[rescue] GNU != Unix (was: PC v. PeeCees? (was: IBM hard

Jonathan C. Patschke jp at celestrion.net
Sun Jul 14 19:12:37 CDT 2002


On Sun, 14 Jul 2002, Kurt Mosiejczuk wrote:

> Really, we've all just gotten soft.  How many remember compiling
> software BEFORE autoconf? *raises hand*

Ick.  I'm just glad that autoconf replaced imake.  Now there was a truly
brain-damaged piece of utility software.

> THAT was a challenge.  I'd plug away at brain damaged makefiles
> (or good makefiles that didn't like my brain-damaged system).

I still have that problem.  Autoconf has a very, very stupid way of
looking for symbols in header files that occasionally breaks in
spectacular ways under IRIX, as possibly other OSes with multiple ABIs.

A patch is forthcoming, as soon as I can wrap my brain around PERL.

> Now we all get cranky if autoconf doesn't work exactly 100%
> right.  And we complain about the quality of the authors.

Well, a lot of the code -is- crap.  A lot of it is inconsistent, uses
"fringe" cases of that language standards, is poorly-formatted, has
stupidly-named variables and functions, isn't documented, and just
generally isn't fit for public consumption.

Now, before I hear "but, what do you expect for free?"...

I would -never- release code like that to the general public unless said
code is written for (and, thus, owned by) someone else, and that someone
else demands it to be released at a particular stage, regardless of its
status.

That said, if the code has -my- name on it, it's a matter of honor and
self-respect.  I Do Not release crappy code.  If it's not good enough for
me to print out and brag about, I'm not going to show it to anyone.

> Oh, and I'm not just tarring everyone else with the brush, I'm
> getting me also.  I don't like slogging through makefiles any
> more either =) (Although I'll do minor work if I feel the
> software is worth the effort).

[1] I need to say this because $previousEmployer pulled this shit, and
    then left my personal email address (instead of the standardized
    $project-bugs@$company.net) as the point-of-contact.  So, there is
    shitty code[2] out there with my name on it, much to my annoyance.
[2] In a piece of server-management software for Solaris, IRIX, and Linux,
    thus reinforcing Murphy's Law.

-- 
Jonathan Patschke
  "gnu: we aim to fuck up everything with the potential to not suck"
                                                   --alex j avriette



More information about the rescue mailing list