The Ultimate Fate of Prime, PRIMOS, CVSI, etc

361 views
Skip to first unread message

Daiyu Hurst

unread,
Nov 27, 2012, 12:14:19 PM11/27/12
to
Google is surely a useful thing!

I spent a couple hours this morning digging around. Most of you may
know that ComputerVision spun off its Prime-related operations in
1997, to CVSI, Inc. Then in 1999, CVSI, Inc., was purchased by 4Front
Technologies, a UK-based "American" company. 4Front Technologies was
in turn acquired by NCR around 2000. At some point, NCR Corporation
transferred some or all of CVSI's rights to Parametric Technology
Corporation. Here's some of that history in detail:

On July 18, 1997, ComputerVision entered into an operating agreement
with CVSI, Inc., and among the details of this agreement, section 5.6
pertains to PRIMOS:

"5.6 CV hereby grants CVSI an exclusive, non-transferable, perpetual,
royalty-
free right and license from the Effective Date to use, copy and
enhance
PRIMOS and associated source code and use, copy, enhance, market,
package,
distribute and license PRIMOS applications and documentation;
provided,
however, that CV and its subsidiaries shall retain any and all rights
to
(i) use, copy and enhance PRIMOS and associated source code,
applications
and documentation for their respective internal operations, and (ii)
use,
copy and enhance PRIMOS and associated source code and use, copy,
enhance,
market, package, distribute and license PRIMOS applications and
documentation for so long as CVSI is in default in its obligations to
provide OSS Services under this Agreement and such default could
reasonably
be expected to result in a material breach of a Customer Contract."

Meanwhile, the "Prime" trademark (US Trademark No. 73123025),
originally filed on 15 April 1977 by Prime Computer, Inc, registered
by them on 9 January 1979, renewed by CVSI, Inc., on 9 January 1999,
and was cancelled on 28 October 2010. The chain of ownership was:

Prime Computer -> CVSI, Inc. -> NCR Corporation -> Parametric
Technology Corporation.

You can see that info here:

http://www.trademarken.com/trademark/73123025?i=PRIME-PARAMETRIC_TECHNOLOGY_CORPORATION_NCR_CORPORATION_CVSI_INC_PRIME_COMPUTER_INC

Similarly, the "PRIMOS" trademark (US Trademark No. 73122880) has been
abandoned. It was originally filed on the same date as above,
registered 20 June 1978, renewed 20 June 1998, except that its
ownership passed directly from CVSI to Parametric, and didn't pass
through NCR Corporation. It expired 28 March 2009.

You can find this info here:

http://www.trademarken.com/trademark/73122880?i=PRIMOS-PARAMETRIC_TECHNOLOGY_CORPORATION_CVSI_INC_PRIME_COMPUTER_INC

and here:

http://www.trademarkia.com/primos-73122880.html


After spinning off CVSI in 1997, ComputerVision was acquired by
Parametric Technology Corporation, in 1998.

Today, CVSI seems to still have a legal existence, but their address,
1700 S. Patterson Drive, Dayton, OH, doesn't exist- it's open
greenspace, and the from satellite pics, that's not a recent
development. The CEO of CVSI, Jonathan S. Hoak, has ties to Teradata
International and of course to NCR.

Parametric is still in business, marketing CAD-related software
products and services. In contrast to the information above, on this
page:

http://www.ptc.com/common/ptccopy.htm

they still list PRIME as a registered trademark. PRIMOS is not listed.

I had contacted CVSI in 2000, inquiring as to the availability of a
license to PRIMOS. I didn't need one really, since by agreement, my
2455 remained the property of Stu Rutkin, and with me operating it
under his license. The CVSI staffer I corresponded with seemed to be
in a hurry to get some cash from me, then suddenly went silent. This
has always seemed very shady to me.

If I recall correctly, ongoing maintenance of PRIMOS had been
contracted out to Peritus, Inc., which is still in business.

The compilers, except for PL/P (written by Jim Cook? of Prime) and
perhaps the original FTN compiler and PMA assembler, had been written
by Language Processors, Inc., which later became Liant, which was
later acquired by MicroFocus, which also bought Ryan/McFarland some
years back.

If anyone can provide corrections or fill in gaps here, please do!

-dai

Bill Gunshannon

unread,
Nov 27, 2012, 12:40:34 PM11/27/12
to
In article <a0c46aba-8a67-4024...@i7g2000pbf.googlegroups.com>,
An intersting missive. I haven't worked with a Prime or PRIMOS since
about the 1986-1987 time frame. I had a small Prime at home for a
while but eventually gave it to the person who was still doing Prime
maintenance at the time.

You didn't mention the "C" compiler which was done by Garth Conboy.

Now, the big question. Is there any chance of PRIMOS and all the
subsidiary pieces being Open Sourced under either a BSD style or
even a GPL style license?

Oh yeah, I guess one more question. What about PRIMIX?

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>

Daiyu Hurst

unread,
Nov 27, 2012, 1:19:16 PM11/27/12
to
On Nov 27, 12:40 pm, billg...@cs.uofs.edu (Bill Gunshannon) wrote:
> In article <a0c46aba-8a67-4024-a375-78c5a5389...@i7g2000pbf.googlegroups.com>,
>         Daiyu Hurst <daiyu.hu...@gmail.com> writes:
..snip..
>
> > If anyone can provide corrections or fill in gaps here, please do!
>
> An intersting missive.  I haven't worked with a Prime or PRIMOS since
> about the 1986-1987 time frame.  I had a small Prime at home for a
> while but eventually gave it to the person who was still doing Prime
> maintenance at the time.

Indeed, I recall your posts from some time back.

> You didn't mention the "C" compiler which was done by Garth Conboy.

Well, I have about as much interest in C as I would a rash in my
nethers.

I have this instinct: when I see everyone running, or even walking, in
one
direction, I turn around and go the opposite direction. I'm somewhat
of an
anti-lemming.

> Now, the big question.  Is there any chance of PRIMOS and all the
> subsidiary pieces being Open Sourced under either a BSD style or
> even a GPL style license?

As soon as someone finds, and then takes possession of the sources,
and assumes ownership of them under the abandoned property laws of
the state in which they reside, then they will be able to do just
as you suggest.

But that's the only way it will happen.

> Oh yeah, I guess one more question.  What about PRIMIX?

Dunno. IIRC, didn't that run on a non 50-series line of boxes?

-dai

Bill Gunshannon

unread,
Nov 27, 2012, 1:43:50 PM11/27/12
to
In article <87333197-dfb5-4350...@qi10g2000pbb.googlegroups.com>,
Daiyu Hurst <daiyu...@gmail.com> writes:
> On Nov 27, 12:40 pm, billg...@cs.uofs.edu (Bill Gunshannon) wrote:
>> In article <a0c46aba-8a67-4024-a375-78c5a5389...@i7g2000pbf.googlegroups.com>,
>>         Daiyu Hurst <daiyu.hu...@gmail.com> writes:
> ..snip..
>>
>> > If anyone can provide corrections or fill in gaps here, please do!
>>
>> An intersting missive.  I haven't worked with a Prime or PRIMOS since
>> about the 1986-1987 time frame.  I had a small Prime at home for a
>> while but eventually gave it to the person who was still doing Prime
>> maintenance at the time.
> Indeed, I recall your posts from some time back.
>> You didn't mention the "C" compiler which was done by Garth Conboy.
> Well, I have about as much interest in C as I would a rash in my
> nethers.
> I have this instinct: when I see everyone running, or even walking, in
> one
> direction, I turn around and go the opposite direction. I'm somewhat
> of an
> anti-lemming.

Funny, that. I hear this about C all the time. One language is the same
as another in general terms. As long as the language is used for what
it was designed for and not for things it is unsuited for I see no problem.
I have never been a language bigot.

>> Now, the big question.  Is there any chance of PRIMOS and all the
>> subsidiary pieces being Open Sourced under either a BSD style or
>> even a GPL style license?
> As soon as someone finds, and then takes possession of the sources,
> and assumes ownership of them under the abandoned property laws of
> the state in which they reside, then they will be able to do just
> as you suggest.

As far as I know, there is no such law. Copyrights will not expire until
long after all of us are dead. What is needed is to determine who the
current holder of those Copyrights are and see if they are willing to do
it. You seem to have traced ownership down to Parametric Technology Corp.
as recently as 2010. Woudl seem likely that they are still the holders
as I can't imagine anyone else buying it. :-)
Wonder if anyone there even knows they are the owners? :-)

> But that's the only way it will happen.
>> Oh yeah, I guess one more question.  What about PRIMIX?
> Dunno. IIRC, didn't that run on a non 50-series line of boxes?

Sure did. It was like EUNICE on VAX/VMS. Ran on top of PRIMOS (very
slowly) and gave what was supposed to be a Unix-like environment. Of
course, porting software from any other Unix system to PRIMIX was a
real challenge. Right of the bat what did Unix think the numeric
value of the letter "A" was? And Prime? :-) I was able to port some
interesting stuff, just the same. Hmmmm... Have to dig out my PRIMIX
documentation. I wnder if Primes could fork()?

matt weber

unread,
Nov 27, 2012, 6:57:47 PM11/27/12
to
Primix was an overlay for Primos to make it look like Unix. It never
worked very well, and Prime/CV de-allocated the resources quite early
in the game. It is what Bob Ollerton used to call a 'flag waver'.

You have a product so you can check off the required boxes on the RFQ,
and wave the flag that you have something,
but make sure it runs so badly that no one in their right mind would
ever use it.

Daiyu Hurst

unread,
Nov 27, 2012, 7:04:18 PM11/27/12
to
On Nov 27, 1:43 pm, billg...@cs.uofs.edu (Bill Gunshannon) wrote:
> In article <87333197-dfb5-4350-aa05-fdf9ba9ce...@qi10g2000pbb.googlegroups.com>,
>         Daiyu Hurst <daiyu.hu...@gmail.com> writes:

> > On Nov 27, 12:40 pm, billg...@cs.uofs.edu (Bill Gunshannon) wrote:
> >> In article <a0c46aba-8a67-4024-a375-78c5a5389...@i7g2000pbf.googlegroups.com>,
> >>         Daiyu Hurst <daiyu.hu...@gmail.com> writes:
> > ..snip..

> > Well, I have about as much interest in C as I would a rash in my nethers.
> > I have this instinct: when I see everyone running, or even walking, in one
> > direction, I turn around and go the opposite direction. I'm somewhat of an
> > anti-lemming.
>
> Funny, that.  I hear this about C all the time.  One language is the same
> as another in general terms.  As long as the language is used for what
> it was designed for and not for things it is unsuited for I see no problem.
>  I have never been a language bigot.

Admittedly my data points don't constitute a random sample, but I've
been
exposed to far more undecipherable code written in C than in any other
language.

It's not bad itself, it just has some quality that tends to attract
the worst
programmers in the world. In fairness, some of the best use it as
well. A good
example of that is the MXLOAD program, written by Olin Siebert, a
utility for
restoring Multics backup tapes to modern systems.

> > As soon as someone finds, and then takes possession of the sources,
> > and assumes ownership of them under the abandoned property laws of
> > the state in which they reside, then they will be able to do just
> > as you suggest.
>
> As far as I know, there is no such law.  Copyrights will not expire until
> long after all of us are dead.  What is needed is to determine who the
> current holder of those Copyrights are and see if they are willing to do
> it.  You seem to have traced ownership down to Parametric Technology Corp.
> as recently as 2010.  Woudl seem likely that they are still the holders
> as I can't imagine anyone else buying it.  :-)

Here's Kentucky's law:

The fundamental principles regarding ownership of abandoned property
are succinctly set forth in 1 Am. Jur. 2nd Abandoned, Lost, and
Unclaimed Property § 24 (2007).

Property which is abandoned becomes subject to appropriation by the
first taker or finder who reduces it to possession. Such person
thereupon acquires absolute ownership in the property abandoned, as
against both the former owner and any person upon whose land it
happens to have been left.

(Citations omitted.)

In other words, [p]ersonal property, upon being abandoned, ceases to
be the property of any person, unless and until it is reduced to
possession with the intent to acquire title to, or ownership of, it.
Such property may, accordingly, be appropriated by anyone, if it has
not been reclaimed by the former owner, and ownership of it vests by
operation of law, in the person first lawfully appropriating it and
reducing it to possession with the intention to become its owner,
provided such taking is fair.

1 C.J.S. Abandonment § 12 (1993). Thus, ownership of abandoned
property vests in the possessor by operation of law and a court need
not resort to its injunctive or equitable powers in order to recognize
that ownership. Although Kelley strongly contests this point, he
offers no legal basis for his contrary position beyond the definitions
of "equitable remedy" and "injunction" contained in Black's Law
Dictionary. Those general definitions are simply insufficient to
contravene established law regarding abandoned property.

> Wonder if anyone there even knows they are the owners?  :-)

Someone could discretely write them and ask them if they knew their
trademarks for "Prime"
and "Primos" were no longer in effect. If I had the money (just a few
hundred for each),
I'd register them myself. Anyway, I would expect the response, if any,
would be along
the lines of "that's for some old stuff we don't know about".

> > But that's the only way it will happen.
> >> Oh yeah, I guess one more question.  What about PRIMIX?
> > Dunno. IIRC, didn't that run on a non 50-series line of boxes?
>
> Sure did.  It was like EUNICE on VAX/VMS.  Ran on top of PRIMOS (very
> slowly) and gave what was supposed to be a Unix-like environment.  Of
> course, porting software from any other Unix system to PRIMIX was a
> real challenge.  Right of the bat what did Unix think the numeric
> value of the letter "A" was?  And Prime?  :-)  I was able to port some
> interesting stuff, just the same. Hmmmm...  Have to dig out my PRIMIX
> documentation.  I wnder if Primes could fork()?

Dunno. What little of *nix I might want on PRIMOS, like TAR and GREP,
I already have. And like you mention, the parity bit thing is not
much of an issue, I was already accustomed to flipping that bit back
and forth doing file transfers back in the 1980s.

-dai

Dennis Boone

unread,
Nov 28, 2012, 10:57:31 AM11/28/12
to
> I wnder if Primes could fork()?

The work done to support PRIMIX made it possible, more or less. PRIMOS
sprouted several new process types, SPAWN$ got some new flags, etc. to
make it happen. We used some of that support to build some PRIMOS tools
for special phantoms.

None of that should imply that it was fast or anything, though I never
had access to PRIMIX to experience it first-hand.

De
Message has been deleted

Dennis Boone

unread,
Nov 28, 2012, 11:35:57 AM11/28/12
to
> The work done to support PRIMIX made it possible, more or less. PRIMOS
> sprouted several new process types, SPAWN$ got some new flags, etc. to
> make it happen. We used some of that support to build some PRIMOS tools
> for special phantoms.

For fun, some code. Here's the entry for SPAWN$ from a Rev 19.4
PRIMOS>KS>SPAWN$.PLP:

spawn$:
proc (xkey, xsptr, xinput_file, xunit, xargs, xuser, xcode)
options (gate, nocopy);

The second argument is a pointer to a structure called sds, which looks
like this:

dcl 1 sds based,
2 version fixed bin, /* Version number of structure */
2 user_id char (32) var,
2 project_id char (32) var,
2 utype fixed bin, /* User type */
2 priority fixed bin, /* Priority level */
2 prvl bit (16), /* Privilege bits */
2 ts fixed bin, /* Time slice */
2 num_grps fixed bin, /* *** Temporary *** number of groups */
2 groups (32) char (32) var; /* *** Temporary *** group names */

The sds.utype field takes one of the following values (well, ok, a
process can have one of these types; actually calling SPAWN$ with a
utype of e.g. u$logo would be pointless at best):

/* User type definitions for UTYPE */

%replace u$logo by - 1, /* Logged out */
u$loip by 0, /* Logout in progress */
u$ltut by 1, /* Low terminal-user type */
u$norm by 1, /* Normal user */
u$trem by 2, /* User gone remote */
u$frem by 3, /* User from remote */
u$thru by 4, /* User logged through */
u$susr by 5, /* Supervisor (SUSR) */
u$tfam by 6, /* Old-style user FAM */
u$htut by 6, /* High terminal-user type */

/* Space for expansion */

u$lput by 65, /* Low phantom-user type */
u$ph by 65, /* Phantom */
u$cph by 66, /* CPL phantom */
u$npx by 67, /* NPX slave */
u$bach by 68, /* Batch phantom */
u$pfam by 69, /* Old-style phantom FAM */
u$net by 70, /* Network process */
u$rts by 71, /* Route through server */
u$fork by 72, /* Unix type Fork phantom */
u$hput by 72; /* High phantom-user type */

Note the u$fork value. Presumably, PRIMIX mapped fork() to a call to
SPAWN$ with sds.utype set to 72.

De

Edward Feustel

unread,
Nov 30, 2012, 9:41:36 AM11/30/12
to
On 27 Nov 2012 18:43:50 GMT, bill...@cs.uofs.edu (Bill Gunshannon)
wrote:
...
>Sure did. It was like EUNICE on VAX/VMS. Ran on top of PRIMOS (very
>slowly) and gave what was supposed to be a Unix-like environment. Of
>course, porting software from any other Unix system to PRIMIX was a
>real challenge. Right of the bat what did Unix think the numeric
>value of the letter "A" was? And Prime? :-) I was able to port some
>interesting stuff, just the same. Hmmmm... Have to dig out my PRIMIX
>documentation. I wnder if Primes could fork()?
>
>bill
The answer is yes. Carl Chan and I filed a patent 4,812,981 on a fork
mechanism for Primix. I don't know if it is still in force. The method
really speeded up the fork, only completing the formal fork if
writing was required.

I do not remember the details -- it was over 20 years ago.
Ed
Message has been deleted

Bill Gunshannon

unread,
Dec 1, 2012, 12:20:54 PM12/1/12
to
In article <sJidncGUlpnwoCvN...@giganews.com>,
But in most cases SPAWN is not the same as fork(). See my previous question.
Message has been deleted

brianh...@gmail.com

unread,
May 20, 2013, 8:19:35 AM5/20/13
to
CVSI was bought by NCR (hence the Dayton address) in 2000. There are 2 people left from cvsi (I was laid off in 2010)
Reply all
Reply to author
Forward
0 new messages