On Sun, Mar 17, 2013 at 10:35:46AM +0100, Corinna Vinschen wrote: >On Mar 17 02:27, Christopher Faylor wrote: >> If we're going to do that then I'd like the actual maintainers to start >> generating packages rather than random other people. Otherwise chaos >> will ensue. > >I didn't know Yaakov is a random person.
I'll attempt to not follow your lead and respond to this matter-of-factly rather than hyperbolically: In cygwin-apps, we have Marco and Yaakov discussiong cmake packages. Neither of them is the cmake maintainer. The cmake maintainer isn't even weighing in, AFAICT. In cygwin-developers, we have people discussing dash and perl who are not the maintainers. There are versions of these programs available but I don't think the maintainers have had feedback into the packages. There is also a 64-bit version of binutils, using a different versioning scheme (a new trend), uploaded by someone who is not the binutils maintainer. If we're thinking about making something that looks like a release then I'd like to do the considerate thing and make sure that the people whose packages you are releasing are ok with what's being done. I, for example, am not ok with making a version of binutils available which uses a different versioning scheme. I don't want to have to worry about that in a future setup.hint file. Downloading and extracting tarballs is, IMO, different than making setup.hint'ed packages available and establishing a new release. If we make a release I think it should be a beta version of an actual 64-bit release rather than something that has to be wiped out and restarted when 64-bit goes live. cgf