First of all I want to be sure that we are referring to the two example i reported in my previous mail... On Mon, Dec 10, 2001 at 08:41:45AM -0600, Colin Watson wrote: > > Yes, I know, but with the two examples i reported a DD do not need to > > explain the reason: in the first case, the change-log file should list the > > changes and their reason; in the second case, the existence of the > > package is the reason. > > Sending an email to the submitter would be glad, sure, but would > > introduce redundancy of information, and would be quite annoying. > > Mail the changelog to nnnnn-done@bugs.debian.org. Please. Otherwise the > submitter has to go hunting around for a changelog they often won't even > have installed locally to find out why their bug was closed. This is > wrong. I do not think it is a real problem: i mean... the submitter of the report receive a message in wich he/she can find the mail address of the one who closed the bug, so can simply ask for explanation (as it happened when i closed some ITP). Many DD never asked, they obviously knew why i closed the ITP. OTOH, using the close command inside a mail to the control server has the same effect as the Closes string inside the change-log file (AFAIK). > This isn't a database normalization problem. Minimizing redundancy is a > *bad* thing when you're interacting with real people who don't > necessarily see all the sources of information available. I do agree with this, but any debian user is able to download the source of a package and check the changelog, if they want to know why the DD closed the bug (supposing that the submitter has not installed the package, and/or do not want to install it). The point of view is different: i'm not minimizing the information, i do not want to bother the submitter sending him/her information he can easily obtain if he/she is interested. Do we want to suppose that any submitter is interested in the reason of the bug closure? I think we do not: at last me. I always suppose that if a DD closed the bug, he solved the problem: if it is not, it's my own interest to ask for a reason, discuss with the DD or whatever... > You can cc control mails to nnnnn-done@ if you want to close bugs too. But this is the same as the use of the close command, with the exception that the submitter will see all my operation on the bts... i'm not sure about it's usefulness. > I've never seen 'close' used in a good way that couldn't be done better > with -done. I do agree with this too: with a -done mail you have an entire mail body to explain the reason, but my small experience as DD showed me that about 90% of the bugs are of technical nature (change to the source package or to the debian files): a change-log row is enough IMHO and a close command should be have the same effect. I'm always referring to the kind of bugs i used as examples. > The alternative, of course, is to cause mails to control@ to be copied > to the submitter. It will not send any explanation too. cheers, -- Luca - De Whiskey's - De Vitis aliases: Luca ^De [A-Z][A-Za-z\-]*[iy]'\?s$
Attachment:
pgpqdPwwwfQtj.pgp
Description: PGP signature