=== ApOgEE__ is now known as ApOgEE === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan === pan1nx is now known as info === info is now known as pan1nx [07:11] hm, any ARM users out there who've tried chromium-browser from lucid-security? i'm seeing a SEGV on launch and wondering if it's a known issue already [07:13] * achiang goes trolling through launchpad [07:15] hm, nothing interesting there [08:04] good morning === hannesw_ is now known as hannesw [08:49] dholbach: howdy [08:49] hi MTecknology [08:50] dholbach: how's it going? [08:50] for some reason I can't compile dumb frotz 2.43. i get: http://paste.ubuntu.com/565279/ when i do 'make dumb' [08:52] good good [08:57] scott__: it seems to use a function name (getline) which is also provided by the libc. You either have to rename it (and all it's calls) or check if it does the same as the one from libc and use it. [08:58] ok i think i try renaming the function. simple enough. [08:58] yep that seemed to be it thanks for the insight === fabo_ is now known as fabo === al-maisan is now known as almaisan-away === yofel_ is now known as yofel [12:17] Hm, what could be different from the buildd and my local sbuilder? I have an app that crashes when it's been built in the PPA, but not when I build it myself [12:18] the buildd seems to pull in more libraries [12:32] n/m [12:33] I was too fast. The newly built package wasn't yet available [12:34] I was too fast, and so I wasted 1 hour :-( === almaisan-away is now known as al-maisan === lionel__ is now known as lionel === Quintasan_ is now known as Quintasan [14:15] hello [14:16] hello ari-tczew [14:18] coolbhavi: how are you? [14:18] ari-tczew, m good mate, you? [14:20] coolbhavi: this was busy week (until today), I'm quite sleepy and tired [14:20] but got some free time for finish some ubuntu-related stuff and learn to warehouse designing [14:21] ari-tczew, same here mate, very tired [15:44] tumbleweed: Sorry for stealing your devweek talk title. ;) === al-maisan is now known as almaisan-away [15:52] Rhonda: lol :) [16:07] Was too tempting. :) [16:14] Rhonda, can i ask about Contribute To debian at here? [16:15] *Debian [16:16] udienz: Rather come to #debian-ubuntu on irc.debian.org (oftc) [16:17] Though, obviously it depends on how deep your question is, in general. [16:35] hi all !! [17:04] hello everyone, I'm starting this motu journey trying to fix LP #706221 [17:04] Launchpad bug 706221 in calibre (Ubuntu) "calibre : More info - Having a typo ( supports is misspelled as upports) " [Undecided,Confirmed] https://launchpad.net/bugs/706221 [17:04] my debdiff right now looks like this @@ -50,7 +51,7 @@ Calibre is primarily a ebook cataloging program. It manages your ebook collection for you. It is designed around the concept of the logical book, i.e. a single entry in the database that may correspond to ebooks in several [17:04] - formats. It also upports conversion from a dozen different ebook formats to [17:04] + formats. It also supports conversion from a dozen different ebook formats to LRF and EPUB. A graphical interface to the conversion software can be accessed easily by just clicking the "Convert E-books" button. . [17:04] @@ -85,7 +86,7 @@ Calibre is primarily a ebook cataloging program. It manages your ebook collection for you. It is designed around the concept of the logical book, i.e. a single entry in the database that may correspond to ebooks in several [17:04] ups [17:04] sorry for that [17:04] my debdiff looks like this http://paste.ubuntu.com/565461/ [17:05] can somebody explain why it has the same parragraph twice??? It's the right behavior [17:07] One is for the "calibre" package, and the other for the "calibre-bin" package. [17:08] persia: thank you [17:09] alucardni: btw, you would be better doing this change in Debian first (as it's quite minor). The branch is actually maintained on Launchpad, so you can get it with `bzr branch lp:~calibre-packagers/calibre/debian' and then propose a merge to that. [17:09] alucardni: that should probably go through Debian, I think the upstream maintainer is pretty responsive :) [17:10] micahg: Laney ok, I'll try to push it to the bzr branch. Thank you guys! [17:12] Sure, make sure you use a Debian version (ending in -2) and use 'unstable' as the distribution instead of 'natty'. [17:12] (and don't change the maintainer) [17:13] "unstable" rather than "UNRELEASED"? [17:14] UNRELEASED indeed, I selected the wrong bzr revision to check this package's policy [17:16] So all I have to do is get the bzr branch, apply my patch (with your recommendations on debian/changelog and debian/control), commit and propose for merging? [17:17] Ideally, yes. === udienz is now known as afghanux === afghanux is now known as udienz [18:02] persia: I followed the steps for Ubuntu Merge Proposal on https://wiki.ubuntu.com/PackagingGuide/Recipes/Debdiff [18:03] Except you used the branch Laney identified? [18:04] persia: yes, the branch Laney identified, only contains the debian directory [18:04] but I got this error when pushing to LP: bzr: ERROR: Invalid url supplied to transport: "lp:~josernestodavila/ubuntu/calibre/debian/fix-706221": No such distribution series calibre. [18:05] You want to push to "lp:~josernestodavila/calibre/debian-fix-706221" [18:05] You aren't making a change to Ubuntu. [18:06] oh! true, thank you againg [18:06] *again [18:08] it's working now :) [18:09] alucardni: you can do 'bzr bd -S' to build a source package, that should work [18:10] and 'bzr bd-do' should give you a complete source tree (in a subshell) that you can work with as normal [18:12] fta: i keep experiencing SEGV when launching chromium-browser on lucid/armel [18:12] achiang, can you get a backtrace? [18:13] fta: no, i tried using the -dbg package last night, but this machine is so small that reading in all the symbols causes the initial fork to fail with -ENOMEM [18:13] Add more swap :) [18:13] hm. highly specialized installer that doesn't create a swap partition... sigh [18:13] can't apport handle that? [18:14] fta: i downloaded the maverick version of the package and it works fine (after also cherrypicking the latest libgconf2-4 from maverick too) [18:15] hm, weird, packaging wise, it's exactly the same [18:15] achiang, Add a swap file post-install. If you're short disk, use a USB disk. The point is just to collect the backtrace. [18:16] i can use USB as swap? [18:16] neat-o [18:16] Sure. [18:16] Either a swap partition or a swapfile on a USB partition. [18:16] cool, i'll try that in a bit [18:17] my afternoon task, i suppose [18:17] Note that USB doesn't tend to be fast, meaning swapping over USB is for the patient. === bcurtiswx_ is now known as bcurtiswx [18:17] fta: i'll try and collect data and file a bug report later [18:17] Laney: both commands return uscan could not find the needed tarball. [18:17] bzr: ERROR: Unable to find the needed upstream tarball: calibre_0.7.43+dfsg.orig.tar.gz. [18:18] persia: this machine isn't fast to begin with, so i'm not too worried. :) [18:18] it seems upstream changed the location of tarballs [18:18] alucardni, apt-get source will give you the tarball, if nothing else works. [18:22] hm, on maverick+, i build with armv7=0, but not on lucid [18:23] What does armv7=0 mean in that context? [18:25] i think my test is wrong, it's supposed to be the opposite, build with armv7 everywhere but lucid [18:25] Lucid requires ARMv7 also: it was only pre-lucid that earlier instruction sets were supported. [18:26] Unless there's something special about the lucid chromium source that makes ARMv7 not work... [18:26] it triggers -march=armv7-a -mtune=cortex-a8 -mfloat-abi=softfp [18:27] Yeah, that should be applied for Lucid+ [18:27] Jaunty and Karmic had different defaults for armel, but they aren't so interesting at this point. [18:27] http://paste.ubuntu.com/565500/ [18:28] There's some work in Debian to define the armhf architecture: I don't know if that will end up in Ubuntu. [18:28] It would have different defaults. [18:28] i guess i need to see the backtrace before changing anything [18:29] I'd recommend building with arm_neon=0 though, as there are a number of cortex-a8 boards that don't support NEON. [18:31] the upstream defaults: http://paste.ubuntu.com/565501/ [18:32] i set arm_thumb=1 everywhere [18:32] and armv7=0 when "ifneq (,$(filter 10.10 unstable development,$(DEBIAN_DIST)))" [18:33] DEBIAN_DIST := $(shell lsb_release -ds | tr -d '()') [18:34] Very odd. Debian armel should *never* be ARMv7, and Debian armhf should *always* be ARMv7. Ubuntu 9.04 and 9.10 should not be ARMv7, and Ubuntu 10.04+ should be ARMv7 [18:34] iirc, asac was happy with those choices [18:35] And there are any number of devices that support ARMv7 but not NEON for armel (although I think armhf expects NEON: I'm not sure) [18:35] and noone complained [18:35] I suspect nobody complained is the important part :) === Amaranth_ is now known as Amaranth === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [21:04] fta: got a backtrace for my chromium SEGV, but not sure how useful it is [21:05] fta: wanna see it? [21:06] achiang, can you upload it in a bug? [21:06] fta: http://pastebin.ubuntu.com/565557/ [21:06] hm, all-threads [21:07] fta: i can make a bug, just wanted to run it by you quickly first to see if it was even worth it with such a weak stacktrace [21:09] fta: i don't have working apport on this machine, is there any information you'd like me to collect? [21:09] achiang, i don't get much arm bug reports, so i'm not sure what to expect for this arch. is it supported by apport? [21:10] oh [21:10] i provide some apport hooks in the package, can you try to run it? (it won't help for the crash though, just the context) [21:11] fta: how do i do that? [21:11] sorry, i'm not very familiar with the tool [21:11] it doesn't produce a core file or anything on the SEGV [21:11] python /usr/share/apport/package-hooks/chromium-browser.py [21:11] kill the gdb instance [21:12] ? [21:12] if you can collect those: https://wiki.ubuntu.com/Chromium/Debugging [21:13] and "info registers" [21:14] ok, thanks === barry is now known as flufl === flufl is now known as barry [21:21] fta: how do i figure out which package's -dbg package to install, given the top stack frame? [21:25] fta: ah cool, apport is running now