From: Austin Ziegler <halostatue@...>
Date: 2005-09-30T21:16:38+09:00
Subject: Re: gems is a language change, not a pkging system (Re: Require Namepaces and RubyGems' effect on LoadPath problem)

On 9/30/05, Mathieu Bouchard <matju@artengine.ca> wrote:
> On Wed, 28 Sep 2005, Austin Ziegler wrote:
> > Again, I disagree, and I think that it's one of the more elegant things
> > that RubyGems has done. The *reality* is that software applications are
> > heavily dependent on version matches. Anything else is DLL hell. I can't
> > express how much I *hate* Linux, libc, and libstdc++ nonsense and
> > soversions. There's an incompatible break between 3.2.5 and 3.2.6 of
> > libstdc++ (I think those are the right versions) -- it was a fscking
> > *nightmare* for us last year.
> BTW, what was that incompatible break and how did it affect you? This is
> just so that we have a concrete example.

I don't recall offhand the specific code that changed, but we had code
that compiled differently under the different versions, and there was
a library call that we had thought we could count on in libstdc++ 3.2,
but there was a difference between 3.2.5 and 3.2.6 that made that
assumption crap. If I can find information about what we found and had
to do with work, then I'll post about it -- but it was one of the
dumbest things I've ever seen in all of my Unix development history
(which is extensive).

I have definitely come to believe that it is important for languages
to help with API version management, and most of that belief has come
from the open source world where a lot of things change API without
proper planning or warning.

-austin
--
Austin Ziegler * halostatue@gmail.com
               * Alternate: austin@halostatue.ca