| tumbleweed | micahg: it was regular FTBFS from nigelb and iamfuzz's list of linker-failure bugs worthy of jamming (nigelb blogged about the jam) | 00:06 |
|---|---|---|
| tumbleweed | (and I was playing with svammel) | 00:06 |
| micahg | what is svammel? | 00:06 |
| tumbleweed | micahg: automated bug filing script (doko used it to file most of the linker related ftbfs bugs in launchpad) | 00:07 |
| micahg | ah, cool | 00:07 |
| alegomaster | I am sort of confused | 01:24 |
| === pfifo is now known as bfifo | ||
| === bfifo is now known as pfifo | ||
| samd | hi, how could i make a deb package to pop a "do you want to reboot?" dialog after installing the deb | 05:24 |
| Rhonda | ScottK: Yes, that's one of those three. | 06:30 |
| arand | When's the last Automatic sync from Debian happining, still time for a package just accepted in Debain to sync over without specific request? | 10:21 |
| micahg | arand: should be tomorrow according to the release schedule | 10:21 |
| arand | micahg: Yeah, I guessed there would be one last sync on the 30th, just making sure :) | 10:25 |
| ScottK | Rhonda: Application endorsed. | 13:32 |
| Rhonda | Thanks! | 13:33 |
| tumbleweed | I wibblymat started ftbfs jamming :) | 15:17 |
| tumbleweed | wibblymat: sagan uses 3.0 (quilt), I think you should make that change with a quilt patch | 15:18 |
| Rhonda | tumbleweed: I thought dpkg-source does create a quilt patch automatically when it finds modified upstream sources | 15:19 |
| Rhonda | From my understanding that's one of the core features of v3? | 15:19 |
| tumbleweed | Rhonda: it does, but unless one tidies those up, they are ugly (even lintian complains about them these days) | 15:20 |
| Rhonda | Your wording sounded differently, thanks for the explenaition. | 15:22 |
| tumbleweed | thanks for making me explain then :) | 15:22 |
| tumbleweed | anyway, as this is wibblymat's first upload (that I can see), I'll take care of it | 15:22 |
| nigelb | I forgot how much I liked looking at pbuilder terminal output scroll by ;) | 15:23 |
| wibblymat | tumbleweed: I'm pretty new to packaging and I'm nto really sure what quilt is... is there a primer somewhere? | 15:24 |
| Rhonda | Is it in green on black in strange hyroglypehes, nigelb ? | 15:24 |
| tumbleweed | wibblymat: https://wiki.ubuntu.com/PackagingGuide/PatchSystems or https://wiki.ubuntu.com/Kubuntu/Ninjas/QuiltMagic | 15:24 |
| nigelb | https://wiki.ubuntu.com/PackagingGuide/Complete | 15:24 |
| nigelb | Rhonda: No :) | 15:25 |
| wibblymat | Yep, saw the notice about the jam the other day, saw the bugs go in last night, just dived in. Didn't see any activity about it in here though so was thinking it was just me :) | 15:25 |
| nigelb | I'm just about to start :) | 15:25 |
| nigelb | well, actually I'm building one of the packages I was working on | 15:26 |
| nigelb | Hopefully, I'll get help to finish this one today | 15:26 |
| tumbleweed | wibblymat: it *was* just you, but we had to start somewhere :P | 15:26 |
| Rhonda | nigelb: Do you understand the reference, or are you too young for that? | 15:27 |
| nigelb | Rhonda: Matrix wasn't that long ago :) | 15:27 |
| Rhonda | It feels like :) | 15:27 |
| nigelb | Rhonda: *gasp*, 13 years. | 15:27 |
| Rhonda | There were all harry potter movies since. | 15:28 |
| Rhonda | See. | 15:28 |
| tumbleweed | eek, I'm feeling old now | 15:28 |
| Rhonda | There is this youngster from over here who back then worked on nmap which shortly was seen in the movie | 15:29 |
| Rhonda | … and he is far from a youngster these days. | 15:29 |
| nigelb | nmap is the port scanner right? | 15:29 |
| Rhonda | yes | 15:29 |
| highvoltage | there are kids who do homework assignments on 9/11/2001... and it happened before they were born! | 15:29 |
| nigelb | Ok, out with nostalgia! | 15:30 |
| LetoThe2nd | hello! I just noticed that the gsoap package is heavily outdated. any chances of bumping the version a bit? | 15:35 |
| tumbleweed | wibblymat: I think LDFLAGS should also have gone before LDADD and LIBS | 15:35 |
| Rhonda | LetoThe2nd: Given that the package has been uploaded twice so far by someone else than the maintainer in Debian - I guess it might be looking for someone new to look after it. | 15:37 |
| * Rhonda . o O ( And it is the same since the lenny release in Debian, yuck ) | 15:38 | |
| LetoThe2nd | Rhonda: yea, thats what i also suspected reading between the lines :-/ | 15:38 |
| * Rhonda . o O ( uh oh, I know the maintainer personally ..... ) | 15:38 | |
| LetoThe2nd | Rhonda: poke him/her with a sharp stick! :-) | 15:38 |
| Rhonda | Haven't heard or seen him since a long time though. He's just also from Austria like me. | 15:38 |
| LetoThe2nd | hm, if the stick is long enough i can poke myself ... (bavaria) | 15:39 |
| Rhonda | There is this strange thing in Austria with DDs, they seem to fade away without a trace at some point. | 15:39 |
| Rhonda | off | 15:39 |
| wibblymat | tumbleweed: I'm not really a C developer so I just did what worked. Is putting the LDFLAGS just tidier or does it make a difference to the output? | 15:40 |
| nigelb | cdbs: ping | 15:41 |
| jtaylor | has somebody already looked at cppcheck? | 15:41 |
| tumbleweed | wibblymat: LDFLAGS should come before the shared objects (LIBS and LDADD). gcc + ld didn't care about this in the past, so many C developers don't know this :) | 15:41 |
| LetoThe2nd | Rhonda: sorry? | 15:41 |
| nigelb | Rhonda: You're in Austria, for some reason I had it my head that you're in Germany :) | 15:41 |
| tumbleweed | yes, some LDFLAGS, like --as-needed act on things that come after them | 15:42 |
| tumbleweed | wibblymat: Here's what I did: lp:~stefanor/ubuntu/oneiric/sagan/ftbfs-jam | 15:42 |
| tumbleweed | wibblymat: what's left now before uploading to Ubuntu is to forward the bug to Debian and/or upstream | 15:43 |
| Laney | anyone know how to make sbuild honour http_proxy? | 15:43 |
| tumbleweed | Laney: you can tell it what environment variables to pass through (I thought) | 15:43 |
| Laney | aha | 15:45 |
| Laney | in schroot.conf maybe | 15:46 |
| Laney | nope :/ | 15:48 |
| jtaylor | cppcheck fixed, forwarding patch to debian | 15:50 |
| nigelb | cdbs: ping | 15:51 |
| nigelb | cdbs: I'm setting up tarmac for community web projects. Let me know if you want it done for hof. | 15:52 |
| * tumbleweed needs to stop getting distracted | 15:53 | |
| nigelb | okay, buildign scratchbox | 15:54 |
| jtaylor | any specific usertags that should be used for as-needed bugs? | 15:56 |
| tumbleweed | as-needed sounds good | 15:58 |
| jtaylor | hm debian-gcc and ld-as-needed exists | 15:58 |
| jtaylor | I'll use that | 15:58 |
| tumbleweed | they weren't auto-tagged, because they aren't the only thing that can create undefined references, I think | 15:58 |
| Laney | User: debian-gcc@lists.debian.org | 15:58 |
| jtaylor | see http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=605839 | 15:58 |
| Laney | Usertags: ld-as-needed | 15:58 |
| ubottu | Debian bug 605839 in fontforge "FTBFS with ld --as-needed" [Normal,Fixed] | 15:58 |
| tumbleweed | neither one already exists | 16:00 |
| tumbleweed | (in ubuntu) | 16:00 |
| wibblymat | tumbleweed: Ok, had a play with quilt. Looks like there is no (easy) way to do it retrospectively if you've already changed the code though? | 16:01 |
| tumbleweed | yeah, no easy way | 16:01 |
| jtaylor | should the forwarded patch be registered somewhere in ubuntu? | 16:02 |
| wibblymat | Ok, will just have to remember to what-patch before I start :) | 16:02 |
| tumbleweed | wibblymat: also, Makefile.in is an auto-generated file, one wants to fix configure.in and regenerate it | 16:02 |
| wibblymat | As my old maths teacher used to say: You only learn from your mistakes, and boy are we learning today! | 16:04 |
| tumbleweed | which gets a little more complicated | 16:04 |
| jtaylor | ah there is a bug in ubuntu I can use | 16:04 |
| nigelb | okay, so I moved things around the makefile for scratchbox2, but it still fails. | 16:04 |
| jtaylor | when stupid launchpad would react! ._. | 16:04 |
| nigelb | Build log http://paste.ubuntu.com/635102/, and the diff of what I did http://paste.ubuntu.com/635103/ | 16:04 |
| nigelb | If anyone can help figure out, that'd be rad | 16:06 |
| tumbleweed | wibblymat: excuse me, I must have had my head in the clouds. The issue here is that configure is being called with a LDFLAGS containing libraries. That's wrong. So, change that to LIBS in debian/rules | 16:09 |
| tumbleweed | no need for any patches | 16:09 |
| wibblymat | tumbleweed: | 16:16 |
| wibblymat | tumbleweed: lp:~wibblymat/ubuntu/oneiric/sagan/ld-as-needed more like it | 16:16 |
| wibblymat | ? | 16:16 |
| tumbleweed | yes, looks good | 16:17 |
| tumbleweed | but aim it at natty rather that oneiric | 16:18 |
| tumbleweed | err other way around :) | 16:18 |
| nigelb | tumbleweed: could you look at my scratchbox build failure? | 16:18 |
| tumbleweed | can do, one minute | 16:18 |
| nigelb | \o/ | 16:19 |
| * jtaylor doing sslsniff | 16:23 | |
| wibblymat | Ok, new merge proposal for sagan (#803194). I'll look at the other two pacakges I did and check I haven't made the same mistakes! | 16:24 |
| tumbleweed | wibblymat: Thanks for your contribution to Ubuntu (uploaded) | 16:27 |
| nigelb | wibblymat: \o/ Congrats! | 16:27 |
| tumbleweed | wibblymat: I also added a space between LP: #XXXX (otherwise it won't close the bug) | 16:27 |
| wibblymat | Woo :) | 16:28 |
| tumbleweed | wibblymat: verified that my xfe suggestions work | 16:28 |
| wibblymat | tumbleweed: Ok, working on that now | 16:30 |
| tumbleweed | nigelb: that's a weird amkefile | 16:31 |
| nigelb | tumbleweed: yeah :| | 16:31 |
| nigelb | Its also weird package | 16:32 |
| nigelb | packaged natively, then made non-native, no orig.tar | 16:32 |
| tumbleweed | that's probably uploader-error | 16:32 |
| jtaylor | was there some change in launchpad? | 16:33 |
| jtaylor | it does not work with opera anymore :/ | 16:34 |
| jtaylor | and firefox also has some weird artefacts | 16:34 |
| tumbleweed | nigelb: you want -I in CFLAGS | 16:35 |
| ruhil | i just upgraded natty to oneiric and could not find the tray. Is it not there in oneiric or have i messed up something? | 16:36 |
| nigelb | tumbleweed: ah, move all the -I flags back to CFALGS? | 16:37 |
| nigelb | *CFLAGS | 16:37 |
| jtaylor | sslsniff patch forwarded | 16:37 |
| * micahg thanks jtaylor for doing sslsniff | 16:38 | |
| tumbleweed | nigelb: yeah | 16:38 |
| tumbleweed | wibblymat: would you mind forwarding the sagan patch to debian, as it was a bug in the packaging? https://wiki.ubuntu.com/Debian/Bugs#Using_submittodebian_to_forward_patches_to_Debian | 16:41 |
| jtaylor | hm sslsniff is in the archive but not sslstrip :( | 16:41 |
| nigelb | hrm, now the debuild -S -sa fails. | 16:46 |
| wibblymat | tumbleweed: Sure thing | 16:46 |
| nigelb | http://paste.ubuntu.com/635132/ | 16:46 |
| tumbleweed | nigelb: looks like a syntax errro | 16:47 |
| * jtaylor alienblaster | 16:47 | |
| nigelb | tumbleweed: doh, right. | 16:47 |
| nigelb | fixed | 16:47 |
| nigelb | Everytime, I type my GPG key password, I remember james_w's really really long password that he tried to type at UDS :) | 16:48 |
| Laney | some of my passphrases include compose key sequences | 16:49 |
| tumbleweed | nigelb: -uc -us ftw | 16:49 |
| Laney | they take a few attempts usually :-( | 16:49 |
| nigelb | tumbleweed: Got too used to signing it :) | 16:50 |
| nigelb | bah | 16:51 |
| nigelb | build failed again | 16:51 |
| jtaylor | alienblaster forwarded | 16:52 |
| nigelb | latest build failure http://paste.ubuntu.com/635135/, and my smaller patch http://paste.ubuntu.com/635136/ | 16:54 |
| jtaylor | nigelb: can you post the compiler line that produces that error | 16:56 |
| nigelb | jtaylor: the makefile line? | 16:57 |
| jtaylor | no the expanded line when executing the makefile | 16:57 |
| jtaylor | you can probably get it with make VERBOSE=1 | 16:57 |
| nigelb | jtaylor: manually? | 16:59 |
| nigelb | like without pbuilder? | 16:59 |
| jtaylor | yes, but maybe the makefile line helps me too | 16:59 |
| nigelb | http://paste.ubuntu.com/635139/ | 17:00 |
| nigelb | this is the makefile line | 17:00 |
| jtaylor | that line has no libraries linked | 17:00 |
| jtaylor | are they placed in LDFLAGS? | 17:00 |
| nigelb | I dont see LDFLAGS defined | 17:00 |
| jtaylor | do you know where that symbol is defined? | 17:01 |
| nigelb | probably in the parent makefile | 17:02 |
| nigelb | looking | 17:02 |
| jtaylor | no I mean the undefined reference | 17:03 |
| jtaylor | urg that package ahs no orig.tar | 17:05 |
| jtaylor | but its not native | 17:05 |
| nigelb | jtaylor: yeah | 17:06 |
| nigelb | jtaylor: its a messed up package | 17:06 |
| nigelb | all I see is "LDFLAGS += $(MACH_CFLAG)" | 17:06 |
| jtaylor | the fix is to reorder the dependencies in that makefiel | 17:10 |
| jtaylor | the .o before the .so | 17:10 |
| nigelb | ahh | 17:11 |
| nigelb | jtaylor: the makefile in utils? | 17:12 |
| jtaylor | yes | 17:12 |
| nigelb | okay, I don't see .so files | 17:12 |
| jtaylor | - $(D)/sb2-interp-wrapper: preload/libsb2.$(SHLIBEXT) $(D)/sb2-interp-wrapper.o | 17:13 |
| jtaylor | + $(D)/sb2-interp-wrapper: $(D)/sb2-interp-wrapper.o preload/libsb2.$(SHLIBEXT) | 17:13 |
| nigelb | ahh | 17:13 |
| nigelb | further proves that I need to learn more about gcc and make :( | 17:13 |
| jtaylor | the .o file needs the symbol in the library (.so) | 17:14 |
| jtaylor | as-needed is positional, so when a so comes before the objects needing it it will drop the library | 17:14 |
| jtaylor | the .o needing the symbols must come before thy library providing them | 17:14 |
| tumbleweed | wibblymat: it's "ld --as-needed" or "-Wl,--as-needed" rather than "--ld-as-needed" | 17:18 |
| nigelb | jtaylor: thanks for the quick help and the lesson :) | 17:18 |
| nigelb | scratchbox fixed \o/ | 17:19 |
| wibblymat | tumbleweed: Ugh, where did I write that? | 17:22 |
| * jtaylor pads | 17:23 | |
| tumbleweed | wibblymat: xfe | 17:23 |
| tumbleweed | wibblymat: I'd also mention that you had to use dh_autoreconf | 17:23 |
| tumbleweed | wibblymat: and $@ comes before the --with (yes, dh is weird) | 17:24 |
| wibblymat | Ah, now thats interesting. When I put it before $@ I got errors | 17:24 |
| wibblymat | About "options before sequence" or somethign similar | 17:25 |
| tumbleweed | wibblymat: that's exactly the error it gives now | 17:25 |
| wibblymat | Oh, wait, you're agreeing with what I said. I think you just have an old branch | 17:26 |
| wibblymat | I've been working on that one | 17:26 |
| hyperair | um $@ comes after the --with. | 17:26 |
| wibblymat | old revision of the same branch, rather | 17:26 |
| hyperair | that's how i've always done it. | 17:27 |
| tumbleweed | hyperair: compat 8 or later doesn't allow that | 17:27 |
| hyperair | tumbleweed: oh seriously, wtf for? | 17:27 |
| wibblymat | Oh, hang on, I may have accidentally regressed | 17:27 |
| tumbleweed | stylistic reasons? :) | 17:27 |
| hyperair | meh =\ | 17:28 |
| wibblymat | tumbleweed: Ok, it was me after all, I hadn't committed that change after all | 17:28 |
| tumbleweed | wibblymat: btw, did you submit the sagan patch to debian? (or did you run into trouble with reportbug?) | 17:30 |
| jtaylor | pads forwarded | 17:30 |
| wibblymat | tumbleweed: It should be forwarded | 17:32 |
| jtaylor | all so simple problems, it takes longer to write the email than to write the patch :/ | 17:33 |
| wibblymat | tumbleweed: I used jtaylor's cppcheck report as a template :) | 17:34 |
| nigelb | okay, scratchbox forwarded | 17:37 |
| tumbleweed | wibblymat: oh, I see it now | 17:37 |
| tumbleweed | great | 17:38 |
| nigelb | and debdiff uploaded | 17:39 |
| jtaylor | oO interesting a new problem | 17:39 |
| jtaylor | tenance needs -fopenmp | 17:39 |
| jtaylor | weird it worked before | 17:40 |
| tumbleweed | wibblymat: OK, xfe patch looks good. This one should be sent all the way upstream (and to debian, if possible) | 17:40 |
| jtaylor | anyone know what fopenmp does besides linking gomp? | 17:41 |
| jtaylor | hm nevermind its a classical as-needed problem afterall | 17:43 |
| jtaylor | tenance forwarded | 17:49 |
| tumbleweed | wibblymat: vacation needs a space between LP: and # | 17:50 |
| tumbleweed | jtaylor: you are picking these off rediculously fast :) | 17:50 |
| jtaylor | its all just library reordering | 17:50 |
| tumbleweed | yeah, that is a fair number of them | 17:51 |
| nigelb | picking up opencity | 17:51 |
| jtaylor | still ought to forward the patches upstream though | 17:51 |
| tumbleweed | looks like upstream already dealt with cppcheck, that's a start | 17:53 |
| jtaylor | so forwarded bugs which have bug trackers wher eI have an account | 18:07 |
| tumbleweed | jtaylor: yeah, that's generally what I do (or mail the upstream). I expect the Debian package maintainer to have a relationship with upstream, so I'm happy to just report there if the upstream isn't easy to contact | 18:08 |
| wibblymat | I've created a bug in sourceforge for xfe. Does it need linking to LP in some way, or is it fire and forget? | 18:08 |
| tumbleweed | wibblymat: I'd mention it in th LP bug. One can't link unless the project is registered in LP | 18:08 |
| tumbleweed | wibblymat: if you do a quilt patch, mentioning where you've forwarded it in the patch header is a good idea: http://dep.debian.net/deps/dep3/ | 18:10 |
| jtaylor | and mark it forwarded in the debian bug | 18:12 |
| jtaylor | so last one for now: mooproxy | 18:16 |
| jtaylor | done | 18:22 |
| tumbleweed | nigelb: seeing that scratchbox2 has no patch system, I think a far more descriptive changelog entry is required (and wouldn't have been a bad idea anyway) | 18:22 |
| tumbleweed | nigelb: also, please link the LP bug to the Debian one :) | 18:23 |
| nigelb | tumbleweed: waiting for BTS to give me the bug number | 18:23 |
| nigelb | tumbleweed: suggestions on what needs to go in? | 18:24 |
| tumbleweed | ah, it's debian bug 632103 | 18:24 |
| ubottu | Debian bug 632103 in scratchbox2 "scratchbox2: Fails to build from source" [Important,Open] http://bugs.debian.org/632103 | 18:24 |
| nigelb | why didn't BTS mail me :\ | 18:25 |
| tumbleweed | probably just e-mail being slow... | 18:25 |
| nigelb | how does this sound? "Fixed FTBFS. Changed the ordering of libs in the Makefile in utils folder." | 18:26 |
| tumbleweed | and it should mention the LP bug | 18:26 |
| tumbleweed | this isn't ordering of libs, so much as libs in CFLAGS | 18:27 |
| james_w | changing the order of libs is what you did, but anyone can see that from the diff, the changelog should explain *why* you did that | 18:28 |
| james_w | why did that order cause the build failure? why doesn't the new one? | 18:28 |
| nigelb | ahh | 18:29 |
| micahg | tumbleweed: unfortunately, I can't help much on the FTBFS jam today, but I can help with syncing from Debian for the fixed packages if you like | 18:31 |
| tumbleweed | micahg: np. I haven't actually fixed any myself today, just been reviewing | 18:31 |
| nigelb | micahg: if you're aroudn toe help minions like me fixing, that helps too ;) | 18:32 |
| micahg | nigelb: nope, sorry, busy with other updates ATM | 18:33 |
| nigelb | np :) | 18:33 |
| nigelb | tumbleweed: I think we can announce that the jam is going to continue for a week. | 18:37 |
| tumbleweed | yeah, I kind of said that in the e-mail | 18:39 |
| nigelb | I'll post a status update onto planet and announce that | 18:40 |
| nigelb | tumbleweed: Updated the patch | 18:40 |
| tumbleweed | nigelb: that's better, although I'd still say "FTBFS with ld --as-needed" instead of toolchain change | 18:41 |
| nigelb | tumbleweed: hang on, updating :) | 18:43 |
| nigelb | tumbleweed: done :) | 18:45 |
| nigelb | ok, so now I'm dealing with opencity which uses autotools | 18:48 |
| Rhonda | nigelb: Don't become insulting :P | 18:48 |
| tumbleweed | autotools packages usually have the mistake in configure.ac (or sometimes one of the Makefile.am s) | 18:48 |
| Rhonda | LetoThe2nd: I meant that I'm going offline. :) | 18:49 |
| jtaylor | usually LDFLAGS where it should be LIBS | 18:49 |
| jtaylor | but its a little problematic to patch when they don't use autoreconf :/ | 18:49 |
| tumbleweed | dh_autoreconf to the rescue :) | 18:49 |
| tumbleweed | nigelb: uploaded, thanks | 18:51 |
| nigelb | \oo/ | 18:51 |
| nigelb | erm | 18:51 |
| nigelb | \o/ | 18:51 |
| nigelb | Rhonda: :D | 18:51 |
| * tumbleweed looks at a concice but ugly jtaylor patch, and decides its time to go home | 18:52 | |
| nigelb | This is going to be tricky. | 18:52 |
| nigelb | I don't see anything problematic | 18:52 |
| jtaylor | ugly in which respect? | 18:52 |
| jtaylor | don't we want to wait for debian maintainer reaction? | 18:53 |
| tumbleweed | jtaylor: ugly in that it's a workaround rather than an upstreamable patch | 18:53 |
| jtaylor | which one? | 18:54 |
| tumbleweed | sslsniff | 18:54 |
| jtaylor | yeah that one is ugly | 18:54 |
| jtaylor | package does not use autoreconf and I think it is up to the maintainer to do that | 18:54 |
| tumbleweed | heh | 18:54 |
| nigelb | is the maintainer Rhonda ? ;) | 18:54 |
| tumbleweed | Also, isn't there a difference between -pthread and -lpthread? | 18:55 |
| tumbleweed | (theoretically) | 18:55 |
| jtaylor | yes, I just pasted what is in the Makefile.am | 18:56 |
| nigelb | tumbleweed: do you know of an autotools-related package fixed so I could take pointers on how to/where to look | 18:56 |
| jtaylor | look for LDFLAGS first | 18:56 |
| tumbleweed | nigelb: today's xfe patch | 18:57 |
| nigelb | jtaylor: not there in any of the makefile.am | 18:57 |
| tumbleweed | probably in configure.ac then | 18:58 |
| * tumbleweed will look when I get home, brb | 18:58 | |
| nigelb | oh | 18:58 |
| nigelb | LDFLAGS="$LDFLAGS $SDL_LIBS" | 18:58 |
| Rhonda | nigelb: pah! I'll tell manoj! | 18:58 |
| jtaylor | thats wrong | 18:58 |
| nigelb | Rhonda: :) | 18:58 |
| nigelb | jtaylor: ah, configure is full of that :| | 18:59 |
| nigelb | move all of that to LIBS? | 18:59 |
| jtaylor | just a note on my patches, they are intentional minimal, just showing whats wrong | 18:59 |
| jtaylor | My intention is that the debian maintainer will adapt it to its needs, the bug usually contains information on what I consider a better fix | 18:59 |
| jtaylor | nigelb: if there are only libs then yes | 19:00 |
| nigelb | jtaylor: mostly. ran into one fugly situation now. | 19:01 |
| jtaylor | I'm not sure if it has consequences when you put a linker flag in the LIBS variable | 19:01 |
| nigelb | LDFLAGS="-L$with_gl_prefix/lib $enable_gprof $LDFLAGS" | 19:02 |
| nigelb | is this real linker flag? | 19:02 |
| jtaylor | -L should be fine in libs | 19:02 |
| nigelb | and there is this | 19:03 |
| nigelb | http://paste.ubuntu.com/635222/ | 19:03 |
| jtaylor | should also be fine | 19:03 |
| jtaylor | good pkg-config --libs should not have any positional ldflags in it | 19:03 |
| jtaylor | expands to something like this: -L/usr/lib/x86_64-linux-gnu -lpng12 | 19:04 |
| nigelb | ah | 19:05 |
| tumbleweed | jtaylor: the problem with that approach is that when the package doesn't FTBFS on Debian, the maintainer often doesn't care. We'd like the upstream to include the right fix as soon as possible, so it makes sense to prepare it | 19:15 |
| jtaylor | lets see if a maintainer reacts | 19:41 |
| jtaylor | if its to slow I might provide proper patches | 19:41 |
| aonyx | tumbleweed: I am officially poking you. | 19:44 |
| aonyx | it is David btw | 19:44 |
| tumbleweed | aonyx: hi | 19:44 |
| tumbleweed | yay for a south african presence :) | 19:45 |
| aonyx | :) | 19:45 |
| aonyx | I was wondering if you could point out a particularly easy package for me | 19:46 |
| tumbleweed | sure, let me find one. In the meantime, here's some background: https://lists.ubuntu.com/archives/ubuntu-devel/2011-June/033619.html | 19:46 |
| tumbleweed | things have just started quieting down, we had a relativly busy last few hours | 19:47 |
| tumbleweed | aonyx: ok, xbill is quite easy to fix but it uses autoconf, which makes things a little more complicated. Let me see if I can find something easier | 20:02 |
| tumbleweed | aonyx: ah, even easier, tabix, bug 803202 | 20:11 |
| ubottu | Launchpad bug 803202 in tabix (Ubuntu) "tabix version 0.2.5-1 failed to build on i386" [Medium,New] https://launchpad.net/bugs/803202 | 20:11 |
| aonyx | awesome, thanks tumbleweed. | 20:11 |
| === yofel_ is now known as yofel | ||
| === paul_ is now known as Elbrus | ||
| nigelb | opencity done, forwarded to debian and upstream. | 22:25 |
| nigelb | bug 771091, if anyone wants to sponsor (its on the queue anyway) | 22:25 |
| ubottu | Launchpad bug 771091 in opencity (Ubuntu Oneiric) "opencity version 0.0.6.2stable-1 failed to build on amd64 with GCC-4.6/oneiric" [High,In progress] https://launchpad.net/bugs/771091 | 22:25 |
| Laney | nigelb: you shouldn't use RC severity unless the FTBFS actually happens in Debian too | 22:26 |
| nigelb | Laney: argh. Wrong key. I was sending them with important. | 22:27 |
| jtaylor | as-needed failures are more minor or even wishlist as debian does not use that default | 22:27 |
| nigelb | I thought debian was also transitioning with us | 22:27 |
| nigelb | Laney: changing it now | 22:28 |
| jtaylor | no, only --no-copy-dt-needed | 22:28 |
| nigelb | ah | 22:28 |
| jtaylor | although its probably possible that that will change | 22:28 |
| jtaylor | especially when everything builds in ubuntu | 22:28 |
| tumbleweed | but generally speaking, if you haven't verified that the package won't build on debian, don't file an RC bug :) | 22:30 |
| nigelb | tumbleweed: I used submittodebian, so I think I pressed the wrong number. | 22:31 |
| tumbleweed | fair enough :) | 22:31 |
| Laney | you can downgrade it | 22:31 |
| nigelb | Anyway, just mailed control to downgrade it | 22:31 |
| Laney | bts severity xxxxxxx minor | 22:31 |
| Laney | nice one | 22:31 |
| nigelb | Laney: thanks for pointing it out. | 22:32 |
| nigelb | Its 3 am, fine details are escaping me :P | 22:32 |
| Laney | someone moaned in #debian-devel | 22:33 |
| jtaylor | a python guy, but with good reason, python apps break easily thanks to as-needed | 22:34 |
| jtaylor | there are several still broken in natty due to the phase with as-needed | 22:34 |
| jtaylor | they don't ftbs but instead break at runtime thanks to dlopen ._. | 22:35 |
| === med_out is now known as medberry | ||
| nigelb | phew, severity downgraded to normal. | 22:38 |
| * tumbleweed tends to use minor for these | 22:42 | |
| nigelb | doh | 22:47 |
| nigelb | next time! | 22:47 |
| wibblymat | Oops, I put the wrong bug number in the changelog for xfe so it hasn't auto closed :s | 23:05 |
| * tumbleweed isn't good about checking such things | 23:06 | |
| jtaylor | the .pc file is in the debian patch | 23:10 |
| jtaylor | you should avoid that in future | 23:10 |
| wibblymat | Ah, ok | 23:12 |
| tumbleweed | that sounds like bzr + 3.0 (quilt) which is messy... | 23:12 |
| wibblymat | A lot of things I've seen in my first week of digging in to Ubuntu has been messy or complicated :) | 23:15 |
| nigelb | I'm staying away from bzr when quilt is involved | 23:18 |
| * tumbleweed wonders about tweaking submittodebian to unapply patches and filter .pc | 23:22 | |
| Laney | does it present the diff to you for review? | 23:25 |
| tumbleweed | the problem I have with it is that the patch it prepares never gets to my mutt, so I always attach a patch I prepared myself | 23:26 |
| tumbleweed | err yes it does let you edit the diff, IIRC | 23:27 |
| hakermania | Hello, till when can we post in REVU? | 23:41 |
| hakermania | And please can anybody explain me how this was advocated? it looks so crazy with so many errors: http://revu.ubuntuwire.com/p/plymouth-manager | 23:44 |
| micahg | hakermania: it's not in the archive | 23:50 |
| micahg | hakermania: and until feature freeze | 23:50 |
| tumbleweed | oh, btw jtaylor: BTS doesn't allow usertags for two users to be set by pseudo-headers, so your origin-ubuntu usertag wouldn't have been applied | 23:54 |
| jtaylor | ._. | 23:55 |
| jtaylor | saw it in a different bug and copied it | 23:55 |
| jtaylor | will fix later | 23:55 |
| tumbleweed | I assume you want to wait for debian maintainter responses, rather than having anyone sponsor them in ubuntu yet? | 23:55 |
| jtaylor | yes | 23:56 |
| jtaylor | unless there is reason for haste? | 23:56 |
| tumbleweed | not particularly for most of those packages | 23:57 |
| jtaylor | how do I clone a github pull request? :/ | 23:57 |
| tumbleweed | clone the source repository? | 23:57 |
| jtaylor | I can't find a link on the page | 23:57 |
| tumbleweed | IIRC github pull requests have links to the source and destination | 23:58 |
| * jtaylor to stupid to find it | 23:59 | |
| jtaylor | just edited to url to get to the source | 23:59 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!