AT&T version.