source access vs dynamism

[Francois-Rene Rideau <I+fare+WANT@tunes.NO.org.SPAM>]

Dear Erik, dear readers,

Erik Naggum <erik@naggum.no> writes on comp.lang.lisp:
>   I'm advocating source access to people who express an actual desire and
>   need for it.
So am I. The question I raise is: "who'll be judge?".
In one case, it'll be an all-mighty centralized marketing department,
and in the other case, it'll ultimately be the person in need oneself.
Of course, in a perfect world, the person in charge will choose well;
but in a perfect world, the person won't need to be in charge, either.
So the question is about the dynamic effects in an imperfect world:
what attitude has most positive dynamic effects?

It looks like to me the "make people responsible and trust them" attitude
is the winning one. You said it many times about the CL vs C++ attitude:
CL trusts the programmer, whereas C++ distrusts them, and the result is
trustworthy CL programmers, and untrustworthy C++ programmers.
The same argument applies to free software vs proprietary software.

There will always be stupid and evil people; no policy will prevent that.
The question is how to make such people harmless to anyone but themselves.
Reminds me quite of F.A.Hayek's "the Road to Serfdom"...

>   of course it helps to deal with non-stupid people.  however, there are
>   lots and lots of license-restricted software products that doesn't need
>   any of this administrative bullshit.
And again comes the question of trust and guarantee:
you may have quite non-stupid partners today that provide
great software with great service.
But what if their marketing department decides that this software
is no more profitable and will no more be supported?
With free software, you just move to another service provider
(who may perhaps hire the employees of the former company).
With proprietary software, you just bite the dust;
no possible long-term warranty.

Former ILOG LISP users and developers unhappily know _perfectly_ well
what I'm talking about...
The motto of the AFUL is "liberté, stabilité, perennité":
liberty, stability, perenniality.
Oh, and if what the company sells that has so much value
is their great service, then it has nothing to fear
from "competitors" who'd just sell unsupported copies of the software...

>   I personally fail to see why people don't take this "getting jabbed by
>   your vendor" thing much more seriously.  if you're afraid of it, and you
>   don't tackle the issue head-on, is it because you _fear_ the vendor?  do
>   you actually _need_ products that will likely cripple you in the future?
>   (I don't think so.)  have you ever talked to the vendor and expressed
>   your concern?  if you haven't, do it now.  if you have and was dissed,
>   why do you still deal with them?  this is the labor union thing all over
>   again, except now with entities you'd expect were able to defend their
>   own interests much better.
Yes, we do fear the vendor. And every vendor behaves the same,
so you don't get much choice with proprietary software.
Think of it as meme stability: the meme of dissing users
is co-stable with the meme of proprietary software,
but not quite so with the meme of free software.
We DO organise in unions to fight vendors who diss us,
and the result is called (surprise!) free software.

> | 5) community-oriented.  From what I can tell, open source projects tend
>      to do an extremely good job of putting "customers" and  developers
>      together.  On the other hand, most commercial companies where I've
>      worked went out of there way to keep developers and customers apart
>      (counter-productive in my view).
>
>   yup, counter-productive in the extreme.  if you use a software tool for
>   developers, and you can't talk to the developers of the tool, you have
>   made a mistake in purchasing it.  however, this is not a function of
>   source access, but of smart people who actually care about what they do.
>
Again, see meme co-stability. Working with the developers
is not co-stable with proprietary software,
all the less as the software spreads and is used by more and more people.
I can't imagine one's favorite C compiler vendor providing developer contact
to all its customers, there are too many of them.
CL dooms itself in being a fringe language
if it claims providing this contact.
With widely spread proprietary software,
developer contact is part of the cost structure,
and is fought against by management.
With widely spread free software, developer contact is a service
that you sell; it's part of the profit structure and sought by management.

>   you have pointed at several issues that point to why people should choose
>   source-based products instead of shrink-wrapped products, and I agree
>   with all of them, but at issue is not source vs shrink-wrap, in my view:
>   at issue is a lot of incompetent people who are mortally afraid that if
>   anyone saw their source, they'd be exposed as the frauds they are, and I
>   actually believe that a certain major software company in Redmond, WA,
>   would be history the day its sources were released in a much more
>   important sense than any other company would fold if its trade secrets
>   were dispersed.  I have argued elsewhere that I think a big motivator in
>   the source-based software world is legitimate rejection of said company
>   and its extremely predatory behavior.

>   however, defense against idiocy
>   and evil is not an end in itself
No, but it's a necessary _beginning_.
Without it, don't even try to go further.

>   -- you have to have a pretty clear
>   picture of what you're fighting _for_.
Indeed. Free software is not the end-all, only the begin-all.
Dynamic software WILL win; it will win WITH free software, not against it.
Proprietary software has brought upon us the domination of
FORTRAN, COBOL, PL/1, C, C++. Static languages.
Free software has always developed its dynamic tools:
LISP (pre-Common; elisp; Scheme), shells, Perl, Python, etc.
Dynamic languages (of various quality).

Of course, there are exceptions: CommonLISP and Dylan are dynamic languages
that have been mostly developed as proprietary systems
(despite heroic free implementations); but they have limited success
among proprietary systems; they don't fit the proprietary software model
of separation between provider and consumer.
On the other hand, there are static free languages (SML, OCAML, Haskell),
but even they have interactive top-levels,
and they have a hard time capturing free software developer mindshare.

> | Why do I get the feeling I'll regret this?
>   beats me.  and you don't come with source, so I can't fix your problem.
And the fact that he does not come with source is just a fact of nature,
so there's nothing we can do, and we should think about real problems.
On the other hand the unavailability of source of computer software
is _not_ a fact of nature; it _is_ a problem, and it can be solved.

Best regards,

[ "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 ]
...so that IBM Java envangelist tells me "nothing spread as fast as Java",
to which I answer: "crack!"...