Re: FYI: `Document' qnx (Was: Various systems)

[Pavel Roskin]

Hello, Akim!

> Pavel> Please undo this patch. We should not document operating
> Pavel> systems in the Autoconf manual.
>
> Why not?  I mean, I understand people who would say it does not belong
> to the Autoconf Reference Manual, but otoh the Autoconf documentation
> aims at providing the user of Autoconf with the answers to the
> question she is likely to wonder.

I'm worried about the political side of the problem. Once QNX is there
other people will want to put other OS'es into the manual.

> Documenting the shells is one good thing, and I think documenting
> systems is good too.  Beginners should be aware there are things named
> AIX, HPUX etc. out there.

The problem is that the documentation for programs, especially for
commercial programs tends not to mention the existing problems, at least
from the point of view of an Autoconf developer, i.e. I don't expect to
see the description of the shell bugs on the HP-UX site.

Notably, the problem with the error status after foo=`bar` is not
described in the sh documentation on the QNX site.

> Pavel> What if QNX is ported to ARM? What if hundreds CPUs become
> Pavel> thousands? What if somebody buys QNX? What if it forks?
>
> I don't get your point.

I mean, you cannot document everything. The world changes. Restrict
yourself to the relevant facts. "Hundreds processors" are not relevant and
sound like advertizing. And I sound like RMS :-)

Document limitations, not achievements.

> That's not the only point, it's also about knowing where you are
> working, not just your tools.  It's about curiosity.

QNX is a good system and many of its features should be emulated by free
software, but the Autoconf manual is just a wrong place for that.

> Pavel> Everything else belongs to "Shellology" or "Limitation of usual
> Pavel> tools" or to yet to be written "Quirks in system headers" and
> Pavel> "Quirks in system libraries".
>
> I very much agree the real content is and will be there.  There must
> be listed and classified the failures we know, and certainly not in
> Shellology nor Systemology.

Ok. If you insist on having the Systemology section I'll try to put more
relevant content there. I'm updating my .plan

Regards,
Pavel Roskin