Is it in the public domain? If not, do you think whoever owns could be
persuaded to release it?
--
*********************************************
Semi-encrypted email address: m_b_p_o_m_i_j_e_ a_t_ i_n_a_v_ d_o_t_
n_e_t_
All non-solicited commercial email will be billed $1,000.
From: Emergent Technologies
Subject: Re: Who owns the LMI source code?
Date:
Message-ID: <u200ln79y.fsf@siteseer.net>
"Martin B. Pomije" <········@inav.net> writes:
> Is it [LMI source code] in the public domain?
No, but....
First let me say that I'm not a lawyer, so this could be complete
garbage. On the other hand, I have actually done *some* research on
this question.
Not all of the intellectual property used by LMI was owned by LMI, for
example, the elements of the source code derived from MIT are
copyright MIT.
Texas Instruments may have some interest in the source code. That
would be covered by the technology transfer contract.
The original work done at LMI was purchased by Gigamos. Gigamos
went bankrupt, and Coopers and Lybrand became the trustees. Coopers
and Lybrand wouldn't ``own'' the code, but would have been required to
sell the assets on behalf of the investors. However, as far as I
know, no one came forward to purchase any part of Gigamos, and the
material assets were sold for scrap.
In theory, the intellectual property would remain the property of
Gigamos and be under the control of Coopers and Lybrand, but when a
corporation goes bankrupt, and the assets are sold, the corporation is
typically dissolved. While I haven't verified this, Gigamos should
have been dissolved by now (LMI was dissolved several years ago.)
The question of ``Who owns the copyright?'' can be rephrased more
pragmatically:
Who would be likely to sue me if I found a copy and used it?
Who would be likely to win such a suit?
and
What would the penalty be if I lost?
The key might be the last question. Liability for copyright
infringement is usually based upon damages. Since no one came forth
to purchase the LMI/Gigamos code for several years, it's pretty clear
that the market value of the LMI code is nil. Since no one has
produced any product whatsoever based upon the LMI lisp machine in
almost a decade, the lost revenue is also nil.
Anyone wishing to claim copyright infringement would have to
demonstrate ownership of the code. This is most likely a costly
proposition.
> If not, do you think
> whoever owns could be persuaded to release it?
I doubt that determining who owns it would be easy or cheap, and I
sincerely doubt they have a copy. The last I saw of LMI/Gigamos was
at Eli Hefron and Sons in Cambridge. They disposed of the material
assets, including the media that contained the source code.
Hope this helps. Out of curiosity, what's your interest?
In article <·············@siteseer.net>,
Emergent Technologies <··········@siteseer.net> writes:
> "Martin B. Pomije" <········@inav.net> writes:
>
>> Is it [LMI source code] in the public domain?
>
> No, but....
>
> First let me say that I'm not a lawyer, so this could be complete
> garbage. On the other hand, I have actually done *some* research on
> this question.
>
> Not all of the intellectual property used by LMI was owned by LMI, for
> example, the elements of the source code derived from MIT are
> copyright MIT.
...
> The question of ``Who owns the copyright?'' can be rephrased more
> pragmatically:
>
> Who would be likely to sue me if I found a copy and used it?
> Who would be likely to win such a suit?
> and
> What would the penalty be if I lost?
Well, since MIT still owns the copyrights on the basis of LMI software and
they do still exist and they're very protective of their property, you might
run a foul of their lawyers. Also, since Symbolics is still in business and
the LMI code came from the same base, I'd think some lawyer would be able to
equate some value to the code, leaving you liable for damages.
> Hope this helps. Out of curiosity, what's your interest?
I curious too!
Mike McDonald
·······@sgi.com
From: Emergent Technologies
Subject: Re: Who owns the LMI source code?
Date:
Message-ID: <u67pw667t.fsf@cape.com>
·······@SGI.com (Mike McDonald) writes:
> Well, since MIT still owns the copyrights on the basis of LMI software and
> they do still exist and they're very protective of their property, you might
> run a foul of their lawyers. Also, since Symbolics is still in business and
> the LMI code came from the same base, I'd think some lawyer would be able to
> equate some value to the code, leaving you liable for damages.
If I remember correctly, there was a lawsuit concerning the Macsyma
code that touches on this sort of issue. I'm sure GJC knows the details.
Obviously you'd have to check out the MIT copyright. Some MIT
software is quite liberal for copyright. I agree that Symbolics might
sue over parts of the code.
There is a fair amount of code that is original to LMI (like the GC
and lexical closures). The LMI Lambda was binary compatable with the
CADR at the microcode level, but the 3600 is not. There is probably
less Symbolics code based on MIT code than there was LMI code, but
that doesn't mean that you wouldn't get sued.
From: Martin B. Pomije
Subject: Re: Who owns the LMI source code?
Date:
Message-ID: <346D0A8A.7089C3C0@inav.net>
Mike McDonald wrote:
(snip)
>
> Well, since MIT still owns the copyrights on the basis of LMI software and
> they do still exist and they're very protective of their property, you might
> run a foul of their lawyers. Also, since Symbolics is still in business and
> the LMI code came from the same base, I'd think some lawyer would be able to
> equate some value to the code, leaving you liable for damages.
>
> > Hope this helps. Out of curiosity, what's your interest?
>
> I curious too!
>
> Mike McDonald
> ·······@sgi.com
Well, my interest was in whether this code could be used as the basis of
a new Lisp based OS, but it sounds like it can't. Darn!
--
*********************************************
Semi-encrypted email address: m_b_p_o_m_i_j_e_ a_t_ i_n_a_v_ d_o_t_
n_e_t_
All non-solicited commercial email will be billed $1,000.
> Well, my interest was in whether this code could be used as the basis of
> a new Lisp based OS, but it sounds like it can't. Darn!
There probably isn't a whole lot from it that would be useful, actually.
It would be faster and better to rewrite anything that it provides.
It would be a good source to look at to see approaches and algorithms,
and perhaps pick out a few little sub-parts.
Copyright issues are not a problem from this useage.
The TI Explorer came with all the source code back when.
Someone must have the tapes around somewhere.
Perhaps someone at UMASS
(which had the second largest TI Explorer LispM installation
in the world if I remember) has a copy they could put on the web?
-kelly
Date: Sat, 15 Nov 1997 17:35:24 -0800
From: Jamie Zawinski <···@netscape.com>
Would that be legal? I doubt TI has given up their copyright on the
version they distributed...
And a lot of it contains MIT copyrights, too. (I just checked some
random files in my copy of Release 6.) Interesting situation: the
product group within TI that handled the Explorer was sold to HP a while
back. Does the copyright move with the group or stay with the company?
Speaking of which, if anyone has any Explorer memory boards they'd
like to sell, let me know.
Got a couple of broken ones. How good are you at surface-mount repairs?
Why hasn't anyone written an Explorer emulator yet?
Believe me, I've been thinking about it, especially as I get back into
the simulation business and learn more about fast simulation techniques.
I mean, at the macroinstruction level an Explorer is just a 32-bit stack
machine with 16-bit instructions. The hard part is figuring out how
much of the underlying nastiness to emulate -- most of GC, some of
scheduling, and all of VM is in microcode. It does sound like a fun
project, though.
Paul Fuqua
Texas Instruments, Dallas, Texas ··@hc.ti.com
In article <··············@elissa.hc.ti.com>,
Paul Fuqua <··@elissa.hc.ti.com> writes:
> Date: Sat, 15 Nov 1997 17:35:24 -0800
> From: Jamie Zawinski <···@netscape.com>
> Why hasn't anyone written an Explorer emulator yet?
>
> Believe me, I've been thinking about it, especially as I get back into
> the simulation business and learn more about fast simulation techniques.
> I mean, at the macroinstruction level an Explorer is just a 32-bit stack
> machine with 16-bit instructions. The hard part is figuring out how
> much of the underlying nastiness to emulate -- most of GC, some of
> scheduling, and all of VM is in microcode. It does sound like a fun
> project, though.
>
> Paul Fuqua
> Texas Instruments, Dallas, Texas ··@hc.ti.com
IF you had all of the sources for the LMIs, wouldn't a better approach be to
retarget the compiler for your Unix machine's processor and then rebuild a Unix
world instead of trying to emulate the LMI microcode? Say, make the
instructions native 32 bit instead of emulated 16 bit? You'd still need the
equivilant of the microcode routines but you wouldn't be so constrained with
how they were implemented. Heck, you might even be able to write some of them
in C. (I'd prefer C to microcode for much of anything.)
A couple of years ago I had thought about emulating the Symbolics 3600 series
but the 36 bit words were too daunting. Also, Symbolics still exists and
they'd get pissed if I did. :-)
Mike McDonald
·······@sgi.com
From: Martin B. Pomije
Subject: Re: Who owns the LMI source code?
Date:
Message-ID: <346D08F7.25F1F97E@inav.net>
Mike McDonald wrote:
(snip)
>
> Well, since MIT still owns the copyrights on the basis of LMI software and
> they do still exist and they're very protective of their property, you might
> run a foul of their lawyers. Also, since Symbolics is still in business and
> the LMI code came from the same base, I'd think some lawyer would be able to
> equate some value to the code, leaving you liable for damages.
>
> > Hope this helps. Out of curiosity, what's your interest?
>
> I curious too!
>
> Mike McDonald
> ·······@sgi.com
Well, my interest was in whether this code could be used as the basis of
a new Lisp based OS, but it sounds like it can't. Darn!
--
*********************************************
Semi-encrypted email address: m_b_p_o_m_i_j_e_ a_t_ i_n_a_v_ d_o_t_
n_e_t_
All non-solicited commercial email will be billed $1,000.