Wouter Verhelst wrote: > A very good example of that is debhelper; nobody ever told anyone to use > it, yet most of our packages do, directly or otherwise.
Parts of Debian encourage experimentation, innovation, and evolution of better solutions: parts don't. That debian/rules is a flexible, standard interface, and the lack of any obstacles to providing code that hooks into it has allowed many approaches to be tried. Compare adding a new suite like testing. The barriers to innovation there are high, including needing set up a copy of the archive (or being one of the few trusted to work on the real one), but also one has to overcome collective innertia and convince everyone they should care about the new suite. I don't know if Debian has become more resistent to innovation. Could be that the easy areas are increasingly tapped out. The exciting potential of dpkg source v3 to me is that it potentially opens an area that had stifled most innopvation, by allowing subtypes of the source format to be developed. But this area is still relatively closed to innovation; dpkg's maintainers still need to sign off on new formats, and the v3 source handling in dak is AIUI unneccessarily limited/hardcoded to only supporting certian subtypes. -- see shy jo -- To UNSUBSCRIBE, email to debian-vote-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100325195143.ga4...@kitenet.net