savannah-hackers-public
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Savannah-hackers-public] Re: Access to GNU projects' webpages by GN


From: Yavor Doganov
Subject: Re: [Savannah-hackers-public] Re: Access to GNU projects' webpages by GNU webmasters
Date: Wed, 22 Dec 2010 00:01:18 +0200
User-agent: Wanderlust/2.15.5 (Almost Unreal) SEMI/1.14.6 (Maruoka) FLIM/1.14.9 (Gojō) APEL/10.7 Emacs/23.2 (i486-pc-linux-gnu) MULE/6.0 (HANACHIRUSATO)

Bernie Innocenti wrote:
> > There's mr
> 
> I tried it out. It's a very nice meta-vcs, but there's one annoyance:
> you need to register repositories before you can update them.

Ah, that's an overwhelming limitation, indeed (I've never used mr,
just thought it's worth mentioning).

> How could we make it to work within our new.py script?

I don't think I fully understand how it works, but one idea is to
probe for the VCS metadata (kinda what VC, or GNUN's GNUmakefile.team
does).  It's not easy as in VC because we're talking about remote
URLs, but since the URL templates are known and are unlikely to change
(or if/when they do, you'll know in advance), it is perhaps doable.

IOW, the script could try

cvs -d :pserver....
svn co ....
git clone ....
bzr branch ....
etc.

...sorted by popularity, perhaps.  For cool speedup a simple db can be
kept -- I guess projects will not change the VCS every day, so once
the right one is determined per project, the expensive per-project
check can be skipped (unless the checkout fails, which may mean that
the VCS has been changed -- but not necessarily).



reply via email to

[Prev in Thread] Current Thread [Next in Thread]