On Wed, 4 Jul 2007, Frank wrote:
I don't think we will transition to svn any time soon. Every time I suggested to do that, the discussion turned into a discussion about how sh*tty Sourceforge is and that we should leave it and go somewhere else, and then everyone suggested his favourite "somewhere else" and then everything stayed as it was.
My big problem with SF for a while was the delay of several hours to the pserver mirror, especially back when that mirroring was completely down for several weeks (was it several months?). Now that's over and last time I checked there was no delay or it was very small.
My only other big problem, afair, is that I can't handle a list of people who have commit access to my particular projects, there has to be one big centralised list for the whole 'pure-data' project. In one sense, that's not much of a problem because I could use another SF project container. That wouldn't be in pure-data:/pd/externals/ anymore, but hey, neither is 'pd-gem'.
_ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal QC Canada
Hallo, Mathieu Bouchard hat gesagt: // Mathieu Bouchard wrote:
My only other big problem, afair, is that I can't handle a list of people who have commit access to my particular projects, there has to be one big centralised list for the whole 'pure-data' project. In one sense, that's not much of a problem because I could use another SF project container. That wouldn't be in pure-data:/pd/externals/ anymore, but hey, neither is 'pd-gem'.
Actually the long and growing project member list to me also is a problem. It's not an urgent problem, but I'd sleep better (as one of the admins) if there was a more fine grained access control. Most people don't need and many don't even want to have write access to everything. It's possible to mess things up even without intending too, just for example with a wrong "cvs import". The fact that this hasn't happened in the past is no excuse for the future.
The main problem with SVN on Sourceforge is, that there isn't even the basic access control for SVN, that SF has for CVS, (AFAIK). I don't think, that Savannah or - god beware: Google has the kind of access control that we could have if we'd roll our own repository e.g. at IEM.
Ciao
Actually the long and growing project member list to me also is a problem. It's not an urgent problem, but I'd sleep better (as one of the admins) if there was a more fine grained access control. Most people don't need and many don't even want to have write access to everything. It's possible to mess things up even without intending Indeed, but there is quite a lot of stuff on pd cvs that requires some programming/makefile skills to get working anyway, that needs "cleaning up" e.g. cxc doesn't compile from cvs (regexp bangfilt needs to be changed to bfilt in makefile).
I won't upload my own working makefile as the external is not my place to do so, but it would be a good idea for those that do not compile, or those that have no makefiles at all, if there were a way of "cvs-suggest" or something. We need an authoritative repository...it's no one person's responsibility to maintain all of cvs, but we could all contribute to its robustness perhaps...
Best, Ed
Lone Shark "Aviation" out now on http://www.pyramidtransmissions.com http://www.myspace.com/sharktracks
--------------------------------- Yahoo! Answers - Get better answers from someone who knows. Tryit now.
-
Ed Kelly -
Frank Barknecht -
Mathieu Bouchard