[Axiom-developer] Re: OpenAxiom-1.2.1 released

[Aleksej Saushev]

root <address@hidden> writes:

>> I'm not going to do any work, that is rejected before planning stage.
>> It is a mere waste of time, I'm not going to live forever or wait forever,
>> even 5 years is too much.
>
> ...[snip]...
>
> Axiom is a long term project. The theme is the 30 year horizon.  The
> point is to make the code live despite the fact that the authors are
> dead.  Just two weeks ago we lost yet another Axiom contributor.

Any other active developer to stay when we lose you?
Judging from commit messages you're the single developer.

> A great deal of thought went into the design decisions and project goals.  
> All of those thoughts have been documented many times in this forum.

Sure, but who are other developers who share most of your views?
Who is going to fix bugs and to develop Axiom (pretty little used) after you 
die?

>> While I agree with some points, other points are simply asking for nightmare.
>> Autoconf may be bad, but what is there now is even worse. And what is going
>> to be is hardly any better.
>
> ...[snip]...
>
> What is there now is being simplified with each iteration.
> A large number of Makefiles are gone. Whole subdirectories are gone.
> The whole Makefile superstructure will disappear.
> Code is being re-structured, refactored, and documented.
> End user documentation is improving with each release.

Does Axiom build with CLISP or ECL on NetBSD as is right now?
It is 15 to 25% of all BSD users by different estimations.

> Exactly the points you raise are the reason FriCAS exists.
> FriCAS developers removed the literate form of files in many cases.
> FriCAS developers use autoconf and traditional code handling.
> FriCAS developers "want it now", not 30 years from now.
>
> If you REALLY want it now, I suggest you switch to Sage.
> Sage is moving very fast, builds everywhere with a single "make",
> (although not autoconf), has a huge number of active authors,
> and is released every week or so. They don't want lisp at all.
> But they bundle everything under the sun.

Do you know any impressions of Sage by real people? All those I know
prefer using individual components of Sage rather than Sage itself.

> I applaud your desire to use a computer algebra system.
> However, the project goals of Axiom clearly don't meet your needs.

This is simply delusion. I object several points of Axiom, which proved
to be wrong in practice, and you declare "project goals ... don't meet."

Was you ever told that you're idealist? that you live in ivory tower?
Honestly, this is the impression of you. I don't talk about your
professional skills those you start complaining of, I only assert that
you seem to lack understanding of target audience needs.

> <begin indignation rant>
>
> P.S. As for that "piece of shit" GCL.... GCL was developed under
> contract to IBM specifically for Axiom. I'm one of its (minor)
> authors.  Bill Schelter, the author, was one of the finest programmers
> I've known.  His death in 2000 is what caused me to call NAG about
> Axiom.  What body of work have you personally authored that so
> outshines GCL that you feel qualified to disparage Bill's work? What
> body of patches have you submitted to GCL that improves it?  His code
> is still used 9 years after his death and generates more efficient
> code for Axiom than any other lisp distribution.
>
> Bill sat at my office desk when he visited IBM to work. I once saw him
> complain about something Emacs did that was failing. He stopped work,
> downloaded the Emacs sources, found the problem, rebuilt Emacs, tested
> it, posted a patch, and went back to the GCL task. He has patches in
> Emacs, written, tested, and posted in the late 80s. He did open source
> work before anyone knew what it was, quietly, efficiently, and correctly.
>
> Bill is unable to defend himself against your remarks but I am.
>
> You seem to be stopped by trivial build problems. 

I have no time to fix "trivial" build problems. You made them not so
trivial, as do other developers.

> You refuse to consider posting even minor patches. 

I do not have much time to invest into problems that should be solved already,
exactly those problems that all other active "PanAxiom" developers consider
to be solved at the first place.

> You feel qualified to disparage work you don't personally understand.

I'm trained to understand real world, I am not a mathematician.

> You feel that wasting your time to help other people is a bad tradeoff. 

Who told you that? How many people did you help lately?

> You feel that the project goals should be redefined to suit your opinions. 

This opinion is based on real world experience. If you have facts to
support your views bring them here.

> You have no sense of vision beyond your own wants and needs. 
> Frankly, I'm not as impressed with you as you seem to be with yourself.
>
> Be that as it may, the Axiom mailing list is NOT the place to lightly
> disparage Bill's work. He was a great programmer and I feel a need to
> defend his reputation.

Like anyone attacks Bill at all.

Look, it is 2009 now, not 1989. GCL is outdated, not quite standard,
the least portable of all open source Common Lisp implementations,
one of the least supported ones too. Until quite recently, noone
answered messages on GCL lists for two years at least. Most projects
have time schedule shorter.

What of these has to do with Bill?
He died in 2000 by your own words, but everything of mentioned above
happens after.

And there're many things that changed in the real world in that time
since 2000: we saw significant glibc changes, we saw significant changes
in NetBSD and FreeBSD projects, which brought them to desktop usage,
same happens to Solaris as we talk now. How is it reflected in GCL releases?
When did the last one happen? 2004? I don't even remember now.

You accuse me in disparaging Bill's work as if he died last week.
Now come and charge me with AI winter of 70s. Get real, wake up.

All I say is that project goals should be adjusted if you want to
receive any acceptance outside yourself and (maybe your) friends.
You have some really nice ideas, but other ones are simply awful.

> <end indignation rant>

-- 
BECHA...
   CKOPO CE3OH...