| voxadam | I see it mentioned that Upstart is going to support timed and scheduled jobs replacing cron but I don't see any Blueprints/specs for it in Launchpad, on the wiki or in the bugtracker, and the last mention of it on uostart-devel was two years ago. Has this functionality been abandon? | 02:24 |
|---|---|---|
| ion | There have been more pressing things to do. | 02:38 |
| voxadam | I can appreciate that. Are timed and scheduled events still planned? | 03:13 |
| ion | Yes. Probably not yet for the next major release (0.10), though. | 03:22 |
| ion | Unless Keybuk has other ideas | 03:22 |
| Keybuk | ion: main piece of work there is the parser | 09:48 |
| keesj | Hi | 11:21 |
| Keybuk | hey, how goes it? | 11:44 |
| keesj | I am doing alright. Thanks for looking at the ini.conf patch | 11:50 |
| Keybuk | keesj: that's ok, sorry I didn't recognise your e-mail address | 11:50 |
| Keybuk | it came through to me as the commit message with your bzr address | 11:51 |
| * Keybuk has just about finally unburied himself from Launchpad e-mail | 11:51 | |
| Keybuk | two years worth of bugs, commits, merge requests, build logs, etc. | 11:51 |
| Keybuk | tip: if you ever become a manager, don't try and hang on to development ;-) | 11:52 |
| keesj | last time I started put time in upstart I hit the dbus version problem and I didn't want to upgrade the system yet | 11:52 |
| Keybuk | keesj: the 1.2.15 dep? | 11:52 |
| Keybuk | 1.2.16 is out now, which fixes a whole bunch of bugs | 11:52 |
| keesj | the bug you found in dbus and sent a patch to them | 11:52 |
| keesj | yes, but there other components using dbus so I can' easly do it | 11:56 |
| Keybuk | oh right, it should be entirely backwards compatible with other things in 1.2.x though | 11:57 |
| Keybuk | you could always build 1.2.16 statically (but not install) | 11:58 |
| Keybuk | then point upstart's build at it, so it'll statically link to it | 11:58 |
| Keybuk | I think it's something like | 12:00 |
| Keybuk | dbus$ ./configure --disable-shared | 12:00 |
| Keybuk | upstart$ ./configure DBUS_CFLAGS="-I../dbus" DBUS_LIBS="../dbus/libdbus.a -lpthread -lrt" | 12:01 |
| sadmac | Keybuk: all the testcases in upstart list the .o files of the code they're testing as libraries and link statically, rather than just listing the .c files as additional sources. What does that accomplish? | 14:24 |
| Keybuk | sadmac: doesn't make any difference to the output? | 14:24 |
| Keybuk | it does mean you're testing the built .o files, of course | 14:25 |
| Keybuk | though in practice, they'd get re-used anyway | 14:25 |
| sadmac | Keybuk: I'd have to look at the output to make sure. In theory it should be functionally identical | 14:26 |
| sadmac | I don't see what /interesting/ differences you'd get out of that, if any. | 14:26 |
| Keybuk | just style mostly | 14:26 |
| Keybuk | _SOURCES for a test are the test | 14:26 |
| Keybuk | under nih it links the library | 14:27 |
| sadmac | Keybuk: might be nice to have libnih provide some M4 macros around that. Its a bit confusing at times. | 14:29 |
| sadmac | UD? | 14:35 |
| Keybuk | Urban Dictionary | 14:42 |
| sadmac | oh yes | 14:43 |
| sadmac | I have a notoriously monotone friend, and we sometimes have him read UD definitions to us for amusement | 14:44 |
| mbiebl | Keybuk: would you mind moving dist_man_MANS = man/nih-dbus-tool.1 inside the if INSTALL_HI endif? | 17:00 |
| Keybuk | it's probably a bug that it isn't ;) | 17:00 |
| Keybuk | patch me! | 17:00 |
| mbiebl | lp oder directly via email? | 17:02 |
| * sadmac fixes a small leak in Keybuk | 17:02 | |
| Keybuk | mbiebl: your preference | 17:03 |
| Keybuk | LP means I'm less likely to forget about it | 17:03 |
| mbiebl | ok then | 17:03 |
| Keybuk | should I vanish before you send it | 17:03 |
| Keybuk | mbiebl: how's getting Upstart 0.6 into Debian going? | 17:09 |
| * Keybuk noticed sid was still on 0.4 | 17:09 | |
| Keybuk | err 0.3 | 17:09 |
| mbiebl | upload today or tomorrow | 17:10 |
| Keybuk | sweet | 17:18 |
| === robbiew is now known as robbiew-afk | ||
| Keybuk | mbiebl: I don't think the extra bit of your patch is necessary | 17:40 |
| Keybuk | automake is smart enough to see dist_* inside an if | 17:40 |
| Keybuk | so patch2 will dtrt anyway | 17:40 |
| mbiebl | Keybuk: sure? | 17:42 |
| Keybuk | yeah | 17:42 |
| Keybuk | I rely on that elsewhere | 17:42 |
| Keybuk | m4/Makefile.am for example | 17:42 |
| mbiebl | tried it, and it wasn't included | 17:42 |
| mbiebl | oh, m4 files have a special meaning | 17:42 |
| mbiebl | you don't need to add them to dist files at all | 17:43 |
| mbiebl | they will end up in the dist tarball in any case | 17:43 |
| mbiebl | the man page won't | 17:43 |
| Keybuk | that seems like a bug | 17:44 |
| Keybuk | dist shouldn't change depending on configure options | 17:44 |
| mbiebl | well, that's how it is unfortunately | 17:44 |
| Keybuk | DIST_COMMON = $(dist_man_MANS) $(srcdir)/Makefile.am \ | 17:45 |
| Keybuk | $(srcdir)/Makefile.in | 17:45 |
| Keybuk | it's included in DIST here | 17:45 |
| Keybuk | oh, hah | 17:46 |
| Keybuk | but dist_man_MANS never gets defined | 17:46 |
| Keybuk | silly Automake | 17:46 |
| Keybuk | is definitely automake bug ;) | 17:46 |
| mbiebl | yeah | 17:46 |
| Keybuk | it's trying to do the right thing, but failing | 17:46 |
| mbiebl | that's why you need the workaround with the variable defined outside of the NIH_INSTALL scope | 17:47 |
| mbiebl | Keybuk: dist_man_MANS = $(manpages) and EXTRA_DIST = $(manpages) is doubled | 19:06 |
| mbiebl | just a minor nit though | 19:06 |
| mbiebl | Keybuk: still around? | 21:13 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!