Re: Increasing regularity of build systems
On Wed, Sep 15, 1999 at 10:30:18PM +0300, Antti-Juhani Kaijanaho wrote:
> On Wed, Sep 15, 1999 at 01:08:22PM -0500, David Welton wrote:
> > I think that as many packages as reasonably possible should
> > migrate towards them. They work pretty well, but I don't believe
> > in forcing them on people if they are really opposed.
> So even if we did decide to do what you suggest, I am free to
> hand-roll my rules files like I've done since I joined?
Of course. I don't think it makes sense to say "you must use this".
More like "well, if you aren't doing anything different than everyone
else, why not use an API that is common and understood.
> > I'm not just talking about debhelper, but about rules files in
> > general. I think using debhelper is one part of this.
> I am with you in demanding well-written debian/rules files. I am
> not with you in saying that debhelper is the way to fix these - it's
> *a* way, sure, but not the only one, and you must be careful in not
> blaming people for not using debhelper.
I'm not blaming anyone. A lot of the broken rules files do use it.
It's not THE answer to the problem, but I think it's a good thing to
use where appropriate (and maybe to explain why you *don't* use it
should you go that route).
> > Policy is not the be all and end all of Debian - frankly, it doesn't
> > interest me much.
> How so? You are here proposing Debian policy, you're just doing it
> on -devel.
There is a technical problem that I wish to get people thinking about.
Policy is more of a mandate, anyway, not a "you ought to do this".
Maybe policy can be changed if we ever get really specific about what
*must* be done, but I dont see that anytime soon.
> > It's a general problem that manifests itself in
> > different ways that are difficult to discuss on a general level.
> Then give us specific examples. You did this in this message, which
> is good.
How can you talk generally when so many packages do so many different
things? It's not easy. We have a lot of bright hackers, and there
are a lot of 'cool hack' kinds of build systems going on, which are in
many cases quite different from one another.
> It is hard to discuss a problem which is not clearly defined (either
> by general definition or by specific examples)
I'll point out more as I see them...
David N. Welton ( Circa mea pectora
email@example.com ) multa sunt suspiria
http://www.efn.org/~davidw ( de tua pulchritudine
debian.org + prosa.it ) que me ledunt misere