2012-02-29 00:27:14 elenst wlad: hi.. any idea who it might belong to? https://bugs.launchpad.net/maria/+bug/942266
2012-02-29 00:27:37 wlad elenst, me:)
2012-02-29 00:28:14 elenst wlad: ok, so you are aware of it? it's just not tagged anyhow, i was afraid it would stay unnoticed
2012-02-29 00:28:29 elenst not tagged as not assigned or milestoned
2012-02-29 00:28:35 elenst even although it's not a word
2012-02-29 00:28:54 wlad elenst, I quickly tried to reproduce yesterday, but could not. Yes, I'm aware.
2012-02-29 00:29:08 elenst wlad: ok, cool
2012-02-29 00:29:45 wlad this should probably go into 5.5, and will be a small fix. the hard part here is to install ubuntu:)
2012-02-29 00:30:43 wlad because on mint made of older ubuntu, and on fedora16 it works ok.
2012-02-29 00:35:31 mkanat joined chan
2012-02-29 01:15:45 hartmut joined chan
2012-02-29 01:30:39 walterheck joined chan
2012-02-29 02:07:56 wlad CareBear\, there is documentation about the feature. It says what is sent
2012-02-29 02:08:04 wlad and probably you can control it
2012-02-29 02:08:17 wlad serg knows more, it is his baby
2012-02-29 02:12:00 wlad CareBear\, send feedback to defend MariaDB credibility on Linux;) with only your contribution, number of Linux users grows by 11%!
2012-02-29 02:12:15 wlad http://mariadb.org/feedback_plugin/stats/os/ :)
2012-02-29 02:17:38 erkules joined chan
2012-02-29 02:24:08 wlad mysqld uptime, which is often the server uptime
2012-02-29 02:24:38 wlad not many people have 166 days
2012-02-29 02:24:46 wlad I think you're the first one.
2012-02-29 02:26:11 wlad CareBear\, you should be able
2012-02-29 02:26:21 wlad install plugin soname...
2012-02-29 02:26:24 wlad http://kb.askmonty.org/en/user-feedback-plugin
2012-02-29 02:26:43 wlad CareBear\, also read "sending data" chapter here
2012-02-29 02:26:51 wlad it is about how to send it manually
2012-02-29 02:29:34 wlad CareBear\, well perhaps it was not there in your version
2012-02-29 02:29:49 wlad but you can build it manually with another  version
2012-02-29 02:29:59 wlad and then install
2012-02-29 02:30:11 wlad should work imo
2012-02-29 02:33:08 wlad CareBear\, yes, perhaps the oldest version where it appeared is about 71 days old:)
2012-02-29 02:34:21 wlad CareBear\, what you can do, is download newer 5.1, extract feedback.so from it, copy to lib/plugin, run "install plugin ..."
2012-02-29 02:35:15 wlad I think it is in newer 5.1
2012-02-29 02:36:42 wlad CareBear\, is this a production database? Wondering because the feature was though to monitor databases in production, but as you can see from CPU count chart or from uptime chart, it is mostly used on development laptops
2012-02-29 02:37:43 wlad great, at least some useful bits;)
2012-02-29 02:47:51 wlad CareBear\, really?
2012-02-29 02:48:11 wlad find . -name "*.so" :)
2012-02-29 02:48:44 wlad why the hell it is static
2012-02-29 02:49:08 wlad did you use --with-feedback?
2012-02-29 02:49:42 wlad the logic of mysql plugins is to make plugins static if --with- is given
2012-02-29 02:49:52 wlad or to make shared if nothing is given
2012-02-29 02:50:02 wlad or not to make, if --without is given
2012-02-29 02:50:13 wlad it is a complicated logic
2012-02-29 02:50:21 wlad sorry
2012-02-29 02:50:54 wlad yes
2012-02-29 02:51:08 wlad it is the same
2012-02-29 02:51:29 wlad alias with --with-plugin-PLUGIN repeated many times
2012-02-29 02:52:00 wlad just do not use --with stuff;)
2012-02-29 02:59:13 wlad CareBear\, we have Kasachstan:)
2012-02-29 03:00:24 wlad this is what I'm mostly proud of, as I was there as kid, teen, and twen several times.
2012-02-29 03:00:37 wlad nice places, great nature
2012-02-29 03:00:56 wlad but was not so much IT back in the days
2012-02-29 03:01:59 wlad CareBear\, sorry to head that cheat did not work. It is bad to have plugin dependencies in server
2012-02-29 03:03:04 wlad CareBear\, no idea, something from sql/* directory
2012-02-29 03:03:37 wlad sounds about right
2012-02-29 03:03:56 wlad time(NULL) ;)?
2012-02-29 03:05:13 wlad convert string to time, with warning
2012-02-29 03:05:20 wlad my guess
2012-02-29 03:05:51 wlad wlad only looked at the function name not at the code
2012-02-29 03:06:47 wlad not so much
2012-02-29 03:09:39 wlad CareBear\, it is only due to inclusion of some header?
2012-02-29 03:10:38 wlad ehh, why do we include all that stuff
2012-02-29 03:13:07 wlad CareBear\, nm <path/to/feedback.so> ?
2012-02-29 03:13:19 wlad does it show something interesting?
2012-02-29 03:39:04 wlad CareBear\, time to sleep:)  it looks like it is not possible to fix it in a simple way. serg may provide better instructions or even a fix :)
2012-02-29 03:40:40 wlad CareBear\, night!
2012-02-29 04:46:48 nils_ joined chan
2012-02-29 05:39:15 jespada joined chan
2012-02-29 05:46:09 montywi CareBear\: is that from 5.3 ?
2012-02-29 05:50:40 spetrunia joined chan
2012-02-29 06:01:19 bar_ joined chan
2012-02-29 06:06:27 shinguz joined chan
2012-02-29 06:16:22 stewart joined chan
2012-02-29 06:47:37 SanjaByelkin I never thought that small box-less computer for geeks trigger such interest. It looks like I missed my chance to buy one...
2012-02-29 06:47:56 SanjaByelkin (I am about raspberry pi)
2012-02-29 06:49:13 SanjaByelkin CareBear\: I hope it will be more, but why everybody need it so that they went down 2 web sites
2012-02-29 06:50:40 SanjaByelkin OK, in any case I was not sure if it possible to order from my country, so lets wait...
2012-02-29 07:01:41 SanjaByelkin All other I saw cost at least 2 times more - too much for big boys games :) (maybe I was looking in wrong places)
2012-02-29 07:11:10 SanjaByelkin CareBear\: se has very expensive beer compared to ua :)
2012-02-29 07:13:21 SanjaByelkin in any case I could wait to get something "main stream" because it always has better support. and as I told it is toy so I do not know yet how to apply
2012-02-29 07:13:36 SanjaByelkin it
2012-02-29 07:18:29 SanjaByelkin I do not like beer in general, but beer made in kulture Bierbrauerei was what I liked...
2012-02-29 07:27:25 SeppoJ joined chan
2012-02-29 07:40:53 hartmut joined chan
2012-02-29 08:05:22 Pierre joined chan
2012-02-29 08:19:08 ipalaus joined chan
2012-02-29 08:27:19 isnnnn joined chan
2012-02-29 08:38:23 kaz_lille joined chan
2012-02-29 08:50:08 Roguecatz joined chan
2012-02-29 08:58:54 ipalaus joined chan
2012-02-29 09:01:56 shinguz joined chan
2012-02-29 09:06:27 knielsen CareBear\: Do you have a recommendation for a good ARM cortex m3 board?
2012-02-29 09:22:52 XL joined chan
2012-02-29 09:34:55 knielsen CareBear\: not sure :) Something simple I guess - I'm not sure what is available
2012-02-29 09:35:07 knielsen yes, I noticed the workshop, maybe I should just go to that
2012-02-29 09:39:43 AL13N small question
2012-02-29 09:39:47 AL13N what is a GA release?
2012-02-29 09:39:58 AL13N is that sort of like between an rc release and a final?
2012-02-29 09:40:10 knielsen GA is final
2012-02-29 09:40:14 AL13N ah okj
2012-02-29 09:40:22 AL13N 5.3 nearly ready then :-)
2012-02-29 09:40:34 AL13N hopefully 5.5 will be soon-ish
2012-02-29 09:40:45 AL13N when you guys release, do you update version?
2012-02-29 09:41:04 knielsen sure - what do you mean?
2012-02-29 09:41:07 AL13N like 5.5.20-alpha => 5.5.21-beta => 5.5.22-rc => 5.5.23-ga ?
2012-02-29 09:41:13 knielsen yes
2012-02-29 09:41:28 AL13N right
2012-02-29 09:41:36 AL13N for me it sounds a bit weird, but ok
2012-02-29 09:41:54 AL13N usually we have 5.5.20-alpha 5.5.20-beta, etc...
2012-02-29 09:41:55 knielsen might be 5.5.22 => 5.5.22a => ... if we release more often than oracle
2012-02-29 09:41:58 AL13N since it's the same version
2012-02-29 09:42:03 AL13N oic
2012-02-29 09:42:25 AL13N i'm asking because we have a version freeze coming up
2012-02-29 09:42:58 knielsen basically, the major version is "5.5". Then each official release bumps the minor version number: 5.5.0, 5.5.1, ...
2012-02-29 09:43:02 AL13N the version freeze restricts us to having the version part frozen (5.5.20) and release bumps
2012-02-29 09:43:29 AL13N there's an exception if it's bugfix releases (no new features)
2012-02-29 09:43:31 knielsen they are all "the same version" in the sense that after GA only bugs are fixed and nothing that breaks compatibility
2012-02-29 09:43:52 AL13N so, between alpha and ga, are there features added,
2012-02-29 09:43:54 AL13N ?
2012-02-29 09:44:07 knielsen between alpha and beta. beta means no new features added
2012-02-29 09:44:14 AL13N hmm
2012-02-29 09:44:24 AL13N any idea when beta is released?
2012-02-29 09:44:32 knielsen 2-3 weeks is the plan
2012-02-29 09:44:38 AL13N hmm
2012-02-29 09:44:44 AL13N AL13N has a quick count
2012-02-29 09:44:46 AL13N crap
2012-02-29 09:44:49 AL13N not gonna make it
2012-02-29 09:45:03 AL13N i'll have to use the 2nd exception about supportability
2012-02-29 09:45:39 AL13N knielsen: is it also a plan to have 2-3 weeks between beta and rc and rc and ga?
2012-02-29 09:46:42 knielsen well - we hope for GA in 5-6 weeks from now, and feature freeze in 2, so kind of yes, though the are no specific plans for rc
2012-02-29 09:46:57 AL13N 5-6 weeks
2012-02-29 09:46:59 knielsen when is your version freeze?
2012-02-29 09:47:06 AL13N generally such things leak a bit
2012-02-29 09:47:16 AL13N so that's more like 2 months before final
2012-02-29 09:47:24 AL13N version freeze is 7march :-(
2012-02-29 09:47:38 AL13N but release freeze is around april i think
2012-02-29 09:47:53 AL13N AL13N goes to doublecheck
2012-02-29 09:48:11 knielsen well, we might still make it - we will see
2012-02-29 09:49:01 knielsen but with respect to "version freeze" - this may work bad with mysql/mariadb versioning. There will never be two releases with the same X.Y.Z version
2012-02-29 09:49:19 knielsen X.Y is the version, .Z is the bugfix release.
2012-02-29 09:49:51 knielsen So if you freeze X.Y.Z you will never get any bugfixes, except what you backport yourself (this is how Debian/Ubuntu does it, or at least used to do)
2012-02-29 09:50:13 knielsen this may or may be what you want - just letting you know how it works ...
2012-02-29 09:50:58 AL13N we backport specific bugfixes/security updates, generally, but there's an except that if no new features are added, AND it's supported upstream we can do it
2012-02-29 09:51:22 knielsen ok, then you can do either, I suppose (after beta)
2012-02-29 09:51:25 AL13N so i'll likely just use the bugfix releases
2012-02-29 09:51:28 AL13N but
2012-02-29 09:51:38 AL13N generally we like to ship our distro with final version
2012-02-29 09:51:47 AL13N but i see that release freeze is 7th of april
2012-02-29 09:51:54 AL13N i don't think the GA will be in time for that
2012-02-29 09:52:04 knielsen well, "final" version concept does not really exist in mysql/mariadb world
2012-02-29 09:52:04 AL13N it's always possible, of course, but we'll have to see
2012-02-29 09:52:17 AL13N knielsen: how do you mean?
2012-02-29 09:52:40 AL13N knielsen: which versions do you support? only the GA, i presume
2012-02-29 09:53:10 knielsen The supported version is the latest version in each major series
2012-02-29 09:53:39 AL13N so, if i lets say ship mariadb-5.5.22-rc, that means that mageia 2 goes with that, and if people don't update, they are going to be using the nonGA version
2012-02-29 09:54:02 knielsen Ie. now we release 5.3.5 GA - we will support this, and the support consists in releasing 5.3.6, 5.3.7, 5.3.8, ... bug fix releases
2012-02-29 09:54:16 AL13N right
2012-02-29 09:54:18 knielsen so the supported version is "5.3"
2012-02-29 09:54:25 AL13N right
2012-02-29 09:54:26 AL13N but
2012-02-29 09:54:33 knielsen to say we support "5.3.5" is kind of meaningless
2012-02-29 09:54:34 AL13N 5.3.4 is not really supported
2012-02-29 09:54:54 AL13N i understand that, but pre-GA is never supported, right
2012-02-29 09:54:55 AL13N ?
2012-02-29 09:55:05 AL13N or am i misunderstanding?
2012-02-29 09:55:29 knielsen sure it is supported. We support it with bugfix releases that people can install.
2012-02-29 09:55:43 knielsen how else would we support it?
2012-02-29 09:55:52 AL13N ok, i see your point
2012-02-29 09:56:06 AL13N but still, i'm quite hesitant to have to ship mageia2 with something not GA
2012-02-29 09:56:10 AL13N knielsen: wouldn't you?
2012-02-29 09:58:53 knielsen There are strong forces to ensure 5.5 GA before percona live in Santa Clare, which means it would meet your April 7 deadline. But we will see
2012-02-29 10:05:05 AL13N heh
2012-02-29 10:05:08 AL13N we will see
2012-02-29 10:06:16 AL13N the whole point of us switching over was on the condition if it would be "ready"
2012-02-29 10:06:30 AL13N so far on our development we've been using some bzr revision from a while ago
2012-02-29 10:07:20 AL13N now we have 5.5.20-alpha
2012-02-29 10:07:43 AL13N the last month or so, i haven't heard about any bugs
2012-02-29 10:08:11 AL13N the only thing is that the KDE/akonadi maintainer having a crash, which was not reproducable or so it seems
2012-02-29 10:08:53 AL13N considering this means that almost all KDE users + testers who are using the development distro....
2012-02-29 10:09:44 AL13N that's likely 40-50% of our development userbase, and since we have about 100 packagers... that would mean 40-50 people estimated
2012-02-29 10:09:44 knielsen we actually do not have any pending development at all on 5.5, haven't had for quite some time. But we have some features finished a while ago sitting in other trees, we need to merge them before we can call it beta. It's just that 5.3 work and other things have caused 5.5 to suffer :-(
2012-02-29 10:10:05 AL13N yeah, i know, i've been looking at it
2012-02-29 10:10:18 AL13N but then the 5.3 parts will be merged in to 5.5 so
2012-02-29 10:10:30 knielsen yes, they are already merged
2012-02-29 10:10:31 AL13N 5.5 will benefit from it as well in the end
2012-02-29 10:10:35 knielsen yes
2012-02-29 10:11:07 AL13N still it means at least 40 people have been using mariadb 5.5 for a month
2012-02-29 10:11:12 AL13N :)
2012-02-29 10:11:42 AL13N and no outstanding bug reports
2012-02-29 10:11:44 AL13N so that's good
2012-02-29 10:11:45 AL13N :)
2012-02-29 10:12:05 AL13N also noone complained when upgrading to 5.5.20
2012-02-29 10:12:05 knielsen yes! :)
2012-02-29 10:12:39 AL13N still, that one guy with the huge akonadi database, i'd like to get him to try it too, but he has hardware issues
2012-02-29 10:21:12 knielsen spetrunia: I looked at rpl_start_stop_slave failure. it seems open_tables() or lock_tables() is returning error without setting proper error code in some case. But I think I need to discuss with monty to proceed.
2012-02-29 10:21:37 knielsen In any case, it's not fatal - it only affects debug builds, afaik, release build will set a valid error code in this case anyway
2012-02-29 10:23:39 spetrunia ok, thanks..  have you caught that "some case" ?  I've tried to narrow it down but was not successfull...
2012-02-29 10:24:14 knielsen spetrunia: indeed, I also have not - that's why I needed monty, I am not very familiar with table open/locking code
2012-02-29 10:24:42 knielsen also I remember monty did some fixes already around slave error codes during STOP SLAVE
2012-02-29 10:26:26 knielsen I think STOP SLAVE sends some non-forced "kill" signal to wake up any sleeping slave - but in this case, slave must not stop immediately, as it is in the middle of an update that cannot roll back - so probably there is some case where this situation is incorrectly handled
2012-02-29 10:27:32 knielsen montywi: ^^^ let me know when you are available to discuss a problem with STOP SLAVE and table locking / error reporting
2012-02-29 10:33:49 ipalaus joined chan
2012-02-29 10:41:14 hakank joined chan
2012-02-29 10:42:14 rasmusj knielsen, I need to borrow your memory a bit
2012-02-29 10:42:25 rasmusj this task: https://mariadb.atlassian.net/browse/MDEV-22
2012-02-29 10:42:44 rasmusj do you recall the discussion about the subject
2012-02-29 10:42:49 rasmusj from Lisbon
2012-02-29 10:43:17 rasmusj I'm confused by my own comment saying "Take an existing patch"
2012-02-29 10:43:25 knielsen knielsen trying to remember
2012-02-29 10:44:09 knielsen rasmusj: sorry, it doesn't ring any bells
2012-02-29 10:47:55 rasmusj knielsen, ok, I'll check with mark if it's still relevant or not
2012-02-29 11:16:08 zul joined chan
2012-02-29 11:46:38 Roguecatz joined chan
2012-02-29 11:47:43 kostja_osipov joined chan
2012-02-29 11:51:02 npmapn joined chan
2012-02-29 12:04:33 pprkut joined chan
2012-02-29 12:43:56 maxmad joined chan
2012-02-29 12:51:21 Jackneill joined chan
2012-02-29 13:04:22 wlad joined chan
2012-02-29 13:06:42 jespada joined chan
2012-02-29 13:09:15 Roguecatz joined chan
2012-02-29 13:11:03 digitalfiz joined chan
2012-02-29 14:20:21 ipalaus joined chan
2012-02-29 14:30:18 DarkStar1 joined chan
2012-02-29 14:48:49 the_wench joined chan
2012-02-29 14:53:31 ipalaus joined chan
2012-02-29 15:06:15 DarkStar1_ joined chan
2012-02-29 15:30:15 dob_ joined chan
2012-02-29 15:32:49 knielsen huh? cmake .. -DCMAKE_BUILD_TYPE=Debug
2012-02-29 15:32:50 knielsen ccache: FATAL: Could not find compiler " gcc" in PATH
2012-02-29 15:33:33 knielsen strange, must be something wierd happened to my tree?
2012-02-29 15:35:00 wlad knielsen, is this really fatal or not?
2012-02-29 15:35:06 knielsen wierd - I had CMakeCache.txt in root, and was trying to build in subdir bld/ - and this happened. Deleting in the root solved it - but shouldn't cmake ignore other directories when I build in bld/
2012-02-29 15:35:25 knielsen wlad: maybe that was not the fatal error, but the build failed
2012-02-29 15:35:45 knielsen but removing CMakeCache.txt in the parent seems to have solved it
2012-02-29 15:36:00 wlad ok I see. building out-of-source after in-source build does not work (never did I think)
2012-02-29 15:36:10 knielsen yeah, I messed up :)
2012-02-29 15:36:22 wlad cmake has ambigious syntax here
2012-02-29 15:36:42 wlad cmake <path-to-source-dir-or-existing-builddir>
2012-02-29 15:36:50 knielsen ah!
2012-02-29 15:37:04 knielsen that explains it I think
2012-02-29 15:37:18 wlad yeah, took me some time to understand as well.
2012-02-29 15:41:40 wlad that ccache thing is something else. IIRC, this comes from the attempt to retrieve  gcc version - something jperkin has introduced,  I do not really recall the reason (perhaps INFO_BIN stuff), but error here was I think not fatal
2012-02-29 15:46:52 knielsen agree, there must have been some other error message that I overlooked
2012-02-29 16:39:26 DarkStar1 joined chan
2012-02-29 16:57:34 kaz_lille joined chan
2012-02-29 16:58:52 laurynas joined chan
2012-02-29 17:10:49 serg joined chan
2012-02-29 17:29:46 HarrisonF joined chan
2012-02-29 18:05:23 fluffypony joined chan
2012-02-29 18:20:40 atz6975 joined chan
2012-02-29 19:12:59 ipalaus joined chan
2012-02-29 19:20:25 spetrunia ../include/my_global.h:1579:0: error: "dlerror" redefined [-Werror]
2012-02-29 19:20:25 spetrunia ../include/my_global.h:1570:0: note: this is the location of the previous definition
2012-02-29 19:20:42 spetrunia this is what I get with compile-pentium-gprof...
2012-02-29 19:36:51 wlad spetrunia, this looks like a static build, no wonder it does not work
2012-02-29 19:38:38 spetrunia and if I get around that error and compile, the binary doesn't support innodb...
2012-02-29 19:38:53 spetrunia it seem like compile-pentium-gprof is unmaintained...
2012-02-29 19:43:41 dbart rasmusj: ping
2012-02-29 20:10:38 ipalaus joined chan
2012-02-29 20:55:38 hingo joined chan
2012-02-29 21:00:43 marialog joined chan
2012-02-29 21:06:25 stewart joined chan
2012-02-29 21:08:28 Dev0n joined chan
2012-02-29 21:08:49 Dev0n hey, "# InnoDB is enabled by default with a 10MB datafile ", I assume this is set to autoextend with no max size limit right ?
2012-02-29 21:09:27 Dev0n yup, cool
2012-02-29 21:10:41 stewart Dev0n, correct. you may want to have a larger file though, even if using file per table.
2012-02-29 21:11:04 Dev0n but if it autoextends, should that really matter ?
2012-02-29 21:16:27 knielsen elenst: mariadb has an empty libmysqlclient16 package that depends on libmariaddclient16. This is necessary so we can satisfy some 3rd-party dependencies on libmysqlclient16. Probably because we have newer version number than the mysql-5.1 one it selects the mariadb one by default
2012-02-29 21:17:09 knielsen elenst: you can avoid this by explicit giving version: apt-get install mysql-server-5.1 libmysqlclient16=5.1.41
2012-02-29 21:17:48 knielsen (replace 5.1.41 with actual ubuntu version number, seen with apt-cache show libmysqlclient16
2012-02-29 21:18:07 elenst knielsen: okay, thanks, i guess it also explains why we attempt to upgrade libmysqlclient16 while installing 5.3 on the clean ubuntu 11.10...
2012-02-29 21:19:02 knielsen yes. libmysqlclient16 is part of the mariadb repos (though it is empty).
2012-02-29 21:19:30 elenst knielsen: yeah i'm having some troubles with this upgrade (i think, still looking)
2012-02-29 21:19:58 knielsen elenst: what upgrade?
2012-02-29 21:21:33 elenst knielsen: i try to install mariadb-server-5.3 on clean ubuntu 11.10, which does not have mysql-server or client, but does have libmysqlclient16 by default. When I install mariadb-server (as in apt-get install mariadb-server-5.3), it says it would *upgrade* libmysqlclient16. Then the installation runs and fails, it takes some dance to get it finished
2012-02-29 21:21:48 elenst knielsen: still looking... would it be you who i should assign bugs about our deb packages if i confirm problems?
2012-02-29 21:21:50 Dev0n is there plans to include sphinx 2.0.3 in 5.5 .?
2012-02-29 21:23:33 knielsen elenst: I am afraid I am the only one left to take the pain, yes :-( However, buildbot runs a test to install on clean ubuntu 11.10, and it does not fail
2012-02-29 21:23:34 terinjokes joined chan
2012-02-29 21:23:52 knielsen but it may be a server install that does not have libmysqlclient16?
2012-02-29 21:24:00 elenst knielsen: i know, i looked at the log. There it does not suggest to upgrade libmysqlclient16, either
2012-02-29 21:24:00 terinjokes noticed lately that MariaDB has been failing when using homebrew on OSX
2012-02-29 21:24:11 elenst knielsen: so i started a VM now, and no, it doesn't have the library
2012-02-29 21:24:30 knielsen however, we also test upgrading from a VM with mysql installed, where it does upgrade libmysqlclient16
2012-02-29 21:24:31 terinjokes configure seems to be bailing out because the version of readline is too new (we have readline version 6)
2012-02-29 21:25:31 knielsen terinjokes: it's a license issue. new libreadline is GPL v3, mysql/mariadb GPL v2. There is an option you can give if you are not distributing the result, so the license is not violated (or you dont care)
2012-02-29 21:25:39 elenst knielsen: upgrade from mysql is what i'm trying to do now, but can't get mysql-server-5.1 installed :( i think it's just not my day
2012-02-29 21:25:47 knielsen terinjokes: alternatively, you can build with bundled libreadline
2012-02-29 21:26:33 terinjokes knielsen: we have the option of libedit version 3, or using that switch
2012-02-29 21:26:42 terinjokes whatever that magical switch is
2012-02-29 21:26:55 knielsen elenst: https://buildbot.askmonty.org/buildbot/builders/kvm-deb-oneiric-amd64/builds/287/steps/test_1/logs/stdio
2012-02-29 21:27:10 knielsen The following packages will be upgraded: libmysqlclient16 mysql-common
2012-02-29 21:27:52 knielsen terinjokes: --disable-distribution
2012-02-29 21:29:26 elenst knielsen: thanks, i will be use it to compare my output with this. Maybe it just works better with 5.3.5, i was installing 5.3.4 from repository
2012-02-29 21:29:45 elenst knielsen: also, i was doing it on 32-bit
2012-02-29 21:30:17 knielsen terinjokes: or --with-readline=yes # use bundled readline
2012-02-29 21:30:29 knielsen elenst: no, it should be the same
2012-02-29 21:30:42 terinjokes knielsen: we want to avoid using the bundled readline
2012-02-29 21:30:55 knielsen elenst: but I've never tested installing mysql-server with maria repos enabled, quite possible it doesn't work
2012-02-29 21:31:16 terinjokes knielsen: giving it a shot with --disable-distribution
2012-02-29 21:31:18 knielsen elenst: just comment maria repos out temporarily and run apt-get update
2012-02-29 21:32:24 knielsen terinjokes: indeed, using bundled is best avoided ... the bailout is just to avoid people unknowingly getting into a gplv2/gplv3 conflict and later complaining
2012-02-29 21:33:48 elenst knielsen: yeah i will, next time. I already tried, but didn't succeed at installing mysql-server anyway, something else went wrong
2012-02-29 21:44:38 svaroqui joined chan
2012-02-29 21:45:28 terinjokes knielsen: i do have another error, although I'm not exactly sure where
2012-02-29 21:47:09 terinjokes see https://gist.github.com/1944688
2012-02-29 21:49:16 knielsen terinjokes: did you modify include/mysql.h (with a patch or something)
2012-02-29 21:49:18 knielsen ?
2012-02-29 21:49:40 knielsen terinjokes: that error is there to catch that a developer inadvertedly changes mysql.h in a way that would break ABI compatibility
2012-02-29 21:51:25 knielsen terinjokes: it looks strange, a bunch of #include's deleted from mysql.h ? maybe your build tree got messed up
2012-02-29 21:52:03 knielsen anyway, the error means - if one changes mysql.h in some way (or a file it includes), one must also change mysql.h.pp (and verify that ABI is preserved)
2012-02-29 21:53:24 _chris_ joined chan
2012-02-29 21:53:30 terinjokes knielsen: nope
2012-02-29 21:53:33 terinjokes o patches
2012-02-29 21:57:11 knielsen terinjokes: another possibility is that the preprosessor works differently on your machine
2012-02-29 21:57:28 terinjokes possible
2012-02-29 21:57:52 knielsen (the check works by running the preprocessor on mysql.h and comparing the output to a fixed files mysql.h.pp)
2012-02-29 21:58:11 terinjokes trying with another preprocessor
2012-02-29 21:58:52 knielsen terinjokes: if you can wait to tomorrow, maybe ask serg here in the channel or mail maria-developers@lists.launchpad.net, he knows more about this ...
2012-02-29 21:59:25 terinjokes i'll probably email the list after this run through the build script
2012-02-29 22:13:00 pento joined chan
2012-02-29 23:06:36 mkanat joined chan
2012-02-29 23:06:46 mkanat I dunnow if anybody pointed out http://code.google.com/p/vitess/ to you guys yet.
2012-02-29 23:15:00 walterheck joined chan
2012-02-29 23:55:50 wlad mkanat, no.. Still we're better in C/C++  rather than Python
2012-02-29 23:56:54 wlad still looks interesting
2012-02-29 23:58:16 wlad why don't people write an OData interface to MySQL I'm wondering..