bleeding money out of clispers
Lyn A Headley
I originally posted this question to the lisa mailing list but
since then have been wondering what other possibilities for a
solution, literature pointers etc the other creative people on
comp.lang.lisp might have in response to my problem, so if you
have any input I would be most appreciative.
I am wondering if lisa is appropriate for solving a
problem which i have been thinking about in a general way
for a while and which specifically has manifested
itelf on the clisp mailing list today in the form of financing the MT
project for clisp. Basically I want to offer money to help the
financing, but only if *others* are also willing to offer some money.
A simple variant of the problem can be stated thusly:
given 3 people, A, B, and C, each of which has the following attitude:
"i will contribute 5 dollars if 2 others do."
Could lisa be used to find an answer which assigns a 5-dollar
responsibility to each of the participants? Of course more elaborate
"bidding strategies" are possible and I would like to explore them,
but first things first. any input would be most appreciated.
-Lyn
Sashank Varma
lahe...@cs.uchicago.edu (Lyn A Headley) wrote:
[snip]
>I am wondering if lisa is appropriate for solving a
>problem which i have been thinking about in a general way
>for a while and which specifically has manifested
>itelf on the clisp mailing list today in the form of financing the MT
>project for clisp. Basically I want to offer money to help the
>financing, but only if *others* are also willing to offer some money.
>A simple variant of the problem can be stated thusly:
[snip]
>Could lisa be used to find an answer which assigns a 5-dollar
>responsibility to each of the participants? Of course more elaborate
>"bidding strategies" are possible and I would like to explore them,
>but first things first. any input would be most appreciated.
[snip]
are you looking for actual human beings to help finance an MT (machine
translation?) project in clisp (clips?), or a production system that
solves your bidding problem?
sashank
Russell Senior
Lyn> given 3 people, A, B, and C, each of which has the following
Lyn> attitude: "i will contribute 5 dollars if 2 others do."
I don't understand that calculus. For me it is simple. Do I want to
help fund a project or not? I don't particularly care if others agree
with me. I am fairly certain that in excess of 5 billion people
disagree with me (sorry, I've lost count). The only concern I have is
that whatever money I contribute be handled responsibly. IMHO,
getting concerned about what others are doing or value as a trigger
for your own action is a recipe for stagnation and death.
--
Russell Senior ``The two chiefs turned to each other.
sen...@aracnet.com Bellison uncorked a flood of horrible
profanity, which, translated meant, `This is
extremely unusual.' ''
Chris Double
> Basically I want to offer money to help the financing, but only if
> *others* are also willing to offer some money.
Have you looked at cousource? That is supposed to be a forum for
bidding for payment of open source projects:
Chris.
--
http://www.double.co.nz/cl
Lyn A Headley
> are you looking for actual human beings to help finance an MT (machine
> translation?) project in clisp (clips?), or a production system that
> solves your bidding problem?
>
> sashank
MT is multithreading. but anyway, for the purposes of this thread I am just
looking for an algorithm, method, platform etc which will help me
solve the posted sample problem.
-Lyn
Lyn A Headley
> >>>>> "Lyn" == Lyn A Headley <lahe...@cs.uchicago.edu> writes:
>
> Lyn> given 3 people, A, B, and C, each of which has the following
> Lyn> attitude: "i will contribute 5 dollars if 2 others do."
>
> I don't understand that calculus. For me it is simple. Do I want to
> help fund a project or not? I don't particularly care if others agree
> with me. I am fairly certain that in excess of 5 billion people
> disagree with me (sorry, I've lost count). The only concern I have is
> that whatever money I contribute be handled responsibly. IMHO,
> getting concerned about what others are doing or value as a trigger
> for your own action is a recipe for stagnation and death.
Generally speaking, I hold the opposite view, and I can cite you any
number of situations which disprove your claim. Take unions, for
instance. Five people walking off the job are fired without a second
thought. A workforce leaving at once (or even a majority of it) is a
crushing blow. Demonstrations. A couple of picketers is a laughing
stock, whereas a great number is again a giant force. Presidential
elections. Ever heard the term "throwing your vote away?" In each of
these cases people act based on prior knowledge they have about how
others are going to act. There is nothing wrong with this. Rather,
in my view formalizing the process by providing potential actors with
more information about the dispositions of the involved parties can
increase the potential effectiveness of the planned action and reduce
the possibility of wasted effort.
-Lyn
Russell Senior
Lyn> given 3 people, A, B, and C, each of which has the following
Lyn> attitude: "i will contribute 5 dollars if 2 others do."
Russell> I don't understand that calculus. For me it is simple. Do I
Russell> want to help fund a project or not? I don't particularly
Russell> care if others agree with me. I am fairly certain that in
Russell> excess of 5 billion people disagree with me (sorry, I've lost
Russell> count). The only concern I have is that whatever money I
Russell> contribute be handled responsibly. IMHO, getting concerned
Russell> about what others are doing or value as a trigger for your
Russell> own action is a recipe for stagnation and death.
Lyn> Generally speaking, I hold the opposite view, and I can cite you
Lyn> any number of situations which disprove your claim. Take unions,
Lyn> [...]
Coordinated action is fine. What I find strange is to say: ``I refuse
to do this desirable thing until I find a bunch of other people to do
it with me.'' I don't think anyone will threaten to fire you for
giving some money to help support a project, so looking for herd
protection in this instance seems incongruous. Perhaps by stepping
forward with some cash you can inspire others with your leadership.
Instead, your vocal reluctance is likely (it seems to me) to have the
opposite effect.
FWIW, I have already privately expressed a willingness to contribute a
paltry $100. I did that without a substantial income at the moment
and without resort to a bureaucratic formula.
who...@wherever.com
>Russell> I don't understand that calculus. For me it is simple. Do I
>Russell> want to help fund a project or not? I don't particularly
To understand it, consider what would happen if you agreed to donate $5 to
each of 1000 different projects. If you agreed unconditionally, you would be
contributing $5000, which you might not be able to afford. But if each
project has a one percent probability of getting enough donations to proceed,
then you're probably only contributing around $50. That would be a much more
effective use of your $50 than to just contribute it to a random subset of 10
of those 1000 projects.
That's one factor, and one way to understand it. But there is also the
factor that a project is more likely to get enough conditional pledges to
proceed than to get enough unconditional ones. If people use the reasoning
in the previous paragraph and this one, they're more likely to contribute
conditional money than unconditional. So the conditional money will add up
faster and is more likely to be enough to proceed.
Tim Bradshaw
> MT is multithreading. but anyway, for the purposes of this thread I am just
> looking for an algorithm, method, platform etc which will help me
> solve the posted sample problem.
I think that this kind of problem is some fairly standard game-theory
thing, and it should either have a known solution or be known not to
have one. It's completely symmetrical - everyone is in the same
position of either contributing x or not and has the same thing to win
or lose, so it smells to me like it should be reasonably approachable.
--tim
Erik Naggum
> Basically I want to offer money to help the financing, but only if
> *others* are also willing to offer some money.
I do not understand this attitude. If I want something, I put money into
it. If I do not put enough money into it to make it viable, I would very
much prefer to know this beforehand rather than in the bankruptcy hearing.
Therefore, I require that my money not be _spent_ before there is enough
financing to ensure that it is not wasted. In the absence of sufficient
funds, I expect to get my money back. Responsible financing takes care
of this kind of condition very easily. In fact, I have never _seen_ any
endeavor where people are asked to put money into something to see if it
might fly. Taking people's money on a known losing bet is known as fraud.
My advice is this: Just put up the money. Show your own interest. Lead
the pack. Require a certain amount before the project starts. Prepare
to lose the money nonetheless. (I.e., do not give away money you need
for basic necessities, regardless of the promised returns.) Those who
are not quite as interested or are sitting on the fence waiting for
someone to tell them what to do, will actually follow any random leader
who steps forward. It is a mystifying human trait that some people will
follow just about _any_ leader. But exploit this pack mentality and just
go ahead and _be_ that leader.
Not that I think multithreading for CLISP is a good idea that _anybody_
should be investing in, but it is their money, so feel free. _I_ think
it is a fantastically bad idea and would discourage people from doing it
for the very simple reason that I expect them to make it gratuitously
incompatible with everything else, tout their superiority by virtue of
being different, and thus make it _harder_ to talk about multithreading
and multiprocessing in Common Lisp. Just like most everything else is
"differently-abled" in CLISP, another different thing to take into
account when discouraging people from using CLISP does not help get
better Common Lisp environments. Thank whatever deities might be behind
this all that CMUCL did it mostly right. If anything, help make CMUCL
better and smaller and everything. Better yet, work on _applications_ in
Common Lisp and show that Lisp is viable as a real language, not one of
those language that are only fun making compilers for.
#:Erik
--
Travel is a meat thing.
Erik Naggum
> Generally speaking, I hold the opposite view, and I can cite you any
> number of situations which disprove your claim.
He made a claim that can be disproved??? Amazing.
> Take unions, for instance.
Somebody leads the unions. Somebody orders the members to act a certain
way. However, those members who disobey are punished. _That_ is the
real source of the "formidable force" of labor unions. Union members do
not commit to do something they want only if others agree. They commit
to do whatever whoever manages to grab the leadership tells them to do.
I have always seen labor unions as the ultimate symbol of weakness, in
that if it were _not_ for their numbers, they would be exactly nothing.
I happen to disagree with this misanthropic weltanschauung, however. I
believe that a man acting alone is a force to be reckoned with. Whoever
had the guts to _start_ organizing labor unions proved _that_ point, as
it was indeed true back then that psychopathic bosses showed reckless
disregard for human life and they _had_ to be stopped in numbers. Had it
not been for the mentally ill people who sought power over other people,
labor unions would have been unnecessary. Since such people are always a
danger in any society that rewards rather than punishes those who seek
power, people must be prepared to walk out on any psychopathic manager.
That _includes_ members of labor unions when their leaders go nuts.
Ideally, this "force of numbers" game would lead groups to dissolve when
their goals had been reached. It is when they do not dissolve that they
cling together and show a fear of standing alone even in the absence of
threats.
But all things considered, what the hell have labor unions got to do with
your unwillingness to put up the money for whatever CLISP trick you want
to finance? Just do it, damnit. Who is going to exert unfair and unjust
force over your life if you do it by yourself that would have made this a
likely comparison with labor unions? Sorry, I don't get it.
Erik Naggum
> To understand it, consider what would happen if you agreed to donate $5 to
> each of 1000 different projects. If you agreed unconditionally, you would be
> contributing $5000, which you might not be able to afford. But if each
> project has a one percent probability of getting enough donations to proceed,
> then you're probably only contributing around $50. That would be a much more
> effective use of your $50 than to just contribute it to a random subset of 10
> of those 1000 projects.
What would happen to the money you gave to a project that did not get
enough funding? Would it just "vanish"? The word "FRAUD" keeps flashing
in red neon sign letters in the back of my head when I hear of such
endeavors. _Obviously_, you would get the money back that went to
projects that did not get enough funding. Such a return policy is
inherent in asking people for money. (Frequently, the return is effected
by not actually transferring the money until ons is asked to do so, but
it cannot be spent on other endeavors while thusly committed.)
> That's one factor, and one way to understand it. But there is also the
> factor that a project is more likely to get enough conditional pledges to
> proceed than to get enough unconditional ones.
And those are the only two options you can see? I am frankly amazed.
40,000 years of human beings on this planet, 10,000 years of systems of
money and finance, and _still_ people are unaware of financial systems.
> If people use the reasoning in the previous paragraph and this one,
> they're more likely to contribute conditional money than unconditional.
> So the conditional money will add up faster and is more likely to be
> enough to proceed.
The way you do this is ask people to commit a certain amount of money to
a project _if_ it gets through. That commitment is not "conditional" at
the level most people understand this term. You _must_ commit that money
when the conditions are met. If the conditions are _not_ met, you do not
commit the money. When you commit the money, it is _committed_. It is
out of your hands, and somebody else has the right to ask for them when
the terms of the contract you have (implicitly) entered require it.
Somebody, please investigate the concept of an Initial Public Offering.
That is _exactly_ what you are doing. You see, amazingly, the world has
had this concept for more than 9,000 years. Actual written records of
what we call "shares" exist that are about as old as number systems. The
Sumerians invented numbers to keep track of their money, as far as
history can tell. Options, forwards, futures, etc, were all invented in
extended farming communities to protect against bad weather, pests,
fires, etc, and were well understood by people who were used to take
risks. Insurance is another way to deal with unknown risks that are
known to happen with a low frequency, but to unknown individuals. There
is archeological evidence of "insurance communities" that are more than
6,000 years old. The first known tax systems were based on compulsory
insurance. Wise rulers also collected a certain portion of the seed to
keep a treasury in case the subjects acted unwisely in dire times.
Yet, in 2001, we have people who do not understand enough finance to even
help launch a simple little hobbyist software project. I'm shocked.
I wonder why this Ballmer dude thinks Open Source is such a threat. I
have yet to see a single person who favors Open Source show an ounce of
business sense, financial understanding, or knowledge of marketing. It
is as if they are trying to invent a whole different kind of economy,
which would have been a smart thing to do if the past umpteen thousand
years of human development was all a giant mistake. Peaceful transition
to world communism will happen before _that_ ever makes sense. *Sigh*
Craig Brozefsky
> Somebody, please investigate the concept of an Initial Public Offering.
> That is _exactly_ what you are doing.
It's probably closer to the Street Performer Protocol than an IPO, both
in purpose and form.
> treasury in case the subjects acted unwisely in dire times.
Yah, that's what it was 8)
> I wonder why this Ballmer dude thinks Open Source is such a
> threat. I have yet to see a single person who favors Open Source
> show an ounce of business sense, financial understanding, or
> knowledge of marketing.
I guess Tim O'reilly and Mike Teimann are just lucky slobs who tripped
into their money.
--
Craig Brozefsky <cr...@red-bean.com>
http://www.red-bean.com/~craig
"Indifference is the dead weight of history." -- Antonio Gramsci
Craig Brozefsky
> * Lyn A Headley
> > Basically I want to offer money to help the financing, but only if
> > *others* are also willing to offer some money.
>
> I do not understand this attitude. If I want something, I put money into
> it. If I do not put enough money into it to make it viable, I would very
> much prefer to know this beforehand rather than in the bankruptcy hearing.
> Therefore, I require that my money not be _spent_ before there is enough
> financing to ensure that it is not wasted. In the absence of sufficient
> funds, I expect to get my money back. Responsible financing takes care
> of this kind of condition very easily. In fact, I have never _seen_ any
> endeavor where people are asked to put money into something to see if it
> might fly. Taking people's money on a known losing bet is known as fraud.
I believe what Lyn is asking for is an informal "escrow" system that
is based on pledges as opposed to the actual holding of funds. The
projects are quite small, and the players on both sides are not
established, and do not have expertise in running businesses
themselves or doing accounting. The use of an informal escrow system
like this is a simple way for a small community to coordinate small
investments. It's akin to you and your freinds pledging 5 bux for a
pizza provided that you got enough people at the party pledging to buy
a sufficient volume. In this case the community is large enough that
a roll-call on a mailing list is no likely to suffice, and the
pledging period could go on for quite some time.
> Not that I think multithreading for CLISP is a good idea that
> _anybody_ should be investing in, but it is their money, so feel
> free. _I_ think it is a fantastically bad idea and would
> discourage people from doing it for the very simple reason that I
> expect them to make it gratuitously incompatible with everything
> else, tout their superiority by virtue of being different, and
> thus make it _harder_ to talk about multithreading and
> multiprocessing in Common Lisp.
What are your grounds for such an opinion?
> help make CMUCL better and smaller and everything. Better yet,
> work on _applications_ in Common Lisp and show that Lisp is viable
> as a real language, not one of those language that are only fun
> making compilers for.
Lyn does that as his day job already.
Paolo Amoroso
> for the very simple reason that I expect them to make it gratuitously
> incompatible with everything else, tout their superiority by virtue of
> being different, and thus make it _harder_ to talk about multithreading
> and multiprocessing in Common Lisp. Just like most everything else is
The job description:
http://sourceforge.net/people/viewjob.php?group_id=1355&job_id=3573
includes some information on the MT interface required for CLISP.
Paolo
--
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://cvs2.cons.org:8000/cmucl/doc/EncyCMUCLopedia/
Erik Naggum
> It's akin to you and your freinds pledging 5 bux for a pizza provided
> that you got enough people at the party pledging to buy a sufficient
> volume.
If you think about it, you just gave me a simplistic version of the
example I gave you that you probably did not understand. I might quip
that this means that I deal with financial instruments and you deal with
pizza at parties, but that would probably make see even redder, right?
Erik Naggum
> > treasury in case the subjects acted unwisely in dire times.
>
> Yah, that's what it was 8)
Is this msianthropic cynicism the result of being at the rule-following
end of the rules for too long? Mankind would not have survived at all
without some people who knew better and who avoided problems caused by
rioting masses. Democracy could not work when people were irrational and
self-destructive to boot. That it (barely) works today is a testament to
the incredible heights to which mankind can rise when the idiots are not
allowed to keep the rest back. Unfortunately, that is what democracy may
be doing when the climax of civilization has been reached. But that is
another kind of misanthropic cynicism.
> I guess Tim O'reilly and Mike Teimann are just lucky slobs who tripped
> into their money.
Well, most business leaders do not actually understand finance, either.
That is why we need specialists. However, most people do not need to
worry about the financial problems of rich people, but they still need to
worry about credit cards, stock prices, food prices, etc. A little
finance comes a long way to dealing with the fluctuations of the economy.
Craig Brozefsky
No, I know it was a simplistic version of your example. My point was
not that it was structurally different, only that the resources of the
players and the risks involved (and therefor formalities required)
where different.
It's true tho that I deal with pizza at parties more than I deal with
these financial instruments.
Lyn A Headley
> But all things considered, what the hell have labor unions got to do with
> your unwillingness to put up the money for whatever CLISP trick you want
> to finance? Just do it, damnit. Who is going to exert unfair and unjust
> force over your life if you do it by yourself that would have made this a
> likely comparison with labor unions? Sorry, I don't get it.
Russell Senior objected to the _general_ notion that acting based on
knowledge of whether others would be acting in concert with you is
to be encouraged. The thing about labor unions was a defense of the
general notion.
It's a bit harder for me to defend my conditional giving to the MT
project. As you say, I should certainly expect that my money will
not be spent unless there are enough funds built up for the project
to succeed. I honestly believe though, that my conditional offer will
make others more likely to contribute than a non-conditional one, since
the value of whatever pledge they might make is now augmented by the
value of my pledge, and the effect of their not pledging not only results
in the project not getting their money, it deprives the project of my
money as well.
-Lyn
Russell Senior
Lyn> Russell Senior objected to the _general_ notion that acting based
Lyn> on knowledge of whether others would be acting in concert with
Lyn> you is to be encouraged.
I think it would be fairer to say that I objected to the general
notion of _not_ acting when you aren't assured in advance that others
will act in concert. I think there is far too much _inaction_ these
days. I think you are more likely to get concerted action when
someone steps boldly and provides some leadership.
Lyn> [...] I honestly believe though, that my conditional offer will
Lyn> make others more likely to contribute than a non-conditional one,
Lyn> since the value of whatever pledge they might make is now
Lyn> augmented [...]
Just to reiterate, I think you are wrong here. What I perceive in
this is a reluctance to contribute, a reluctance that I think is
psychological poison.
I think you would do much more for the chances of success of the
project by _actually_ narrowing the gap between funds collected and
the goal, independent of what others choose to do or not do.
Erik Naggum
> It's a bit harder for me to defend my conditional giving to the MT
> project. As you say, I should certainly expect that my money will
> not be spent unless there are enough funds built up for the project
> to succeed. I honestly believe though, that my conditional offer will
> make others more likely to contribute than a non-conditional one, since
> the value of whatever pledge they might make is now augmented by the
> value of my pledge, and the effect of their not pledging not only results
> in the project not getting their money, it deprives the project of my
> money as well.
For what it might be worth, when I hear somebody make a conditional
commitment to something, I also hear strong doubts about its success.
When I hear someone asking for X amount of money in order to keep
something going (in the case of the only leftist newspaper in this
country) or to start something, I may want to commit as much as I can
spare in the expectation that if the whole thing is overcommitted, I
receive notification of such overcommitment and do not have to commit all
that I was prepared to. If, however, somebody gives me the feeling that
he will only commit money if everybody else fail to commit sufficient
funds, I feel like I would give him all that money. So to me, there is a
very sharp distinction in impression left by someone who commits money
before or after the condition is set. If you commit the money and want
it back if not enough funds are secured, that is fine with me. If you
tell me that you withhold funds until the condition is met, I actually
interpret that to mean that I should not commit at this time, either.
I also believe that marketing a project is about communicating a sense of
trust in its success. Nothing says that better than having put up some
real money of your own. Of course, you could be lying, but you could
also be lying about putting up your share when others have put up theirs.
Given the general lack of knowledge of other people's trustworthiness, I
am more likely to trust someone who (says he) has already done what he
would like others to do, than he who asks others to do it first.
Now, more useful than MT in CLISP in my view is support for multiple
listeners abd background streams in ILISP. So far, Allegro CL has such
an advantage over the others in terms of multiprocessing by virtue of
this feature in the Emacs-Lisp Interface alone that they have no real
competition. That would be doing something useful with the available
multiprocessing support in existing Common Lisp implementations.
And how about writing up some specification-quality material for this
that could go into some sort of standard? That is also sorely needed.
Lars Lundback
wrote:
>
>The job description:
>
> http://sourceforge.net/people/viewjob.php?group_id=1355&job_id=3573
>
>includes some information on the MT interface required for CLISP.
That page led me to the clisp doc page:
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/~checkout~/clisp/clisp/doc/
Some of the listed documents (by sds - Sam Steingold) have the .xml
extension. The top tag in the source says <?xml version="1.0"
encoding="UTF-8"?>
Is the content and layout such that xml is required? Only MS IE 4.0
is available on the various PC-s at this site, so they are in effect
unreadable. I got curious about the GC overview, and may look at it
from home, but it doesn't really matter; CLISP is about the only
Common Lisp system I have abstained from.
But since general information about CLISP is published as
xml-formatted documents I ask sds:
why, oh why do you do this to us?
/Lars
Lars Bjønnes
> That page led me to the clisp doc page:
>
> http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/~checkout~/clisp/clisp/doc/
>
> Some of the listed documents (by sds - Sam Steingold) have the .xml
> extension. The top tag in the source says <?xml version="1.0"
> encoding="UTF-8"?>
> Is the content and layout such that xml is required? Only MS IE 4.0
> is available on the various PC-s at this site, so they are in effect
> unreadable. I got curious about the GC overview, and may look at it
> from home, but it doesn't really matter; CLISP is about the only
> Common Lisp system I have abstained from.
Well, you're browsing the _source_ in the CVS. It seems like sds is
using DocBook (or something based on DocBook) to write the
documentation. You could probably download the files and use a DSSSL
or XSL stylesheet to transform it something more appropriate, like PDF
or HTML. Or the clispers could transform the current documentation and
put it under the docs-section.
--
Lars
Lars Lundback
<lars.b...@fredrikstad.online.no> wrote:
>
>Well, you're browsing the _source_ in the CVS. It seems like sds is
>using DocBook (or something based on DocBook) to write the
>documentation. You could probably download the files and use a DSSSL
>or XSL stylesheet to transform it something more appropriate, like PDF
>or HTML. Or the clispers could transform the current documentation and
>put it under the docs-section.
You mean they should _sink_the _source_ :-)
Further in the Clisp CVS, you find:
"
CLISP implementation notes, converted to DocBook (SGML/XML) by Sam.
impnotes.xml.in -- template for XML (xp, xt) processing
impnotes.sgml.in -- template for SGML (nsgmls, jade) processing
impent.xml -- entities (= shortcuts) for the other files
imptoc.xml -- introduction and table of contents
impbody.xml -- body of notes, parallel to CLHS
impissue.xml -- issue index
impext.xml -- extensions
impbib.xml -- bibliography
"
I guess that the incredibly long, already existing "Clisp
implementation Notes" (available in html) with all the exceptions and
deviations, forced that cry out of me. Of course an author is free to
select his documentation tools.
Sam Steingold
> * On the subject of "Re: bleeding money out of clispers"
> * Sent on Tue, 12 Jun 2001 11:54:15 GMT
> * Honorable lars.r....@telia.com (Lars Lundback) writes:
>
> That page led me to the clisp doc page:
>
> http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/~checkout~/clisp/clisp/doc/
this is not the CLISP doc page - this is the CVS repo over WWW.
do you actually _read_ the URL you cut and paste?
the CLISP page is http://clisp.cons.org/
> But since general information about CLISP is published as
> xml-formatted documents I ask sds:
> why, oh why do you do this to us?
CLISP general information is published in HTML on the aforementioned web
site. In particular, the XML files you found in the CVS are compiled
into http://clisp.cons.org/impnotes.html via DocBook XSL.
This file, impnotes.html is present in all source and binary
distributions.
Your attitude is as if you got a source distribution and complained that
it did not contain the executable for your platform, and you tried to
run the C files and got errors like "this file is not executable".
--
Sam Steingold (http://www.podval.org/~sds)
Fighting for peace is like screwing for virginity.
Lars Lundback
>
>Your attitude is as if you got a source distribution and complained that
>it did not contain the executable for your platform, and you tried to
>run the C files and got errors like "this file is not executable".
I stand corrected - I was careless and did not examine the URLs
mentioned before using them, and so plead guilty to not noticing that
the documents list was on a CVS page. See my previous response.
/Lars
Simon András
> Now, more useful than MT in CLISP in my view is support for multiple
> listeners abd background streams in ILISP. So far, Allegro CL has such
> an advantage over the others in terms of multiprocessing by virtue of
> this feature in the Emacs-Lisp Interface alone that they have no real
> competition. That would be doing something useful with the available
> multiprocessing support in existing Common Lisp implementations.
Wouldn't implementing (perhaps a subset of) LEP in CMUCL be even more
useful? I don't know if it's feasible, given that LEP is undocumented,
but it would be nice to have the same emacs interface (eli, with its
MP support) to ACL and CMUCL.
[No, I have nothing against ILISP, I depend on it and got to like it
too, but I still prefer eli.]
Andras
Bulent Murtezaoglu
[...]
SA> Wouldn't implementing (perhaps a subset of) LEP in CMUCL be
SA> even more useful? I don't know if it's feasible, given that
SA> LEP is undocumented, but it would be nice to have the same
SA> emacs interface (eli, with its MP support) to ACL and CMUCL. [...]
Eli would be my preference also. Actually Franz's licence on the
elisp side appears liberal enough that one could conceivably extract
the details of the le protocol from the elisp code and use eli as is
with cmucl. My worry would be that it might cause headaches for
Franz when 'enhanced' versions of the elsip code start floating
around and they lose the flexibility of conrolling both sides of the
protocol. (We like our vendors here.)
BM
Mike McDonald
I don't actually see a separate license for the emacs interface. At least
not with the 4.3 version, which is what I use when interfacing to ACL or
CMUCL. If I could find such a liberal license, I'd post my hack version of the
CMUCL side.
Mike McDonald
mik...@mikemac.com
Simon András
mik...@mikemac.com (Mike McDonald) writes:
> I don't actually see a separate license for the emacs interface. [...]
I wouldn't have suggested using CMUCL with eli if I hadn't noticed the
following in
http://www.franz.com/support/documentation/6.0/doc/eli.htm#licensing-2
Users of the Emacs-Lisp interface are bound by the GNU Emacs copyright
agreement. Note that all of the Emacs code may be redistributed.
[...]
Files whose type is `el' may be distributed with GNU Emacs under the
terms of the GNU Emacs license agreement. Files without the extension
`el' are not covered by the GNU Emacs License agreement and use is
restricted according to the copyright notice above.
and
;; Permission is granted to any individual or institution to use, copy,
;; modify, and distribute this software, provided that this complete
;; copyright and permission notice is maintained, intact, in all copies and
;; supporting documentation.
in the .el files.
> If I could find such a liberal license, I'd post my hack version of the
> CMUCL side.
If you find this license liberal enough, I'd love to try your code!
Andras
> Mike McDonald
> mik...@mikemac.com
Mike McDonald
asi...@math.bme.hu (Simon András) writes:
>> If I could find such a liberal license, I'd post my hack version of the
>> CMUCL side.
>
> If you find this license liberal enough, I'd love to try your code!
Looks like I need to upgrade to 6.0 from 4.3 for the eli code. When I verify
that my hacks still "work", I'll post them here.
Thanks!
Mike McDonald
mik...@mikemac.com
Simon András
>
> Looks like I need to upgrade to 6.0 from 4.3 for the eli code. When I verify
> that my hacks still "work", I'll post them here.
Great!
Have you implemented (a subset of) LEP in CMUCL, or tweaked the eli
side?
Andras