/srv/irclogs.ubuntu.com/2012/02/12/#ubuntu-motu.txt

=== JackyAlcine_ is now known as JackyAlcine
=== sagaci_ is now known as sagaci
=== almaisan` is now known as almaisan-away
=== almaisan-away is now known as al-maisan
=== al-maisan is now known as almaisan-away
AmpelbeinWith this http://paste.ubuntu.com/838760/, I still get the hardening flags in $CFLAGS. According to dpkg-buildflags' man-page, "export DEB_BUILD_MAINT_OPTIONS = hardening=-all" should disable hardening. What am I doing wrong?08:19
micahgwe have some default hardening in our toolchain, I wonder if that's supposed to be removed in this case08:20
Ampelbeinmicahg: If I call 'DEB_BUILD_MAINT_OPTIONS=hardening=-all dpkg-buildpackage' CFLAGS is set correctly so I guess it should.08:22
micahgyou have extra space around the first = in the paste, idk if that would make a difference though08:23
AmpelbeinIt's a makefile so that shouldn't matter.08:24
AmpelbeinIt doesn't.08:24
tumbleweedAmpelbein: to get the effect of DEB_BUILD_MAINT_OPTIONS you need to export CFLAGS et al yourself08:25
Ampelbeintumbleweed: Ah, ok. I was under the impression that dpkg-buildflags handles that for me.08:27
Ampelbeindebian bug 644412 sounds related08:27
ubottuDebian bug 644412 in dpkg-dev "dpkg-buildflags: use DEB_BUILD_MAINT_OPTIONS when including buildflags.mk" [Normal,Fixed] http://bugs.debian.org/64441208:27
tumbleweedAmpelbein: it handles it for you, if you use it08:28
tumbleweedsure, you can use buildflags.mk08:28
tumbleweedor just do it by hand (I do that in pypy)08:28
AmpelbeinWhat I don't get is: It works correctly if I do 'DEB_BUILD_MAINT_OPTIONS=hardening=-all dpkg-buildpackage'08:29
tumbleweedthat's because dpkg-buildpackage runs dpkg-buildflags08:37
tumbleweed(although that  behavior has been dropped in Debian)08:38
tumbleweedwe'll presumably follow suit in q08:38
micahgAmpelbein: don't forget to document the disabling of hardening flags on the security wiki page (I assume you trying disabling some of them before all)08:48
Ampelbeinmicahg: It's about a fpc using package (gearhead), which should not have hardening enabled in the first place as the programming language is pascal.08:49
micahgAmpelbein: hmm, if that's true, then we probably need a more generalized solution for this that disabling on a per package basis08:52
Ampelbeinmicahg: The build failure in question: https://launchpadlibrarian.net/92025261/buildlog_ubuntu-precise-i386.gearhead_1.100-2_FAILEDTOBUILD.txt.gz08:53
AmpelbeinThat's another thing I don't understand. According to the manpage of dpkg-buildflags, only in dh compat level 9 should the flags be exported. The package uses compat level 5...08:54
micahgwe have -fstack-protector in our toolchain by default amongst other things08:55
micahgdid you try just without that?08:55
AmpelbeinYes, fpc doesn't understand any -format, -format-security, -fstack-protector08:55
tumbleweedAmpelbein: we export the build flags in dpkg-buildpackage08:55
* tumbleweed digs out the e-mail announcing that08:55
micahgstill, there are about a dozen packages that use fpc, I'm wondering how any of them built befre08:57
tumbleweedAmpelbein: https://lists.ubuntu.com/archives/ubuntu-devel/2011-November/034351.html08:57
Ampelbeinmicahg: In the case of gearhead, the last build was october 201008:58
AmpelbeinWhere CFLAGS didn't include any of the hardening flags (https://launchpadlibrarian.net/48191679/buildlog_ubuntu-maverick-i386.gearhead_1.100-2_FULLYBUILT.txt.gz)08:59
Ampelbeintumbleweed: thanks, reading.08:59
micahgah, ok08:59
tumbleweedour gcc also does hardening flags by default, doesn't it09:00
* tumbleweed forgets09:00
micahgyeah, but this is fpc which is why it worked before, now the hardening has moved09:00
tumbleweedmicahg: thanks for reminding me tha tI should document the lack of hardening in pypy09:01
Rhondahmm, so what's the proper version for a lucid-backport?  (heading to the wiki to check herself …)10:42
Rhondahmm, either I missed it, or https://wiki.ubuntu.com/UbuntuBackports doesn't state so …10:44
Rhondaduh10:49
RhondaMake people more aware of backports, and make the backport request process less transparent10:49
Rhondashouldn't that read MORE transparent instead of LESS transparent?!?10:49
* Rhonda . o O ( from https://wiki.ubuntu.com/BackportsPromotion )10:49
Rhonda… not that the page looks in any sense relevant, or current. :)10:50
Zhenechoh hai Rhonda -- want to do some syncs for me again? :) https://bugs.launchpad.net/ubuntu/+source/pokerth/+bug/930906 https://bugs.launchpad.net/ubuntu/+source/monkeystudio/+bug/93090711:00
ubottuUbuntu bug 930906 in pokerth (Ubuntu) "Sync pokerth 0.9.3-1 (universe) from Debian sid (main)" [Undecided,New]11:00
ubottuUbuntu bug 930907 in monkeystudio (Ubuntu) "Sync monkeystudio 1.9.0.1-2 (universe) from Debian sid (main)" [Undecided,New]11:00
Zhenech(yes, yet another pokerth update, sorry, upstream is active)11:00
=== yofel_ is now known as yofel
Rhondahmmm, start to wonder what the build1 of monkeystudio was for11:02
ZhenechRhonda, libqscintilla 2.6.0 i guess11:02
Zhenechit bumped soname some month or two ago11:02
Rhondaah, that build1 was also available in debian, and not only in ubuntu?11:03
Zhenechyepp11:03
Zhenech+b1 there of course, but you get the point11:03
Rhondasubject: [ubuntu/precise] monkeystudio 1.9.0.1-2 (Accepted)11:04
Rhondasubject: [ubuntu/precise] pokerth 0.9.3-1 (Accepted)11:04
Zhenechthanks a lot11:04
Rhondade nada11:05
Rhondanow tell me what version I need to use :)11:05
Zhenech~bpo60+111:05
Zhenech:P11:05
Rhondadoes 1:1.10-1~lucid1 look proper?11:06
RhondaZhenech: ~bpo60+1 is done already ;)11:06
micahgRhonda: looks right for the official version, backportpackage can help though with an unofficial version to be superseded if backported in the archive11:07
micahgRhonda: why didn't you sponsor the syncs? (syncpackage -s)11:08
Rhondaoh11:08
Rhondaerm, because I forgot about that :)11:08
micahgheh, ok :)11:09
RhondaI would have thought that is done implied when I have the rights to11:09
RhondaBecause it does use launchpad anyway?11:09
micahgwell, you have rights to upload, but -s gives whoever you pass to it the credit for the sync11:10
Rhondaah, so -s zhench?11:10
Rhondaor rather, sargentd11:11
Rhondawill try to remember11:11
micahgyep11:11
Zhenechmh, reminds me I wanted to get ~evgeni on launchpad some time…11:12
Zhenechbut that means i have to kill my ppas11:13
Rhondacant you transition them?11:13
Zhenechcreating a new user and then moving around? yeah could work11:14
Zhenechlp is able to rename users though, they just cannot have ppas11:14
Rhondano, get the accounts merged11:14
Rhondaah11:14
Rhondaonly recently started with ppas11:14
Zhenechmh, the only really used ppa is the monkeystudio one and we could just move it to the monkeystudio group…11:17
Rhondawould make it more official looking, yes12:06
=== tsimpson_ is now known as tsimpson
dupondjeStill ok to do sync requests ?20:11
Ampelbeindupondje: Feature Freeze is on thursday, so yes.20:13
dupondjeok :) lets do some20:16
broderto be clear, FeatureFreeze doesn't mean we can't do syncs/merges anymore. it just means that either they have to only contain bugfixes (no new features) or they have to be approved by the release team (which isn't that hard, especially immediately after FF)21:10
dupondjebut its better to do it asap ofc ;)21:11
tumbleweedas long as you don't break enerything, sure21:12
tumbleweedeverything21:12
ajmitch'break everything' would be starting a new major library transition this week, without serious coordination21:13
dupondjenew packages also allowed? http://packages.qa.debian.org/d/daq.html21:37
dupondjewould like to get the prehistoric snort gone :)21:37
tumbleweeddupondje: nothing wrong with new packages (in moderation)21:41
MikerHello23:33
MikerAnyone there23:34
bregmano23:34
MikerNobody is chatting so I was just wondering23:34
MikerWhats up with you Mr Webb?23:35
MikerGrammys is on tonight23:37
=== ripps_ is now known as ripps

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!