[MPlayer-cvslog] CVS: main configure,1.1196,1.1197

[Ivan Kalvachev]

2006/5/12, Derek E. Lewis <dlewis at solnetworks.net>:
> On Fri, 2006-05-12 at 07:01, Ivan Kalvachev wrote:
> > There is no such policy.
> > And Diego is not the person who can decide this kind of stuff. We've
> > always tried to do our best to write the most portable code and
> > support as many architectures as possible.
>
> >From DOCS/tech/MAINTAINERS:
>
> <snip>
> MPlayer code:
>     * build system: Diego Biurrun
> </snip>
>
> Regardless, you should've griped then and not now when a serious issue
> is trying to be remedied.  I'll be submitting a patch today to the
> MAINTAINERS file, as well, to remove my name from the status of 'Solaris
> maintainer'.  I nominate you for that position.

I don't see what I have to do with Solaris. I am worried for the
non-Solaris cases.
Besides, maintainers _are_ developers, so you _must_ be able to commit
this by yourself.
(yes, I am urging Diego to give you write access.)

> > As for sysi86, was it standard (SysV) or customized (Solaris10 or
> > something like that)?
>
> I wouldn't know, but I see it as irrelevant.  Solaris is the only true
> SVR4 that MPlayer 'officially' supports, as I know it.  By the language
> of the code that was in question (main/loader/ldt_keeper.c), that was
> true, as well.

What we "officially" support means only that somebody have
successfully compiled MPlayer on that arch at least once and survived
to tell us. It is equally possible that it works out of the box and
nobody to have noticed that it is not "officially" supported.

It is not good practise to put artificial obstacles to the people that
may want to port it to their own weird system. (well, we may do one
exception for SCO case;)