- getting a free[1], working, portable DEFSYSTEM.
- adding extensions to CLX. At least big-request and shape extensions
  should go in.
- working on _any_ of the GUI's: CLUE, Garnet. I don't care which one,
  at the moment I'm trying to package up Garnet, but I wonder why, as I
  don't see a community that uses it. I would like that a uniform GUI
  based environment would suddenly appear. Get Hemlock to work with CLUE
  for instance, and get CLUE to _finally_ use normal windows. <rant>The CLX
  default windows have no place anymore on a modern desktop.</rant>
- Writing better bignum routines. You don't have to hack CMUCL for this,
  just define your own bignum-tng type and write and debug all the code.
  Integrating debugged, working code is not so difficult. (I hope) And
  this is a better use of the few people who can integrate this into
  CMUCL.
- Updating and integrating some of the PD code that is "out there". A
  lot of this stuff just needs minimal cleaning up, or better
  documentation. (e.g. Chess In Lisp. It works, it's nice and useful as
  an example)
- Updating the documentation.
- Reporting bugs (no kidding, people sometimes complain to me that CMUCL
  stinks, but they don't report it!)

Only recently I heard that there is a lisp-based web-browser "out
there". Don't you think a browser would look better as an example
program than a program that only uses "format" and "read" (and most of
the time even SHOUTS BACK :-))

Groetjes, Peter

[1]: See http://www.debian.org.

:   tb> What I meant was that it's possible to share some static parts
:   tb> of the heap, either by lazy copying (the good way) or by the
:   tb> emacs trick.

: What implementations are able to do this?

Poplog (which includes a CL implementation) does this.

: I assume that this implies
: storing a part of the heap in text pages of the executable,

As far as I understand, Poplog does not work in this way, but the heap
images can be tree structured.

On OSes which support shared memory, the part of the heap which is in
common is shared.

I'm sure comp.lang.pop folks may add further details.

On the contrary, it seems to me that those who claim an entitlement are
the hoarders who will sue people on the grounds of copyrights and patents.
Note that keeping source code secret (not publishing) is indeed a right,
but that it is not quite the same as forbidding anyone to arbitrarily use
software they have as they please (including running it, copying it,
redistributing it, reverse-engineering it, and modifying it).

You may well disagree with arguments in favor of free software, but please
admit that there is more to it than just A PRIORI ideological belief.

On the *very* long run, of course, barriers will fall,
just because it is the common interest,
and then, the technically superior solutions may prevail
(whether it is still LISP, or something else).
But it is our responsibility to contribute to good things happening
within our life-time.

Again, I will gladly accept criticism on my article.
        "Metaprogramming and Free Availability of Sources"
        http://www.tunes.org/~fare/articles/ll99/index.en.html

Please help me understand better if I'm mistaken.

Respectfully,

[ "Faré" | VN: Уng-Vû Bân | Join the TUNES project!   http://www.tunes.org/  ]
[ FR: François-René Rideau | TUNES is a Useful, Nevertheless Expedient System ]
[ Reflection&Cybernethics  | Project for  a Free Reflective  Computing System ]
Free Market is not the end of every large-scale economical problem;
but it's the beginning to any long-term solution to anyone of them.

This is a general problem of thoughtful movements.  After the founder
dies people lose track of why it started and a "good and thoughtful idea"
becomes "dogma".  Choice is replaced with non-choice, and the ability to
respond to changing circumstance is removed.  This is the problem with
organized religion today, for example.  It started out well-intentioned
and has drifted because it has lost track of the problems it was trying to
solve or even the fact that it was a problem-solving device.

You have obviously though hard about this issue and I won't try to
dissuade you from your personal theory.  But it seems to me it remains
a theory.  It is possible to disagree.  I have also thought hard about
this and I am trying to stake out a right to a counterposition in the
same space.

My position is based on the classical construction of copyright law in the
first place, which is that it exists to assure people will share.  The
fundamental non-legislatable right is the right not to speak.  Weak copyright
law is not about sharing ideas freely, it's about "sharing freely those ideas
spoken by people who understand that speaking is equivalent to `giving away'
or by people who don't understand the effect of speaking".  That set of
people does not exhaust the space of possible speakers, and the fact is that
under weak copyright law, some would not speak, and that would be a loss.

Strong copyright law is clearly correct in the trivial sense that
using it you can implement weak copyright law.  Anyone who wants to
give away their ideas can do so.  It is the right to control
information that allows one person to implement powerful systems like
GPL.  I choose not to subscribe to GPL as a creation vehicle for my
own thoughts, but strongly copyright law gives me the ability to
implement an alternative.  Weak copyright law just gives big companies
the ability to roll over me because they can take what I use and produce
product and compete with me on marketing, which I can't afford and they can.
Weak copyright law if it ever comes will be the death of the "little guy".

I do think it's an error that patent law protects against independent creation.
That seems nutty and requires ESP to know you're not duplicating someone
else's process.  I won't discuss patent and copyright in the same discussion.

I also think code copyright is too long.  Three years seems adequate.
And I think ownership should revert to the people who did the actual work
in the case of disuse (the legal doctrine of "abandonment", I believe).
But I think text copyright being a fair fraction of the life of the author
is appropriate for the case of books and it will be a disaster if that
ever changes.

I can tell you for a fact that I would never post to this newsgroup or the
web at the rate that I do if posting meant giving up my rights.  You would
have the right to receive my writings privately at great cost under specific
contract not to reveal my writings and that would be it.  An elite subculture
that never spread.  Too bad.  It is exactly strong copyright that leaves
me free to post first and think about whether to formally publish what I
write second because my posting is not an automatic grant of license to
someone else to use my words.  I post to share my ideas, not my words.
Ideas are not subject to copyright.

Copyright is an incentive to do great works and share them with the public
because it separates the notion of "sharing" from the notion of "giving
up right".  "free software" is a step backward because it ties up the two
concepts and assures that anything one doesn't want to give up rights to
will also not be shared.

My purpose in my remarks before was not to rehash this old tired
issue, but to say that for SOME (not ALL) people there is a reasonable
concern that SOME (not ALL) other people want "free software" just
because they see it as an entitlement and they are insensitive to the
need of specific extant companies to stay in business while they
migrate to a future path.  I am not a person who is insensitive to
these issues.  I think there are such people, though I could be wrong.
I do not think all people are of that kind.  I do think that those
people who think it's generally good to give away free software have
not done their duty to explain in dollar-for-dollar spreadsheet form
why Harlequin and Franz should just give away software they are now
selling.  Absent such an explanation, I completely understand a
hesitation to proceed.  It looks dicey to me.  An interesting claim,
nothing more.  And nearly all good ideas in business risk accidentally
crossing the $0 working capital line, so you can't just blunder forward.
My remarks were/are intended to provide a sense of my feeling of where
there might be a compromise between free(er) software and continued cash
flow that might help navigate that delicate line.  I also don't have
spreadsheets.  But then, I'm not advocating quite as extreme a point of
view.

I've always felt that Borland way back when had the right approach.
They were selling nice fast Pascal compilers for $50 when everyone
else was selling theirs for $500.  After a while, I didn't know anyone
without Borland Pascal.  I think this also greatly reduced losses due
to piracy.  I was certainly willing to pay $50 for the disks and
manuals.  If I had pirated it, then I'd have to go out and get one of
the How-To books for $50 anyway.

But perhaps this won't work for Lisp?  I have no idea of the cost
structure.  I also note that it's pretty hard to find tools that cheap
anymore.  Things that cost more the $100 require more thought.

  I think you just made anybody interested in providing professional Common
  Lisp systems anxious and less willing to help you reach your goals,
  because your goals clearly include the elimination of a livelihood for
  Common Lisp vendors.  this is not smart of you.  in fact, it may cause
  what source access we have today to be retracted because people like you
  are likely to steal it and violate license agreements as you see fit.

| That there would be a tall wall between the lucky ones who can access
| LISP for their work, and those who mostly can't.

  the existence of all life forms is based on whether they can get what
  they need to thrive and survive.  there is nothing in the human condition
  that invalidates this.  "have vs have not" is a fact of life as such, and
  as unchangeable as the earth orbiting the sun.  humans, however, tend to
  worry about it and want to change it, and this is perhaps good for some
  of us, but we can't get rid of the fundamental principle of life: there
  will _always_ be a "tall wall" between the "lucky" ones and the rest.

  moreover, "luck" is not what you think it is.  success is exploiting
  timing.  luck is _only_ timing.  luck is like winning the lottery.
  success is seeing and taking advantage of any break you can get.  that
  both have to do with timing is a mere coincidence of timing (luck :).

  however, what you are doing is actually _depriving_ people of what they
  need to thrive and survive: if writing something in Common Lisp means it
  can only be given away, people with financial responsibilities will flee
  in panic as all they can see is non-recoverable expenses, and everybody
  _but_ the people who do the real work will be rewarded, in much worse
  ways than "capitalists" are supposed to exploit "workers" today.

  in my view, Common Lisp provides _values_ that far exceed what other
  languages do, but I fully realize that people won't understand this until
  they have actually experienced the problems that it solves for them.
  now, what's _really_ sad is that those who want free Common Lisp systems
  have _not_ seen what the commercial systems provide in just this way, and
  they will go on to make free Common Lisp systems that just _don't_ offer
  the serious advantages that solid Common Lisp systems do that people
  can't _afford_ to provide for free.

| That's the disappearance in irrelevance that I fear.

  one man's niche market is another man's irrelevance.

| This would be (and this actually is) an enormous waste or human
| resources: people (LISPers and non-LISPers) who do things independently,
| and ever restart from scratch, instead of building on each other's code.

  geez.  welcome to planet earth.

| So much human resources spent for no progress!

  you _really_ have to be new here.

| Granted, nature lives on infinitesimal yields.

  yes, this is a _fundamental_ premise.

| Still, unless I am proven that this is a inevitable doom, I won't be
| quite satisfied.

  some people will never be satisfied no matter what you give them or do
  with them.  it is important to know who those people are, so we don't
  waste any time or other resources on them.

  doom is inevitable, BTW.  mankind will die out, planet earth will be
  vaporized when Sol goes nova, if not sooner, and then Common Lisp will
  have to acknowledge defeat to the unwavering hostility of the universe.
  for those of us who plan to become immortal, this is a serious concern.

| Well, maybe I'm just unlucky, and not introduced in the right places, but
| here in France

  depression is all about focusing on the negative things you observe,
  about expanding the impression of being "unlucky" to some kind of cosmic
  condition and from then on doing the opposite of what normal people do,
  which is to ignore the negative and the hardships because they _expect_
  every ounce of positive experience to have a pound of cost, and they
  _know_ that errare humanum est, and that errors have costs, too.  on top
  of this, the depressed tend not even to _see_ anything positive.

  instead of blowing your own dismal experiences out of proportion, look at
  what made people change their minds to use clearly inferior languages in
  somehow superior _tools_.  see if you can do something about this and
  make a couple franc on it.  instead of doing the stupid "if everything
  were free, we'd never be hungry and there'd be peace on earth" routine,
  think about what _you_ need to _supply_ a Common Lisp environment to
  people, what you would need to do to make people _buy_ from you to keep
  you alive, and what your _customers_ would want to do.  think business.

  "free" works very well for the hobbyist market.  it works not at all for
  the professional market.  now, you might argue that all programmers are
  hobbyists, but I'd argue that that's about as relevant as your experience
  with Lisp "dying" in France.

| While I'm happy for you, and convinced that the positive activity of
| people like you will only increase the number of people using CommonLISP,
| I fear that, thanks to your own value, the world around you is much
| better than it is elsewhere, and I'm still worried about the future of
| LISP.

  this _really_ sounds like clinical depression.

| Feats of exceptional individuals are not enough to make prosper large
| economies; economies work as the result of a network of normal people,
| with a normal bell-curve distribution of proficiencies.

  I'll take that as a compliment, but I think you exaggerate my abilities
  and what I have done so you don't have to consider me as part of "normal
  people".  this is not new: if normal people do something exceptional or
  out of the ordinary, they somehow have to be exceptional individuals, and
  whatever remains as "normal people" can feel better about not doing
  anything exceptional.  I think this is fundamentally false.  first, this
  delusion gives normal people a psychological barrier to break out of the
  pattern.  instead of facing hardships and surviving them, they meet
  hardships with "am I sufficiently exceptional to _ever_ win?", and go on
  to lose, which only serves to prove their point.  second, it makes it
  much harder for normal people do to a _few_ exceptional things, because
  they then have to defend why they, as exceptional individuals, do _only_
  normal stuff the rest of the time.  these are fundamentally _dangerous_
  psychological issues and make individual success into a serious hazard
  for people who _aren't_ able to take the expectations well.  (you need
  look no further than to the many musical artists who turn to drugs to
  deal with the pressure.)

| And if there's any reason to worry about LISP being more widely used
| (which you seem to do, since you're affirming gladly that the Lisp market
| is growing), we need understand the phenomena at work, and not just blame
| people for not being virtuous enough.

  that didn't quite parse, but I don't blame people for not being virtuous
  enough.  nobody does that as far as I can see.  I'm saying that you need
  to mature to appreciate the value of Common Lisp.  being immature is just
  another fact of life, but maturing doesn't happen to people who expect to
  have reached their goals and set no further goals for themselves.  add to
  this that reaching your goals sometimes means abandoning what cannot be
  reached -- going on and on forever is also immature.

| To prosper does not only mean survive, but also grow.

  survival is easy to measure.  so is decline.  growth, however, is hard to
  measure, since in order to do so, you need to know the goal and see the
  small steps on the way (and why certain setbacks are _not_ decline), and
  absent such often secret information, you can't see growth because you
  don't know what to look for.  this is why whiners will point to the
  decline of Lisp (which they think they can measure by any form of failure
  to live up to _their_ expectations) and ignore the growth because they
  have some pretty bizarre ideas about what growth means in the first place.

| I can very well imagine Perl 7 having just every semantic feature from
| LISP ...

  I'm sorry, I can't.  people said the same about GUILE a few years ago,
  and what we have today is a hopeless mess that's only getting worse.

| If LISP is to prosper, it must drop proprietary software barriers; it
| must enter the free software model.

  yeah, if Lisp is to grow in _your_ mind-set, it has to do that.

  however, your mind-set is a recipe for economy-wide disaster, as the free
  software world will find out the very hard way when they have succeeded
  in defeating Microsoft, which is the _real_ purpose and driving factor
  behind Open Source.  Free Software has a higher goal, but it appears that
  the success of Open Source projects are jeopardizing the goals of Free
  Software, which I think is quite clear if you watch what RMS is saying
  and doing lately.  if you _understood_ his reaction to Linux a few years
  ago (instead of just ridiculing it), you will appreciate that his concept
  of Free Software is in jeopardy.  Open Source is much _more_ in jeopardy
  than Free Software is, because its success is defined in terms that do
  not stand up against a maturing public.

  yes, you are seriously confused and you spread lies and confusion to
  others, as well.  what you're doing should be illegal.  it may well be.

  I do what you say is impossible in a production system.

| How about this simple CLOS test to add another slot to an existing
| stream.  It generates an infinite loop, and quickly kills off the lisp
| entirely.  Should I file a bug report, or
|
| USER(2): (defclass mystream (BIDIRECTIONAL-TERMINAL-STREAM)
|          ((myslot)))
| #<STANDARD-CLASS MYSTREAM>
| USER(6): (setf x (change-class *standard-output* 'mystream))
| #<MYSTREAM [initial terminal io] fd 0/1 @ #x200061d2>gc: done
| gc: done
| <<<lisp dies>>>>

  I just tried this in both 4.3.1 on SPARC and 5.0 on Intel/Linux, and
  _nothing_ unexpected happened.

  I sincerely think you should stop doing something that clearly creates so
  many problems for yourself.  instead of being fun and profitable, Common
  Lisp is hurtful to you.  go do something else and make yourself happy.

By the way, with respect to CL being OO:

 1. There are some comment on this at
http://www.elwood.com/alu/table/objects.htm.     I welcome any
suggestions for changes/enchancements.

 2. For my money, the big win with CLOS is often not in clases, but in
generic-functions, and this is what sets it apart from other "OOLs."
Given that structure-class and builtin-class are, in fact, classes that
can participate in dispatch, the trick with developing a MOP for
something like streams is in coming up with the generic function
protocol that can be extended by the user.  The actual metaclass of the
user-defined classes which can participate in the protocol is less
important.  In fact, I could imagine a streams protocol which defined
the generic functions and the classes involved, but left it up to the
implementation to choose and document the metaclasses of these classes
--- or even the metaclasses of the generic-functions!!!

 Come up with a business plan you like that you are willing to bet real
money on.

Anyone who is serious -- and by that I mean that they are willing to
devote real resources (e.g., cash) -- should go ahead and contact a
vendor with their proposal.  Do so by contacting them directly with at
least a summary of your specific proposal.  (I do not recommend just
leaving message on c.l.l saying "why don't you guys do XXX".)  If you
don't know how to write up a business plan, there are resources on the
net.

I'm sure that all the vendors would be glad to hear from you, but
speaking for Elwood, you can contact me through how...@elwood.com.

You make some compelling arguments, Erik.  I think I agree that the
rise of Open Source owes much of its strength to anti-Microsoft
sentiment.  However, I think the movement is somewhat more complex.

Given the assumption that Open Source hasn't proved itself yet, I
think it does offer longer term benefit.  For example, the company I
work for (www.homedome.com) sells fiberglass observatories to amateur
(and pro) astronomers.  I work on the remote computer control systems
for them.  These systems include client software for the PC and an
embedded controller which goes in the dome.  (Please pardon my use of
VB on the PC- it was convienent, but I do use Emacs for the embedded
stuff).  As soon as its viable (and I learn enough), we will write a
Linux client.

We release the source code for the PC client software along with the
binaries under the Open Source model, and although we haven't put it
on the web site yet, we will do so when the code becomes more stable.
Its evolving so rapidly that its hard enough to manage the version
control alone.  We have 2 control systems and have shipped 15 of one
and over 20 of the other.

We regard the PC client software as something of a lost leader.  In
our opinion, the more people who write programs for our dome
controllers the better so releasing the source might help spur
purchases of our control system.  If someone chooses to reverse
engineer the embedded controller, and maybe use our protocol then we
must compete on technical grounds- but thats what people do in a
competitive market.  If we cannot compete on technical grounds, then
we loose.  I guess the "natural" tendency is to put all sorts of
license restrictions in place and keep a bunch of lawyers chained up
and fed with raw steak, but we're a small company and can't afford it-
and I think our customers would be better served in an open
enviornment with competition (or at least the opportunity for
competition).

I can't comment on the other Open Source models, but I think if
companies positioned to do something like this do so, it will make
many systems of various kinds more available and interoperable -
simply because the interfaces are opened and documented.

Having said the above, we do not plan on releasing the embedded code-
we want the systems to be open for people to use how they choose, but
in the end we must protect our technology investment.

Its possible that "pollutes" me with respect to Open Source, but I
think the ultimate significance and form of Open Source is not yet
clear.  It may be that Open Source is only effective for particular
classes of software, if so, I think thats to everyones benefit because
I think the computer related industry is far too complex for a single
business model.

in keeping with the analogy, it ought to be mildly annoying to take
the safety off in order to let you be a bit more relaxed during normal
operation but easy enough not to be prevented from taking your foot
off should you really want to.

i mean you do use lisp, you could be running assembler.  why use a
high level language in the first place?

[...]

(And memory goes back to a Spider Robinson short story where two
foolish burglars make the mistake of thinking a house is empty.  When
they regain consciousness, they realize that they have made the severe
mistake of provoking an author, who cyranoacrylated them together, and
now that they are no longer disturbing him, has gone back to his den
and forgotten that they exist...  "Help!  Police!  WE GIVE UP!")

Dear Erik, dear Howard, dear comp.lang.lisp readers,
   I would prefer to resume my work rather than post yet another article
on the topic, but there are many misconceptions about free software
that seem to creep in your posts, that I'd like to dispell,
even though odds are low that will suffice to convince anyone.
As I do, I'll try to draw the sketch of a business plan
for a LISP vendor producing free software LISP systems,
as suggested by Howard R. Stearns in another message.

Other topics addressed in another message.

First, I want to clarify that I do NOT want the death of LISP companies.
It's all the contrary. I only wish them success and prosperity.
Only I'm convinced that the proprietary software model is flawed,
and that we'll only see LISP flourish
when there is a solid free software infrastructure to back it,
be it based on current LISP vendor capital, or on new capital
(I have no doubt that the people doing the underlying technical work
will be mostly the same in any cases).

Under the free software model that I'm convinced is economically sound
(and much sounder than the proprietary software model),
LISP development companies make money by selling *services*, not licenses.
Services include:

* trust (availability, reliability, branding, of code),
* support (correction of bugs in the customer's code or the core system),
* development (ports and extensions to the system),
* customization (development and deployment of customer-specific solutions),
* education (training and documentation).

Companies that already prosper thanks to free (non-LISP) language software
include Cygnus (www.cygnus.com) and AdaCore Technology (www.gnat.com).

The real, worthy, asset of these companies
is NOT their "intellectual property",
it is their *human capital*.
They prosper because they value *people* rather than software.

To them, non-licenseed users are growing their overall market,
and the opportunity for future services being sold,
not "thieves stealing important company asset".
So the marketing is: advertise solutions, advertise support,
give away as many copies of the software as possible (including by FTP),
to create an ever growing market for *services*.
I invite you to read the AdaCore site to see that these guys mean business.

Of course you may disagree that this model is sound,
but at least I'd like you to acknowledge that even if they be mistaken,
free software advocates are NOT against business.
I am deeply convinced of the essential morality of free market economy;
now if you read classic texts (Turgot, Bastiat) you'll see
that market economy is free exchange of *services*
and elimination of *priviledges*.
I rank any Lisp related *work* (see above) as a *service*,
and licenses as *priviledges*.

FR> Harlequin (and maybe Franz) already earns most money from services;
FR> maintaining their LISP implementations is part of their cost structure,
FR> not of their benefit structure.
EN>   this looks like a statement of fact, but it is mere opinion.
Yes, I'm sorry, and I apologize.
Although I'm as convinced of it as can be without seeing actual figures,
this was only an extrapolation,
and I shouldn't have made it appear that factual. MEA CULPA.

However, I'm not extrapolating from nothing,
and since you pinpoint the above fact,
I feel compelled to explain my reasoning:

* commercial licenses from LISP vendors include support,
 a support that you've demonstrated was highly valued.
 This I count as a service, even if in its current form,
 its sale is tied to a licensing agreement.
 I'm convinced that people who purchase licenses with support
 would equally buy that same support that they appreciate,
 even though the bare license was free of cost.
 In other words, the monopoly value of the license seems little to me
 before the service value of the support.
 Even though the sources be free software, the vendor,
 as principal author and maintainer of it, would still have
 a competitive advantage over competing providers for support service.

* It is the announced policy of Harlequin to move
 from the language vendor industry as such into the service industry;
 it looks like they already have quite a business in that field,
 and are only expanding more into it,
 because they (at least) feel that the big bucks is definitely there.
 In such a setting, the software is useful as a tool,
 but it's a burden as "intellectual property",
 since it enters the cost structure and becomes a burden to market entry;
 making it free software both allows outsourcing of part of the development
 and removes fear from potential customers of being tied with proprietary
 tools that however technically good are guaranteed anti-perennial.

All in all: the market, even the current one, is mostly in services,
and the monopoly value of proprietary licensing does not look to me
like it benefits much to the vendors, all the less in a niche market,
while it prevents market growth.
Can someone at a LISP vendor tell me how right or wrong I am,
even without giving any figure?

FR> Their IP barriers only protect them from each other, leaving them on a
FR> small market.  Dropping these barriers will allow them to attack the real
FR> enemy, and grow a mass market.
EN>   so "growth" coincides with "mass market" in your view.
Well, since we're talking about a programming language,
that'll more be "mass of average developers"
than "mass of average consumers".
But yes, I'm convinced that growing the LISP market
means attracting more developers to LISP.
It'll benefit vendors by lowering the customer's cost of entry
in a LISP system, and lowering average wages of LISP programmers.
It'll benefit good LISP programmers by giving them the opportunity
to rise in a market of LISP programmers.
It'll benefit (currently) non-LISP programmers by giving them
the opportunity to find a decent job where to program with a decent language.
It'll benefit end users, because LISP will yield
better cheaper more reliable software.
Everyone wins. Only C++ and Java crooks lose. Unless they become LISP crooks.
If you stay a niche market, invert those benefits.

I am deeply convinced that the interest of LISP vendors
is to make their core system free software,
so as to gain as many developers as possible, and sell services.
So as to get a stable hold on the heart of developers,
they should also encourage the development of a large developer community,
by creating a body of software that covers their need
(i.e. removing any possible "need" for tools like make, perl, and
all the unix infrastructure, yet allowing unix tools to seamlessly
interface the LISP system so as to incrementally penetrate the market;
having a native Emacs port).
That's the kind of developer-directed marketing I'd do.
Once developers are hooked on a technically superior system,
they won't leave it, and will become the seeds that grow
into demand of more services from their employers.

Remaining customers promptly decided to migrate
all their development to C++, Java, Smalltalk, or whatever they could.
All their code base was technically lost.
Even though they could negociate the use of sources for Talk as customers,
and even though their maintenance contract will last a few more years,
they would have to do all maintenance in-house afterwards,
and not be able to reuse the resulting code for other projects
or share it with others. Economically, their code is DEAD,
and it will soon be in every other meanings of the term.

Even the healthiest vendor can die or change its business plan
(should I cite names of LISP/LISPM companies?).
Even the wealthiest software company on earth eventually ceases to maintain
its development platforms (my employer knows that *very* well!).
With proprietary software, the consumer is at mercy.
His development software is not perennial.
When it's a lame low-level language like C/C++/Java (because of JVM),
that's fine with him, because the abstraction level required
for compatibility is low (like the whole of the software),
and he can easily switch vendor.
When it's a high-level language like LISP that has special dependencies
on a large proprietary runtime, he's stuck, and too bad for him.

By making code free software,
you give customer the guarantee of perenniality.
Customers are beginning to understand that, and the more they do,
the hardest time vendors will have at not publishing free software,
unless they are die-hard big-time monopolies.
Moving into a free software service company means you don't rely
so much on the artificial monopoly value of licensing your software,
but rather on the natural workforce value of your employees.
That's what I imagine as a sane industry,
due to generate sane revenues for a long time.
The more your customers understand that, the more you explain it to them,
the more they come to you rather than to proprietary rivals.
That would be a good marketing plan towards general management:
not only are these technically SUPERIOR solutions
that work NOW for EFFECTIVE development COST,
but they are solutions that will last FOREVER,
and that even if the vendor's archives are nuked,
and all its developers poisoned by eating rotten meat at a company lunch.

Again, I may be wrong, but it sounds quite consistent to me,
and I'd like to be *proved* wrong if I am wrong
(together with all the many serious free software advocates I know).

Dear Erik,

EN> depression is all about focusing on the negative things you observe,
Well, I can see *lots* of positive things around me,
and particularly things happening in the software industry.
I regret to say none of these positive things are related LISP,
whereas I have see lots of distress about people having to not use LISP.
Again, it's only my own limited experience (although I've been in contact
with developers from a variety of different backgrounds,
it's still a limited experience).

As for focusing on the negative things people observe, it seems that
you have a talent for doing just that on comp.lang.lisp posts:
in your replies, you tend to inflate the (actual) defects
of other people's postings, and ignore their positive content.
Of course, this all may have to do with your only replying to what you
want to correct, and not with what you agree with :)
Now, the same phenomenon is perhaps present in most usenetters.

FR> [...]
EN> I'll take that as a compliment,
It is. And it isn't meant to bribe your @{%[#! opinion.

EN> but I think you exaggerate my abilities
It's not about anyone's unique outstanding abilities,
but I believe that, to simplify,
there are combinations of various positive factors
independently developed in everyone;
hence the notion of a normal bell-curve distribution of people
with more or less positive combinations of factors.

I otherwise agree with most of what you say
and welcome your advice about maturing,
but I don't feel that my fundamental arguments
are contradicted by your post.

EN> [about Open Source vs Free Software]
Open Source(TM) is but Free Software in a business suit;
it may use a different language from a different point of view,
but despite the different form evaluated from within a different package,
it has the same semantics (well, as far as I and many understand it).

EN> if you can't make [LISP vendors] see things your way,
EN> buy them up with your own money and make them
EN> do what you think they should.
Well, I do intend to build a free software reflective LISP operating system
(CommonLISP or not), allowing for orthogonally persistent distributed
programming, and make money out of *services around it*,
as soon as I finish my PhD. If would sure be glad to raise capital
from LISP vendors, since it would give me the opportunity of working
with elite programmers, but I'm pretty sure I'll find capital and programmers
somewhere else if they are not convinced.

EN> here's a scenario that you should consider: remove intellectual property
EN> from broadcast sports shows.  explain how it does not kill the sports
EN> industry.
Oh, it may kill the sports industry *as it is*,
but will greatly benefit a reorganized sports industry.
Instead of a corrupted industry where plays
a little elite of players on steroids,
it would be a sane industry where many people earn honest money
by having more population do sport instead of watching it on TV,
so that sport becomes a lively activity as it should have always been.

And I'm convinced that if you wait 20 years,
you'll see that the *sports* industry itself
(I don't mean the parasitic show-business around it)
will have grown, and be much more prosperous than it is now,
while the population will be more sportslike and healthier.
A few people gain much less money (they are still quite decently paid),
but a lot of people live honestly doing a useful job.
There will still be shows, but they won't be as much
of a mediatic event as they are now.

Of course, if you only remove intellectual property *from the sports shows*,
then the show business will move somewhere else,
and sports will greatly suffer at first.
But if you also remove intellectual property from everywhere else,
then it will be a general world-wide regeneration of all industries and arts.
Instead of have bribed arts that are worth mostly nothing,
with a premium being given to information in as much as it is new
and unattached to previous work (hence not building up richer structures),
and with technical and ethical criteria coming last,
artists would be paid for their actual work, not their face value:
creating, performing, composing, designing, recording, etc.
There won't likely be as many multi-billion dollar superstars
as there now are, but instead the serious, working, talentuous, artists,
will be recognized for their real work, and live decently;
the really good may even become rich,
though most likely not as much as can be seen today.

I could also talk about property of information inducing a producer-driven
economy, whereas license-Free Information means a consumer-driven economy.

Of course, this is only me considering your scenario.

EN> you support my conclusion that Open Source is only a means of fighting
EN> Microsoft.
Well, I know a lot of people from free software associations,
and I have contact with developers in many different sites,
who all support the idea of free software. I can tell you
their involvement in free software is *quite* independent
from Microsoft or any particular vendor (criticisms include LISP vendors,
though they are appreciated for their technical merits).

Best regards,

FRR> You seem to imply that free software advocate all start
FRR> from an A PRIORI dogma of "entitlement" or "right".
KMP> You inserted the word "all".  If you remove the word all from your
KMP> statement, and infer instead the implied word "some" which is the
KMP> default usage, I do mean to imply THAT.
Yes but then it's a more of straw man argument, and in as much as it holds,
it could very well be used one way as well as the other,
since as far as I can tell, tenants of proprietary software
are no less no more religious than tenants of free software.
Hum. Looks like I re-launched a discussion for a petty argumentation detail.
Oops. Sorry. I certainly didn't want to accuse you of fallacy.
After re-reading the posts, I admit that I have reacted inappropriately.
I will try refrain from posting on such subjects on comp.lang.lisp anymore.

The problem at stake is not with the opinions held by anyone,
but only with the fact that these opinions be held A PRIORI vs A POSTERIORI.
I *might* claim that freedom to FOO or not to FOO is a right,
yet not pretend it's a god-given right of which I had a revelation,
but rather than it finds its justification A POSTERIORI
in the global nuisance associated with refusing that right,
and the global welfare associated with acknowledging it.

KMP> You have obviously though hard about this issue
I thank you a lot for acknowledging it.

KMP> and I won't try to dissuade you from your personal theory.
Although I appreciate that,
I think this proposition should be quite independent from the former.
The latter should follow only
from our having found *where* to agree to disagree,
from my theories not being dangerous,
or from our mutual enlightenment otherwise not being worth its cost.

What the fact that we both have thought hard on the subject implies
is that we likely have internally consistent point of views,
and that it should be easier to find where they match not.

KMP> My position is based on the classical construction of copyright law in the
KMP> first place, which is that it exists to assure people will share.
And I admit my thoughts have led me to question this very construction,
that I had long before taken for granted, unquestionned.

Since we already found a fundamental divergence,
we can but acknowledge the further divergences.
But we can still explore and compare our conclusions
and let each other decide whether they reveal some contradiction
in one's own explicit and implicit assumptions.
We can also understand each other better.

KMP> under weak copyright law, some would not speak, and that would be a loss.
There are also people who won't speak under strong copyright law.
So the whole problem is:
"which law encourages most publication of useful information?"

KMP> Strong copyright law is clearly correct in the trivial sense that
KMP> using it you can implement weak copyright law.
While this of course isn't enough to break the conclusion of it,
I refuse this argument.
It's as if you said "right to government subsidies is clearly correct
in the trivial sense that you can implement lack thereof (by refusing them)".

Copyrights ARE a privilege granted by government
(which even the US constitution acknowledges),
and the indirect subsidies induced by this privilege
is the very encouragement to publishing done by copyright law.
These subsidies are paid directly and indirectly by all the citizens,
and that even if they don't want it; just because a citizen
doesn't claim privileges for himself won't discharge him
from having to pay for the privileges claimed by others.

KMP> I also think code copyright is too long.  Three years seems adequate.
Indeed, the duration of the copyright is the price paid by State.
The amount of this price is not an eternal truth,
but has to be decided so as to optimize publication of useful information.
In particular, the amount should adapt to technological environment,
but this seems too difficult for legists to do.

My opinion is that there shouldn't be such thing as
automatic granting of privileges by the government,
for price should be decided by the market:
if people want an idea to be published rather than kept secret,
they can organize in private groups who'd buy *access* to the information
and republish it (or not). There is no need of forcibly slowly-adaptating
state law for that, only of competing private infrastructures.

A free market means that information will be paid a balanced price:
worthless information won't be paid
(instead of being given a systematic premium as is the case now),
while worthwhile information will be paid as high as can get
(instead of being systematically undervalued as it is now).
Maybe the state itself, as an economical actor among others,
may participate in bids, offer prizes, etc, for particular information;
it needn't systematically enforce "egalitarian" price to all information.

This is an elementary free market argument to me.
I understand that it might not be compelling to you, though.

KMP> I can tell you for a fact that I would never post to this newsgroup or the
KMP> web at the rate that I do if posting meant giving up my rights.
What rights exactly do you want to keep? And whence do they come?
Are they natural rights, or State-given privileges?
Is that protection against copy, or against misuse and liability?
There is no A PRIORI reason why you should be liable
for people copying your work, or why their activity
be not subject to prosecution for libel, perjury,
misrepresentation, and crookery, independently of any copyright.
I'm not convinced that Copyright as it exists protects the Right Things (TM).

It is great that on usenet,
we may find people capable of civilized discussion;
I will try harder to be among of them.
Thank you, Kent, for your living lesson,
and sorry for wasting your precious time.

And now, back to LISP!

"Howard R. Stearns" <how...@elwood.com> writes:

If you want me to undertake a precise business plan,
I'm willing to do it, but I require to be paid accordingly,
and since I'll need help from an accountant or other business consultant,
this means you'll have to pay the consultant, too.
And so as to build this business plan, I'll need access
to your company's business records, too, and likely to
consulting relative to your country's national business legislation.

If you're interested in hiring my services for this matter (or another),
you may contact my employing laboratory [1].
We have experience in the design, development and deployment
of very-large long-lasting high-reliability real-time information systems,
for which we're more and more convinced that proprietary software products
are part of the problem set as much as of the solution set.
Although defining business plans for foreign companies
is not our primary activity (again, we'll require external consultancy),
we are always looking for diversification of financial resources,
as opposed to reserving our expertise our mother-company.
Your offer, if firm, will be considered seriously.

Now, if you want me to undertake such work on my own spare time,
for a company in which I'd invest personal capital, I even might.
But only when my current contract is over,
and then, I'm afraid it will be for my own company,
not for a foreign company I only know from Internet hearsay.

[1]     DTL/ASR  -  CNET  -  France Telecom
        38-40 rue du Général-Leclerc
        92794 Issy-les-Moulineaux Cedex 9 FRANCE
        Phone: (+33) (0)1 45 29 54 91
        Fax: (+33) (0)1 45 29 66 04
        e-mail: francoisrene.rid...@cnet.francetelecom.fr

Best regards,

Talk is cheap - particularely in your case it's all you ever do, be it
the LispOS list, the Tunes list, comp.lang.lisp, whatever.

For all your philosophical spewing about morality and ethics and
"Regeneration day when all Lisps are free again", this post to me is
the pinnacle of hypocrisy.

Franz has been extraordinarily generous to the Lisp community in
providing the Allegro CL Linux Trial Edition, which is virtually their
Enterprise edition (at least thinks it is), for $0 for non-commercial
use.

Harlequin, with Kent Pitman, provided us with the HyperSpec, a
document that clearly took an extraordinary ammount of time and effort
to produce, and as far as I can tell never directly planned to or did
in fact make any money from doing this.

I have a feeling you aspire to become one of these people like Eric
Raymond, that like to talk a lot but shouldn't be mistaken for anyone
that actually writes a whole lot of code[2]. Eric Raymond though has
spent a lot of time consulting with corporations and vendors and the
media laying out how they could profit and succeed with an Open Source
model, without charging for his time. Whether he does this for his own
ego or not is not so important as the fact that good results are being
made.

Have fun writing your little papers,
Christopher

[1] If your bizarre Curriculum Vitae is any indication
http://www2.tunes.org/~fare/cv.html

  what you have posted, Francois-Rene Rideau, is basically on par with
  going to an investor and saying "I am deeply convinced that it is in your
  interest to give me X amount of money, so how about it?"

  the more insane drivel I read from proponents of Free Software and Open
  Source, the more convinced I am that this once noble movement will go
  away and not even become a footnote in the history of mankind when their
  real goal has been achieved: the destruction of Microsoft.

  in my view, Microsoft has never been particularly relevant.  they are
  very good at defrauding people who don't know what to look for every time
  they part with a small amount of money, but if I were to be fazed by such
  people, I'd be seriously bothered by any and all politicians, commodity
  advertising, televised propaganda, etc, and I'm not.  I'm bothered by the
  fact that stupid people don't spontaneously combust, which they should.
  no, seriously, I'm fighting the view that Microsoft is relevant to anyone
  and want people to look over yonder knoll when Bill Gates is in jail or
  in a mental institution and none of the mindless droids in Redmond know
  what to do with their lives when fraud is no longer a viable option.
  Free Software is a protest movement.  Open Source is a weapon.  I can see
  a significant constructive element to Free Software, but I can't see any
  with Open Source.  Linux is _not_ a success to be repeated, just like
  Microsoft is the last of its kind.  in both cases, however, men in suits
  with their ties so tight blood supply to the brain is cut off believe
  they know what caused these successes and attempt to emulate them through
  some incidental quality that has nothing to do with the success.  some
  other fanatics also crawl out of the woodwork to predict doom and
  disaster if everything doesn't follow the Grand Business Model de jour.

* Francois-Rene Rideau <f...@ZhengHe.augustin.thierry>
| I would prefer to resume my work rather than post yet another article on
| the topic, but there are many misconceptions about free software that
| seem to creep in your posts, that I'd like to dispell, even though odds
| are low that will suffice to convince anyone.

  some day, I hope you will start to listen to people who have worked with
  Free Software for the better part of a decade and have wondered (1) why
  people get much enthused but then leave disappointed and disgruntled, (2)
  why free software programs gain a very high quality as long as the goal
  is very clear, and (3) why it then goes on to accrete crappy features
  nobody needs but are fun to add by the less competent people who are
  unable to accept that an idea has fully matured.  (watch GNU ls acquire
  "human" sizes, for instance: multiples of 1000, instead of 1024.)

  Free Software or Open Source is _not_ a panacea, nor a religion, and it
  should be possible to argue against it without someone who is still wet
  behind his ears coming out to "dispell misconceptions" because he fails
  to understand what people are talking about.

  as I have said previously, immaturity is a fact of life, which it is not
  smart to deny or ignore, but which also should not be catered to.  people
  who think they are already mature _enough_ are unlikely to mature _more_,
  but part of the process is to realize that some of one's core beliefs may
  be wrong or have underlying goals that are reached only in part or not at
  all, but which it does not make sense to pursue further.

  unencumbered access to other people's intellectual property is one such
  _immature_ goal.  the same argument could be made for other property, yet
  isn't, because the glaring insanity of such requests cannot be ignored.

  the very first step on the way to convince anyone is to accept that they
  may be right, and you have to figure out how to talk to them in terms
  they can understand and relate to to make them change their ways.  people
  do whatever they do because it has worked better for them than the other
  stuff they have tried, but at some point, they gave up looking for better
  ways and were "satisfied".  your second step on the way to convince
  anyone of your own views is to make them _want_ to continue to look for a
  better way and _maybe_ that is your way.  if they don't want to continue
  to look for a better way, they should be ignored completely as a waste of
  space and human potential.  since the third and following steps are by
  now completely irrelevant to Francois-Rene Rideau, who has found the
  solution to all the world's problems, I'll save myself the effort of
  writing them down.  this has gotten long enough already.

* Francois-Rene Rideau <f...@ZhengHe.augustin.thierry>
| While I'm giving away my ideas about a general free software business
| model because I'm personally convinced they might make a better world,
| I'm will not be giving away my work time, because I value it much;

  so you value your "work time" more than actually _doing_ something to
  make a better world?  geez.  this translates into "I don't believe what
  I'm saying, but it would be nice if some other people did".

| If you want me to undertake a precise business plan, I'm willing to do
| it, but I require to be paid accordingly, and since I'll need help from
| an accountant or other business consultant, this means you'll have to pay
| the consultant, too.  And so as to build this business plan, I'll need
| access to your company's business records, too, and likely to consulting
| relative to your country's national business legislation.

  let me get this straight.  you are unwilling even to invest enough of
  your own time and money to create a business plan that might help reach
  your goals if doing so could help other people believed in it?

| If you're interested in hiring my services for this matter (or another),
| you may contact my employing laboratory [1].

  not bloody likely.  after the above stunt, I cannot even assume you know
  how to do the job you are paid to do and I certainly would _not_ employ
  you in a business that planned to make money.

| We have experience in the design, development and deployment of
| very-large long-lasting high-reliability real-time information systems,
| for which we're more and more convinced that proprietary software
| products are part of the problem set as much as of the solution set.

  so get better people.  that probably means firing you, so I can see why
  you argue in favor of giving sources away so others can do your work for
  your for free if you just give them the sources.

| Your offer, if firm, will be considered seriously.

  except that you will not fund it yourself, which is what "seriously"
  _means_ in a discussion of business issues elsewhere.  I'll bet the
  saying "put your money where your mouth is" sounds really alien to you,
  but it's a very common saying towards people who are suspected of
  spouting hot air and not actually _wanting_ anything.

| Now, if you want me to undertake such work on my own spare time, for a
| company in which I'd invest personal capital, I even might.  But only
| when my current contract is over, and then, I'm afraid it will be for my
| own company, not for a foreign company I only know from Internet hearsay.

  so you now admit to know nothing at all about either Harlequin or Franz
  or Digitool or Elwood or any of the other players in the Lisp field.

  I'm glad my suspicions about you were not disproven.

  you'll note that I ignored the Followup-To to gnu.misc.discuss, which is
  just another of your despicable cop-out tactics.  you're just a nutball
  like the majority of the proponents of stealing other people's work, and
  when you can't talk your way out of it, you certainly will not _work_.
  and this is the commonality of people on gnu.misc.discuss, too, so thank
  you for telling us where you really belong.

  I hope Free Software and Open Source are considered less viable means of
  doing business after this, and that anyone who considers using their
  licenses to publish software in source form (which I still think is a
  good idea, but not for any of the reasons the shallow morons think, and
  most definitely _not_ to the general public), will think twice and talk
  to a good intellectual property lawyer about how to secure the rights of
  the _creators_ of intellectual property, not just those of its _users_.

  I'd like to get back to the future of Lisp now that we have shown what
  kind of people are likely to argue in favor of unencumbered access to
  other people's investments.

Kent M Pitman asked:

Fortunately, the situation has changed a lot in the last few years.
Many of my clients now have machines with at least 32 MB RAM.
And on the Internet, it doesn't matter how much memory the users
have as long as they can run a browser.

So I bought Harlequin Lispworks in April 1998. Since then
I've finished three small (but paying) projects with Lispworks
and I'm close to finishing a fourth one. I am very happy that
I can earn (part of) my living using the language that I fell
in love with 14 years ago. For me, Lisp is not dead, dying or
doomed. On the contrary: it's finally starting to bear fruit
after a 14 year germination period.

Getting back to Kent's question, in my current situation
the most important barriers are:
1. Users still need more than 16 MB for running a GUI program in
   Harlequin Lispworks on Windows 95.
2. Some of my clients still require Windows 3.1.
3. Lack of libaries for stuff like HTML generation/parsing, regular
   expressions, TCP/IP, POP, SMTP, different character sets,
   graphics formats like GIF or JPEG, etc.

That sounded like a pretty big 'barrier to Lisp acceptance'.
So I had almost resigned myself to the prospect of spending part
of my Christmas holidays on learning yet another scripting language,
when it dawned on me that there was a better way: I could write
a Lisp program to analyze the database and generate the necessary
HTML-files plus some very simple Javascript-stuff for the search
interface. That way, the server wouldn't have to run any scripts
at all.

The general strategy is: if you can't write it in Lisp,
generate it with Lisp. Do as much of the processing as possible
at compile time instead of run time. That way, you have a good
chance of using Lisp for the interesting stuff at compile time.
The requirements ("you can only use C, or Perl, or ...") on the
run time part are much more acceptable this way. First, there's
less of it. Second, when you're lucky, you can actually generate
it in Lisp.

This strategy works quite well for me.

The software world is a very complicated system with lots
of different niches. I think I'm slowly growing into a niche
where I'll be able to use Lisp for almost all my work and
I'm very happy with that prospect. It would be interesting
to hear more about Lisp niches that other people have found.