Re: Nomination for Common Lisp dictator

Matthias Buelow <mkb@xxxxxxxxxx> writes:

jimbokun@xxxxxxxxx wrote:

But by making Edi dictator, and Starter Pack the defacto standard
library, social pressure could be brought to bear on all the
distributions to include the Starter Pack libraries in one form or
another.

Just because you say so doesn't make it so.

Indeed.

And while it's important to understand that the suggestion may be made
with all good intent, there is another side to this, which I assume
is what Matthias is doing, in a more brief fashion than I ever seem to
manage myself.

Among other things, and with no intent to be either approving or
disapproving of any particular suggestion of a person or site or
library, comp.lang.lisp is not known to be representative of the Lisp
community, so asking here doesn't have any effect in terms of consensus.
A bunch of positive responses here mean nothing, just as a failure of
responses means nothing. It's interesting, of course, but it's not
proof of consensus or lack thereof, nor necessarily even a hint at what
might be.

Also, the question presupposes that appropriateness of saying what is
de facto standard. The normal definition of de facto standard is that
it doesn't require saying it, because people just know.

And then there's the question of whether a single standard is good, which
many (myself included) have raised as an issue--more on that below.

So as far as I can tell, the actual proposal here is just to afford the
name "standard" to a particular person's approach. Why not just let
that person (or any person) give a unique name to their item, as happened
almost by accident with David Gray and "Gray Streams", and say that the
Weitz Library has <whatever>.

The whole idea of the substandards thing I suggested a while back
(which has thus far not taken off for complicated administrative
reasons that could be oversimplified as either "available time" or
"economics") boiled down to this for me:

True standards are about two things:
(1) writing down the meaning of something and putting it in a publicly
visible place that won't change and will be easy to find
(2) getting people to agree

It used to be important to do both of these, but I came to believe
that the second part was less necessary in today's day [due to various
issues of how the business and computational worlds work] and is a
very expensive thing to avoid. [Not that skipping the step but saying
there was agreement, as I understand is the suggestion here, is a
substitute. Rather, when I say "getting people to agree" is
unnecessary, I don't mean we should fiat things, I mean we should
allow people to disagree and support multiple paradigms.]

Basically, I'm pretty sure the "getting people to agree" is a false
choice. There is very little cost of having multiple ways of doing
something. If people gravitate to one thing, great, but if they
don't, there is little cost of offering alternate options. And the
cost of getting them to agree can be quite high in many ways, not
always apparent.

So, to me, the cost of saying "dictator" and "standard" is one of
disrupting people who are already using something else and making them
bear useless peer pressure about what they're using, rather than
worrying about interchange standards for DATA. Data is really what
matters, not programs. There are myriad libraries for XML, for
example, but it doesn't matter. It matters that there is some library
that each person likes, or even two or three that can co-reside, and
it matters that everyone agrees on a data interchange model. Things
like that are more important in practice, but are not generally
language issues.

Standards are contentious and they are in many cases unnecessary.

(Someone once observed to me that the word "science" gets added to
things that people wish were sciences but aren't, like "computer
science". True sciences mostly don't need the word "science" because
people just know physics is a science. I see a corollary for
standard, especially when no standards body has been activated to
manage the standard--but even sometimes otherwise.)

If it turns out that all vendors are already including the relevant
items, then indeed, you have a case to make that it's a de facto
standard. But if not, then I don't think it's a fair initial
presumption...

Incidentally, something I am NOT interested in doing is holding things
back because my substandards idea doesn't have currency. I'm not
proposing that as an alternative and saying "wait". What I am saying
is that what bothers me is not people using these things if they are
getting used, it's the notion that a non-deliberative process that is
not obviously representative of the community is suggesting speaking
for a community that is not even clearly defined. There are several
missing steps there. The idea behind substandards was to side-step
all of that too, but the way I wanted to do that was to not declare
itself to be unique. If anyone, including Edi, with serious resources
to muster and interest wants to contact me privately about making that
work, I'll be happy to discuss it in more detail.

The two key things I'd want out of any non-deliberative standard site
would be:

- Pluralism/Inclusiveness. Permitting more than one way to do things.
Not locking out everyone after the first person to drive a stake in
the ground.

- Politics-Neutral. Fairly accommodating BOTH free / open source AND
also commercial interests [including those incompatible with GPL and
friends].

Any process that purports to speak for the community without identifying
the community and documenting a fair polling of the community it purports
to speak for is suspect.

Of course, anyone who makes a newly named item like Graham's ARC is
implicitly doing this because anyone is free not to rally around. As
long as he doesn't call his process a standard for what he detached
from in order to do his thing, I'm all for it. Then it's voluntary.
Otherwise, it's forced on people--as "dictator" correctly implies.

And I didn't even get into the philosophy of dictators. Benevolent and
correct dictators are always good. It's just that they can decay into
Malevolent or wrong, and then that can be less good...

This is just my personal opinion, and in addition is my opinion today.
I'm certainly open to discussion and it's not like I never change my
mind. I also emphasize that I claim no special right to dictate how
things go; I didn't even get to do that as the editor the spec, and I
certainly wouldn't do it now when my status is even less official.
But all of that said, I have put a fair deal of thought into this, so
I do have some fairly strongly held opinions that I'm open to sharing.
.

Relevant Pages

  • Re: Python does not play well with others
    ... I think this is also a good insight into why things are as they are within the core development section of the community, ... one aspires to contributing to the standard library. ... Libraries, on the other hand, ... The committee was captured by the template fanatics, ...
    (comp.lang.python)
  • Re: C89, size_t, and long
    ... libraries which have useful stuff). ... So because others here can't think of a good reason to use long instead of size_t and you can't prove there is never such a reason, we must all assume there could be a good reason to not use the mechanisms provided by the standard to deal with a problem? ... If you are claiming there is a reason to avoid the mechanisms the standard provides to allow portability then it is up to *you* to prove your point, not up to others to disprove it. ... but you can hardly blame the consequences on the Standard headers. ...
    (comp.lang.c)
  • Re: C89, size_t, and long
    ... libraries which have useful stuff). ... So because others here can't think of a good reason to use long instead of size_t and you can't prove there is never such a reason, we must all assume there could be a good reason to not use the mechanisms provided by the standard to deal with a problem? ... want to have standard headers included in their own headers which ...
    (comp.lang.c)
  • Re: Any Clojure users here?
    ... ABCL and I've been contributing to it for a while (though since some ... from the CL standard or from the Java API, ... You can use CFFI-based libraries in ABCL without ... Yes, but without a Clojure standard, it's hard to know what the "core" ...
    (comp.lang.lisp)
  • Re: comparing doubles for equality
    ... AFAIK there is no ANSI/ISO standard for it. ... Fortran 95 seem to have some built-in operators but it's not clear ... LIA,GIA,ICE libraries for interval methods in C++ from Delisoft ... Actually I've been for quite some months proposing getting together ...
    (comp.programming)