[04:54] <CIA-19> ubiquity: evand * r1506 ma/: Merged with trunk.
[05:01] <CIA-19> ubiquity: evand * r1507 ma/ (d-i/manifest debian/changelog):
[05:01] <CIA-19> ubiquity: * Set the partition model after populating it in the KDE UI
[05:01] <CIA-19> ubiquity: * Automatic update of included source packages: console-setup
[05:01] <CIA-19> ubiquity:  1.16ubuntu4, kboot-installer 0.0.1ubuntu3, partman-newworld
[05:01] <CIA-19> ubiquity:  18ubuntu1.
[05:01] <cjwatson> hmm, odd commit message
[05:02] <evand> whoops
[05:02] <cjwatson> (not worth fixing)
[05:02] <evand> I probably should've stripped that first changelog item from it.
[05:03] <cjwatson> it shouldn't do that unless the line changed
[05:03] <cjwatson> maybe trailing whitespace or something
[05:28] <evand> curious, my debsign appears to be broken, but I have no problem signing things normally.
[05:30] <evand> nevermind
[05:32] <cjwatson> I whack it over the head with DEBSIGN_MAINT='Colin Watson <cjwatson@debian.org>' in ~/.devscripts
[05:35] <evand> hrm, should I bzr export before debuild'ing?
[05:35] <evand> to avoid including .bzr
[05:37] <thom> evand: -i should do you, shouldn't it?
[05:37] <cjwatson> you *can*, but it's easier to use -I.bzr -I.bzrignore -I.shelf to debuild
[05:38] <cjwatson> thom: native package so you need -I too
[05:38] <evand> ah, thanks thom and cjwatson
[05:38] <thom> ahr, right
[05:38] <cjwatson> (though for others, yes)
[05:38] <cjwatson> what I do is this in my shell startup scripts:
[05:38] <cjwatson> export DH_ALWAYS_EXCLUDE='CVS:.svn:{arch}:.arch-ids:.bzr'
[05:38] <cjwatson> and this in ~/.devscripts:
[05:38] <cjwatson> DEBUILD_DPKG_BUILDPACKAGE_OPTS="-i'(?:^|/).*~$|(?:^|/)\..*\.swp|DEADJOE|(?:/CVS|/RCS|/\.svn|/\.deps|\{arch\}|\.arch-ids|\.arch-inventory|\.bzr|\.bzrignore|\.shelf)(?:$|/)' -ICVS -I.svn -I\{arch\} -I.arch-ids -I.arch-inventory -I.bzr -I.bzrignore -I.shelf -uc -us"
[05:38] <cjwatson> DEBUILD_PRESERVE_ENVVARS=DH_ALWAYS_EXCLUDE
[05:38] <cjwatson> that gets the job done nicely and I don't think about it
[05:39] <evand> noted
[05:39] <cjwatson> might want to sync that -i nightmare rune up with whatever's currently in /usr/bin/dpkg-source - unfortunately there's no easy way to say "just add to the list of stuff you're ignoring rather than replacing it"
[05:45] <CIA-19> ubiquity: evand * r1508 ma/debian/changelog: releasing version 1.5.3
[05:54] <CIA-19> ubiquity: evand * r1509 ma/ (configure configure.ac): bump to 1.5.4
[06:45] <CIA-19> ubiquity: cjwatson * r2088 ubiquity/ (configure configure.ac d-i/manifest debian/changelog): merge from Evan
[06:47] <btm_> busybox mkswap supports only 2Gb?
[06:48] <cjwatson> btm_: according to mkswap(8), util-linux is the same
[06:48] <cjwatson> The
[06:48] <cjwatson>        maximum useful size of a swap area now depends on the architec
[06:48] <cjwatson>        ture.   It  is  roughly  2GiB  on i386, PPC, m68k, ARM, 1GiB on
[06:48] <cjwatson>        sparc, 512MiB on mips, 128GiB on alpha and 3TiB on sparc64.
[06:49] <cjwatson> (of course, the man page could be out of date)
[06:50] <btm_> Yeah, I saw that too and saw '99 in the man page.
[06:50] <btm_> xm3adm@dev3:~$ sudo mkswap /dev/system/swap
[06:50] <btm_> Setting up swapspace version 1, size = 8589930 kB
[06:54] <cjwatson> but can you swapon that?
[06:57] <btm_> yeah.
[06:57] <cjwatson> and when you do, how much does the kernel say it's actually using?
[06:59] <cjwatson> could be that busybox mkswap needs to be updated from util-linux though
[07:00] <btm_> well, it lists 8G in proc/swaps and proc/meminfo. As far as being used, it's unclear to me if that much swap is no longer effective.
[07:10] <cjwatson> hmm, ok, I think you're right, looks like a legit bug
[07:10] <cjwatson> could you please file it?
[07:10] <cjwatson> current upstream is totally different, and while it fixes the bug it looks pretty hard to backport
[07:11] <cjwatson> r15704 | landley | 2006-07-16 09:17:03 +0100 (Sun, 16 Jul 2006) | 2 lines
[07:11] <cjwatson> Changed paths:
[07:11] <cjwatson>    M /trunk/busybox/util-linux/mkswap.c
[07:11] <cjwatson> New mkswap.  Small and simple.  Might even work. :)
[07:11] <cjwatson> (also won't do UUIDs etc. so is not very useful for our purposes right now. silly upstream.)
[07:12] <cjwatson> though I guess the old version didn't either, that was a feature I added
[07:14] <btm_> Sure.