Re: error messages in the raw lintian.log
Lucas Nussbaum lu...@lucas-nussbaum.net writes: There are quite a lot of error messages in lintian.log: $ grep -v -E ^(X|O|W|N|I|E): /org/lintian.debian.org/logs/lintian.log Argument 4\n4 isn't numeric in numeric lt () at /org/lintian.debian.org/root/checks/debhelper line 249. Argument 4\n4 isn't numeric in numeric lt () at /org/lintian.debian.org/root/checks/debhelper line 285. Use of uninitialized value in numeric lt () at /org/lintian.debian.org/root/checks/standards-version line 139. Use of uninitialized value in numeric lt () at /org/lintian.debian.org/root/checks/standards-version line 139. Argument 5\n# $Id: compat,v 1.3 2007-07-29 12:49:04 joostvb Exp $... isn't numeric in numeric lt () at /org/lintian.debian.org/root/checks/debhelper line 249. Argument 5\n# $Id: compat,v 1.3 2007-07-29 12:49:04 joostvb Exp $... isn't numeric in numeric lt () at /org/lintian.debian.org/root/checks/debhelper line 285. tar: d4x-2.5.7.1.orig/share/themes/gnome/popup/remove.png: implausibly old time stamp 1970-01-01 00:00:00 Argument 4\n4 isn't numeric in numeric lt () at /org/lintian.debian.org/root/checks/debhelper line 249. Argument 4\n4 isn't numeric in numeric lt () at /org/lintian.debian.org/root/checks/debhelper line 285. tar: guile-1.8.5/build-aux/ltmain.sh: implausibly old time stamp 1970-01-01 00:00:00 tar: guile-1.8.5/guile-readline/ltmain.sh: implausibly old time stamp 1970-01-01 00:00:00 Have you considered redirecting them elsewhere, so they don't end up in the log? Well, sending them to the log was an improvement so that we could find them. However, a further improvement may be to use a separate log just for them. There was some discussion a while back about whether it was okay to just stick them in the log for now and we decided it was, but there's really no reason why we can't have a separate stderr log. (Raphael provided patches for all of those except the tar errors, I think, so they should be fixed once Lintian on lintian.debian.org is upgraded to 2.2.2.) -- Russ Allbery (r...@debian.org) http://www.eyrie.org/~eagle/ -- To UNSUBSCRIBE, email to debian-lint-maint-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: error messages in the raw lintian.log
Russ Allbery wrote: Lucas Nussbaum lu...@lucas-nussbaum.net writes: There are quite a lot of error messages in lintian.log: $ grep -v -E ^(X|O|W|N|I|E): /org/lintian.debian.org/logs/lintian.log Argument 4\n4 isn't numeric in numeric lt () at [...] Have you considered redirecting them elsewhere, so they don't end up in the log? Well, sending them to the log was an improvement so that we could find them. However, a further improvement may be to use a separate log just for them. More specifically, so we could find the errors and associate them with the package responsible; the latter constraint is what makes this more complicated than simply redirecting stderr to a separate log. Adam -- To UNSUBSCRIBE, email to debian-lint-maint-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: error messages in the raw lintian.log
Adam D. Barratt a...@adam-barratt.org.uk writes: Russ Allbery wrote: Well, sending them to the log was an improvement so that we could find them. However, a further improvement may be to use a separate log just for them. More specifically, so we could find the errors and associate them with the package responsible; the latter constraint is what makes this more complicated than simply redirecting stderr to a separate log. Oh, ack, yes. I completely forgot about that. Thank you! -- Russ Allbery (r...@debian.org) http://www.eyrie.org/~eagle/ -- To UNSUBSCRIBE, email to debian-lint-maint-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: error messages in the raw lintian.log
Raphael Geissert atomo64+deb...@gmail.com writes: Russ Allbery wrote: (Raphael provided patches for all of those except the tar errors, I think, so they should be fixed once Lintian on lintian.debian.org is upgraded to 2.2.2.) You are right; I didn't see the tar ones. I don't know how we could suppress those, maybe a bug report against those packages is the only option, as tar doesn't seem to have a ultra-quiet option (and even if it had, it doesn't sound like the best solution here). We could easily duplicate the check that tar is doing. We already do for binary packages; we just don't for source packages. Unfortunately, it's one of those annoying upstream bugs where there isn't much the Debian package maintainer can do to fix it unless they repackage the upstream source. -- Russ Allbery (r...@debian.org) http://www.eyrie.org/~eagle/ -- To UNSUBSCRIBE, email to debian-lint-maint-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: error messages in the raw lintian.log
Russ Allbery wrote: [...] We could easily duplicate the check that tar is doing. Sure. We already do for binary packages; we just don't for source packages. I have always wondered what the exact relation between that check and dak is. Unfortunately, it's one of those annoying upstream bugs where there isn't much the Debian package maintainer can do to fix it unless they repackage the upstream source. Yes, but maybe the maintainer did *something*. On guile's case, only the -1.8 tarball has that problem, 1.6 which is also in sid does have correct time stamps. I couldn't find upstream's 1.8 tarball, only the ones for the previous releases, so I downloaded the snapshot which has 1.8 in its name; as you may guess from my comment, the tarball I downloaded also had correct time stamps. By the way, bit off topic, shouldn't the source-contains-* checks better be severity: pedantic? There's no much maintainers can do about those but repack the tarball and/or bug upstream, not to mention that they usually don't affect the final .deb and when they do another check should catch those. Cheers, -- Raphael Geissert - Debian Maintainer www.debian.org - get.debian.net -- To UNSUBSCRIBE, email to debian-lint-maint-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: error messages in the raw lintian.log
Russ Allbery wrote: Lucas Nussbaum writes: [...] tar: d4x-2.5.7.1.orig/share/themes/gnome/popup/remove.png: implausibly old time stamp 1970-01-01 00:00:00 [...] tar: guile-1.8.5/build-aux/ltmain.sh: implausibly old time stamp 1970-01-01 00:00:00 tar: guile-1.8.5/guile-readline/ltmain.sh: implausibly old time stamp 1970-01-01 00:00:00 [...] (Raphael provided patches for all of those except the tar errors, I think, so they should be fixed once Lintian on lintian.debian.org is upgraded to 2.2.2.) You are right; I didn't see the tar ones. I don't know how we could suppress those, maybe a bug report against those packages is the only option, as tar doesn't seem to have a ultra-quiet option (and even if it had, it doesn't sound like the best solution here). Cheers, -- Raphael Geissert - Debian Maintainer www.debian.org - get.debian.net -- To UNSUBSCRIBE, email to debian-lint-maint-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: error messages in the raw lintian.log
Dropping Lucas since we're drifting afield of his original report. Raphael Geissert atomo64+deb...@gmail.com writes: Russ Allbery wrote: We already do for binary packages; we just don't for source packages. I have always wondered what the exact relation between that check and dak is. See check_timestamp in dak/process_unchecked.py (which in general is a good bit of code to be familiar with). git clone http://ftp-master.debian.org/git/dak.git to get the repository. The exact cutoff that dak uses is configurable. The value for Debian is: FutureTimeTravelGrace 28800; // 8 hours PastCutoffYear 1984; Yes, but maybe the maintainer did *something*. On guile's case, only the -1.8 tarball has that problem, 1.6 which is also in sid does have correct time stamps. I couldn't find upstream's 1.8 tarball, only the ones for the previous releases, so I downloaded the snapshot which has 1.8 in its name; as you may guess from my comment, the tarball I downloaded also had correct time stamps. Weird. By the way, bit off topic, shouldn't the source-contains-* checks better be severity: pedantic? There's no much maintainers can do about those but repack the tarball and/or bug upstream, not to mention that they usually don't affect the final .deb and when they do another check should catch those. Yes, I think you're right. I'll make that change. -- Russ Allbery (r...@debian.org) http://www.eyrie.org/~eagle/ -- To UNSUBSCRIBE, email to debian-lint-maint-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org