On Jun 12, 2008, at 8:43 PM, David R. Morrison wrote: > Greetings, fink developers. > > There have been a number of threads over the past few months about the > fate of the stable branch. I want to start a new one, in which I will > make a different proposal than the ones made in the past. > > From my point of view, the stable branch might well be named the > "bindist-candidate branch". This is where we can collect versions of > packages which we hope will be eligible for the next bindist, and work > on them as a group until they all build, have no missing dependencies, > validate, and so on. > > Now it is true that there hasn't been a bindist in two years, and > there has never been one for 10.5. However, I've heard the opinion > expressed that if we simply abandon the current stable branch at the > time pango-cairo is merged back to HEAD (i.e., dump all of unstable > into stable), this will bring us closer to the next bindist. > > In my opinion, it will bring us farther away. > > What's fantastic about the work that the pango-cairo-branch-team has > been doing is that the packages in that branch have been very > thoroughly vetted -- everything validates, the dependencies are all > there, and so on. We definitely need a coherent set of packages with > these characteristics if we want to release a bindist some day. > > However, there are lots of parts of the unstable tree which have *not* > been worked on by the pango-cairo-branch-team. Many of those packages > are truly unstable, and were we to dump them to a "bindist-candidate" > tree (i.e., the stable tree), we would set the clock back on a bindist > until such time as all of the bugs could be ironed out of *those* > packages. > > So here's my proposal. When the pango-cairo-branch is merged back, we > put all of the *active* files from that branch immediately into the > stable tree. Now this is going to cause problems for stable-only > users, since not all of the dependencies are there. However, on the > one hand folks on this list assure me that nobody uses stable-only > anymore, and on the other hand a few weeks work can restore all of the > needed dependencies. So I believe this is a sound strategy. > > Your feedback is welcome. > > -- Dave > >
I have a stable-only alternate Fink tree available. The following caveats apply: 1) My machine is a PowerPC, so I can't test Intel-specific issues. 2) It's a G4, so no G5-specific issues. 3) It's a desktop, and I'm going to be away from home for six weeks on assignment, so if it goes down it's a multi-hour drive for me to go back and fix it (my wife's not here due to economic realities) ------------------------------------------------------------------------- Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php _______________________________________________ Fink-devel mailing list Fink-devel@lists.sourceforge.net http://news.gmane.org/gmane.os.apple.fink.devel
