Re: Automatic closing of bugs
On Sat, 3 Dec, 2005 at 17:15:58 +, Colin Watson wrote: yaclc provides this. I also have bug #316385: [process.in] allow for process.in commands to restrict themselves to a specific package (like service.in). Presently the BTS doesn't seem to have this functionality, but it would probably not be a big change, since this already exists elsewhere. The source package thing mentioned is another small problem. -- Clear skies, Justin References [0] http://bugs.debian.org/316385 [1] http://lists.debian.org/debian-devel/2005/12/msg00033.html -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Automatic closing of bugs
On Sat, Dec 03, 2005 at 02:10:05AM +0100, Simon Richter wrote: The problem here would be that said test requires network connectivity, while the rest of lintian does not. Indeed. Adding such a test to dput seems to me a better idea. dput already has the need of network connectivity, it can simply perform the test before uploading and refuse to upload if the bug belongs to someone else. -- Stefano Zacchiroli -*- Computer Science PhD student @ Uny Bologna, Italy [EMAIL PROTECTED],debian.org,bononia.it} -%- http://www.bononia.it/zack/ If there's any real truth it's that the entire multidimensional infinity of the Universe is almost certainly being run by a bunch of maniacs. -!- signature.asc Description: Digital signature
Re: Automatic closing of bugs
On Fri, Dec 02, 2005 at 02:01:28PM -0500, Joey Hess wrote: A lintian-like test to see if the listed bugs match the package before uploading seems more useful to me. It would have prevented this particular problem. yaclc provides this. -- Colin Watson [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Automatic closing of bugs
On Fri, Dec 02, 2005 at 12:53:51AM +0100, Alexander Schmehl wrote: * Peter Samuelson [EMAIL PROTECTED] [051202 00:33]: This has been suggested before; the standard counterargument is what about closing an ITP? Then why not make a check (source package of bug and changelog are the same) or (bug to be closed is an ITP)? Or an ITA, O, or maybe any bug against wnpp or other pseudo packages as was suggested elsewhere in this thread? -Roberto -- Roberto C. Sanchez http://familiasanchez.net/~roberto pgp6zoopRPd6Q.pgp Description: PGP signature
Re: Automatic closing of bugs
On Fri, Dec 02, 2005 at 10:45:31AM +1100, Matthew Palmer wrote: Your mission, should you choose to accept it, is dig through the Perl code in merkel:/org/bugs.debian.org/scripts and work out how to add this functionality. grin - Matt Maybe it is a good thing that I am neither a DD (yet) and that I have no knowledge of Perl (other than to be able to execute a Perl script). grin I will defer to those with the proper access and expertise. -Roberto -- Roberto C. Sanchez http://familiasanchez.net/~roberto pgpx8r8c419wX.pgp Description: PGP signature
Re: Automatic closing of bugs
Hi, Matthew Palmer wrote: Your mission, should you choose to accept it, is dig through the Perl code in merkel:/org/bugs.debian.org/scripts and work out how to add this functionality. grin You can use package foo as a command to control@ to tell it ignore everything that does not affect bugs against foo. I am unsure whether comma notation is allowed (so katie could generate package bar,wnpp at the beginning of bug closing emails). Simon ---BeginMessage--- Processing commands for [EMAIL PROTECTED]: package test Ignoring bugs not assigned to: test tags 318229 = wontfix Bug number 318229 belongs to package xserver-xorg, skipping. kthxbye Stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) ---End Message--- signature.asc Description: OpenPGP digital signature
Re: Automatic closing of bugs
Simon Richter wrote: Matthew Palmer wrote: Your mission, should you choose to accept it, is dig through the Perl code in merkel:/org/bugs.debian.org/scripts and work out how to add this functionality. grin You can use package foo as a command to control@ to tell it ignore everything that does not affect bugs against foo. I am unsure whether comma notation is allowed (so katie could generate package bar,wnpp at the beginning of bug closing emails). Yes, but for multi-binary source packages, the package changelog doesn't specify which binary package the bug applies to, so katie would have no way of knowing whether to say (e.g.) package bar, package libbar1, or package libbar-dev. I think this functionality would have to be implemented on the BTS side. -- Kevin B. McCarty [EMAIL PROTECTED] Physics Department WWW: http://www.princeton.edu/~kmccarty/Princeton University GPG: public key ID 4F83C751 Princeton, NJ 08544 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Automatic closing of bugs
On Sat, 3 Dec 2005 03:03, Kevin B. McCarty wrote: Simon Richter wrote: Matthew Palmer wrote: Your mission, should you choose to accept it, is dig through the Perl code in merkel:/org/bugs.debian.org/scripts and work out how to add this functionality. grin You can use package foo as a command to control@ to tell it ignore everything that does not affect bugs against foo. I am unsure whether comma notation is allowed (so katie could generate package bar,wnpp at the beginning of bug closing emails). Yes, but for multi-binary source packages, the package changelog doesn't specify which binary package the bug applies to, so katie would have no way of knowing whether to say (e.g.) package bar, package libbar1, or package libbar-dev. I think this functionality would have to be implemented on the BTS side. Logically package foo should work where foo is a source package. (Because its quite common to eg tag multiple bugs pending at the same time). Andrew V. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Automatic closing of bugs
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Kevin B. McCarty wrote: Simon Richter wrote: Matthew Palmer wrote: Your mission, should you choose to accept it, is dig through the Perl code in merkel:/org/bugs.debian.org/scripts and work out how to add this functionality. grin You can use package foo as a command to control@ to tell it ignore everything that does not affect bugs against foo. I am unsure whether comma notation is allowed (so katie could generate package bar,wnpp at the beginning of bug closing emails). Yes, but for multi-binary source packages, the package changelog doesn't specify which binary package the bug applies to, so katie would have no way of knowing whether to say (e.g.) package bar, package libbar1, or package libbar-dev. I think this functionality would have to be implemented on the BTS side. It could however say package bar,libbar1,libbar-dev and any package from the list will do :-) Cheers Luk -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.2 (GNU/Linux) iD8DBQFDkIAc5UTeB5t8Mo0RAlk4AJ0SHO6jk/rkkvDotYUh2WlxexVK7wCfZBBf ycGt76qfSA+A2GMtjnraHUM= =FBlm -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Automatic closing of bugs
Hi Matt,
if that would help, I'd include a (n option that allows to) check via
bts2ldap + the attached script into dput.
That'd be less intrusive than changing the behaviour of Closes:, for
better or worse.
Kind regards
T.
--
Thomas Viehmann, http://thomas.viehmann.net/
#!/usr/bin/python
import sys, rfc822, ldap
from apt_listchanges import ttyconfirm
def parse_changes(changes):
chg_fd = open(changes)
check = chg_fd.read(5)
if check != '-':
chg_fd.seek(0)
else: # found a PGP header, gonna ditch the next 3 lines
chg_fd.readline() # eat the rest of the line
chg_fd.readline() # Hash: SHA1
chg_fd.readline() # empty line
if not chg_fd.readline().find('Format') != -1:
chg_fd.readline()
changes = rfc822.Message(chg_fd)
return changes
ch = parse_changes(sys.argv[1])
closedbugs = map(int, ch.get('Closes','').split())
pkgs = ch.get('Binary','').split()+[ch.get('Source','')]
if closedbugs:
print Querying ldap...
l = ldap.open(bts2ldap.debian.net, 10101)
l.simple_bind()
resid = l.search(dc=current,dc=bugs,dc=debian,dc=org,
ldap.SCOPE_SUBTREE,
(|%s)%(''.join(map(lambda x: (debbugsID=%d)%x, closedbugs))),
[debbugsPackage,debbugsID,debbugsTitle])
res = []
r = l.result(resid, 0)
while r and r[1]:
res.append(r[1][0][1])
r = l.result(resid, 0)
otherpkgsbugstouched = filter(lambda a: a['debbugsPackage'][0] not in pkgs, res)
if otherpkgsbugstouched:
print 'Changelog closes bugs of other packages:'
print ' '+'\n '.join(map(lambda y:
' '.join(map(lambda x: ' '.join(y[x]),['debbugsID','debbugsPackage','debbugsTitle'])),
otherpkgsbugstouched))
print Do you want to continue? [y/N],
if sys.stdin.readline()[1:] not in ['y','Y']:
sys.exit(1)
Re: Automatic closing of bugs
Thomas Viehmann wrote: if sys.stdin.readline()[1:] not in ['y','Y']: For added utility I might want to improve on that. Kind regards T. -- Thomas Viehmann, http://thomas.viehmann.net/ 'What'll we drink to?' Nick asked, holding up the glass. 'Let's drink to testing,' Bill said. 'All right,' Nick said. 'Gentlemen, I give you testing.' 'All testing,' Bill said. 'Everywhere.' 'Testing,' Nick said. 'That's what we drink to.' -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Automatic closing of bugs
A lintian-like test to see if the listed bugs match the package before uploading seems more useful to me. It would have prevented this particular problem. Perhaps I am sloppy, but I often find it useful to close some bug in a changelog that I might not necessary have taken the time or remembered to reassign to the package that actually ended up fixing it, especially when large refactorings (like the apt-setup split out of base-config) are going on. -- see shy jo signature.asc Description: Digital signature
Re: Automatic closing of bugs
Joey Hess [EMAIL PROTECTED] writes: A lintian-like test to see if the listed bugs match the package before uploading seems more useful to me. It would have prevented this particular problem. IMHO, is the best and easier alternative. -- O T A V I OS A L V A D O R - E-mail: [EMAIL PROTECTED] UIN: 5906116 GNU/Linux User: 239058 GPG ID: 49A5F855 Home Page: http://www.freedom.ind.br/otavio - Microsoft gives you Windows ... Linux gives you the whole house. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Automatic closing of bugs
On Fri, Dec 02, 2005 at 02:22:41PM +0100, Simon Richter wrote: Matthew Palmer wrote: Your mission, should you choose to accept it, is dig through the Perl code in merkel:/org/bugs.debian.org/scripts and work out how to add this functionality. grin You can use package foo as a command to control@ to tell it ignore everything that does not affect bugs against foo. I am unsure whether comma notation is allowed (so katie could generate package bar,wnpp at the beginning of bug closing emails). Except that katie doesn't send to control@, it sends to nnn-close@ (indicated in the header of the bug-closing messages in the BTS). I'm fairly sure that the BTS doesn't read commands off the top of messages sent to nnn-done@ (that could cause some killer confusion) but it does parse a pseudo-header (to get the Version: information to drive the version tracking feature). - Matt signature.asc Description: Digital signature
Re: Automatic closing of bugs
Hi, Joey Hess schrieb: A lintian-like test to see if the listed bugs match the package before uploading seems more useful to me. It would have prevented this particular problem. The problem here would be that said test requires network connectivity, while the rest of lintian does not. Simon signature.asc Description: OpenPGP digital signature
Automatic closing of bugs
I just had a bug that I opened (#339832) closed by a changelog entry in a new debconf upload. This is apparently a typo, as the changelog entry claims that the bug it was closing was related to a Swedish translation update. My bug was a wishlist bug against gmessage asking for it to become an alternative to xmessage. Is there a way to not allow changelog entries to automatically close bugs assigned to other packages? This seems like it might require modifying some infrastructure, but I am not sure what are the affected components or what I can do to help. -Roberto -- Roberto C. Sanchez http://familiasanchez.net/~roberto pgpX02ULlQWZ0.pgp Description: PGP signature
Re: Automatic closing of bugs
On Thursday 01 December 2005 23:45, Roberto C. Sanchez wrote: Is there a way to not allow changelog entries to automatically close bugs assigned to other packages? This sounds like a usefull restriction. I've seen enough cases where the wrong bug was closed to see the benefit of this. If the involved developers don't spot this a bug will be closed incorrectly and another remain open equally incorrectly. Maybe the check could be defined a bit broader as (not) belonging to the same source package. If a bug is not closed because of this rule, a message should probably be sent to the uploader suggesting to check his changelog and correct and close manually where needed. Cheers, FJP pgp6fFsqQyL7d.pgp Description: PGP signature
Re: Automatic closing of bugs
[Roberto C. Sanchez] Is there a way to not allow changelog entries to automatically close bugs assigned to other packages? This has been suggested before; the standard counterargument is what about closing an ITP? signature.asc Description: Digital signature
Re: Automatic closing of bugs
On Thu, Dec 01, 2005 at 05:45:53PM -0500, Roberto C. Sanchez wrote: I just had a bug that I opened (#339832) closed by a changelog entry in a new debconf upload. This is apparently a typo, as the changelog entry claims that the bug it was closing was related to a Swedish translation update. My bug was a wishlist bug against gmessage asking for it to become an alternative to xmessage. Is there a way to not allow changelog entries to automatically close bugs assigned to other packages? This seems like it might require modifying some infrastructure, but I am not sure what are the affected components or what I can do to help. Aaah, changelog typos. Brings back memories or trampling all over people's bugs with a quick slip of the finger... If you have a look at the message sent to 339832-close@, the close message does actually state the source package which closes the bug, so I think the fix is to make debbugs recognise Source: pseudo-headers in messages sent to -close@ and only close the bug if the values match. I could have sworn something like that was already done, but it fairly obviously isn't (at least not for Source: pseudo-headers). Your mission, should you choose to accept it, is dig through the Perl code in merkel:/org/bugs.debian.org/scripts and work out how to add this functionality. grin - Matt signature.asc Description: Digital signature
Re: Automatic closing of bugs
On Thu, Dec 01, 2005 at 05:33:11PM -0600, Peter Samuelson wrote: [Roberto C. Sanchez] Is there a way to not allow changelog entries to automatically close bugs assigned to other packages? This has been suggested before; the standard counterargument is what about closing an ITP? There can be exceptions for virtual packages (like wnpp). It's not like there's that many of them out there. Regards Javier signature.asc Description: Digital signature
Re: Automatic closing of bugs
* Peter Samuelson [EMAIL PROTECTED] [051202 00:33]: This has been suggested before; the standard counterargument is what about closing an ITP? Then why not make a check (source package of bug and changelog are the same) or (bug to be closed is an ITP)? Yours sincerely, Alexander -- http://learn.to/quote/ http://www.catb.org/~esr/faqs/smart-questions.html signature.asc Description: Digital signature
