On Wed, May 01, 2002 at 12:48:43AM +0200, Josip Rodin wrote: > Well, I think it is serious, and not merely because someone somewhere has > made a document that says so, but because of the de facto standard that you > don't mesh binaries and configuration files. Of all seven hundred packages > on my system, there was one single "black sheep" that placed a binary among > the configuration files. In the large scheme of over a thousand > configuration files in my /etc/, there was this little deviant binary. > It crossed the line no other package crossed. This made me think it was a > serious bug, rather than normal. I already said I wasn't arguing about the severity of the bug vis a vis the official definition list. I'll stipulate that the problem is as gross, inelegant, and personally offensive as you please. The issue is for me is twofold: about the package maintainer being empowered to manage his own bug list for triage purposes, and about the limits of the Release Manager' or another developer's power to make decisions for another developer. > That's not true, katie sends out the closing messages when the package > is in the Debian archive. In my experience, it sends out the closing messages roughly contemporaneously with the time the package is moved into queue/accepted. Or has this changed very recently? Earlier this week, it was still the case. > It's not propagated to all the mirrors of the same archive, but it > nevertheless is in the archive. Even before NI, it sent the > bug-closing messages before the mirror sync was started. Which subdirectories of queue/ are part of the "archive" and which are not? -- G. Branden Robinson | I have a truly elegant proof of the Debian GNU/Linux | above, but it is too long to fit branden@debian.org | into this .signature file. http://people.debian.org/~branden/ |
Attachment:
pgpDaFuQQQcSW.pgp
Description: PGP signature