**** BEGIN LOGGING AT Wed Jul 10 02:59:59 2013 Jul 10 03:00:11 night all Jul 10 05:11:21 clear Jul 10 05:28:36 Did my battery calibration last night. How to get constant drain on battery if keeping the backlight on is not possible? Solution: connect to wlan and wget stuff. Jul 10 05:28:41 for i in 1 2 3 4 5 ; do wget -O /dev/null ftp://ftp.funet.fi/dev/100Mnull --limit-rate=50K; done Jul 10 05:29:04 I think the rate limiter is essential, otherwise one gets too much data and the load is much higher Jul 10 05:29:44 CSOC: 5 % RSOC: 5 % Jul 10 05:29:45 Average Current: 109 mA Jul 10 05:30:14 Last Measured Discharge: 1271 mAh Jul 10 05:31:01 Now my battery is reporting more than 1100mAh when it's fully loaded Jul 10 05:54:25 Last Measured Discharge: 1034 mAh Jul 10 05:54:27 :( Jul 10 06:16:46 I have genuine Nokia battery, purchased less than year ago. Jul 10 06:17:53 The previous battery was really bad. Putting some load from system and the battery which appeared to be quite full went empty in no time. Jul 10 07:46:38 whats up with the wiki.maemo.org? Jul 10 07:46:49 semms like there are no styles Jul 10 07:47:05 and graphics are missing Jul 10 08:29:13 Pali: what does your battery applet read for its info Jul 10 08:40:03 Lava_Croft: battery applet read data from BME or BME replacement or from battery kernel drivers (via HAL) Jul 10 08:40:37 priority: bme replacement; bme; kernel drivers (via HAL) Jul 10 08:41:11 when you stop bme it automatically try to use any other active input source Jul 10 08:41:27 and when nothing is possible it will show something line no data or 0% Jul 10 08:42:34 Pali: I ask because I ran Doc's battery calib script and both it and bnf report a different value than the applet Jul 10 08:42:38 this is normal? Jul 10 08:42:55 seems the applet still thinks its on a 1250ish mAh battery Jul 10 08:43:05 while it actually is a 1500ish mAh battery Jul 10 08:43:40 if battery is not calibrated it calculating capacity against design capacity reported by 3rd battery pin Jul 10 08:44:02 it is calibrated, ran the cycle twice with nearly same results Jul 10 08:44:11 also in last version there is gconf value which configure when to use design capacity Jul 10 08:44:28 make sure you have last version from cssu-devel Jul 10 08:45:34 jan09 Jul 10 08:46:37 above was fixed in version 1.0-5 (27 Mar 2013) Jul 10 08:46:54 see info in git commit: https://gitorious.org/community-ssu/status-area-applet-battery/commit/3d3599743a598a833c1ecfa886ebf53fb6d11a12 Jul 10 08:46:56 ok, the one on skeiron is old then Jul 10 08:47:42 http://maemo.merlin1991.at/cssu/community-devel/ Jul 10 08:47:57 http://maemo.merlin1991.at/cssu/community-devel/pool/free/s/status-area-applet-battery/ Jul 10 08:48:03 yeah, am there Jul 10 08:48:04 thanks:D Jul 10 08:48:33 see git commit for configuring calculating percentage Jul 10 08:48:33 silly how attached you can get so that simple applet Jul 10 08:48:40 thank you Jul 10 09:04:56 i wonder where im failing this:| Jul 10 09:05:37 gconftool-2 --set /apps/osso/status-area-applet-battery -t int 2 Jul 10 09:06:25 Pali: ^^ that is correct? Jul 10 09:07:41 2 will always use design capacity reported by battery and will ignore calibrated max capacity Jul 10 09:08:07 but command could be correct Jul 10 09:10:06 do i need to reboot or something Jul 10 09:10:25 trying to see any kind of change in the reported value Jul 10 09:12:38 after installing new version of plugin you need to reboot device Jul 10 09:12:50 but changing that gconf key is instant (no reboot needed) Jul 10 09:13:08 hm, i didnt see any changes when changing the value to any of the valid settings Jul 10 09:13:48 just keeps reporting 1252mAh, while it should be 1557mAh Jul 10 09:25:38 well, i have no clue why it doesnt work Jul 10 09:25:57 Ill just remove the applet, since getting the wrong values is annoying :< Jul 10 10:55:59 Pali: using bme replacement solved it Jul 10 10:58:29 wonder if it can charge yet Jul 10 11:00:00 it can, fantastic Jul 10 11:02:55 Lava_Croft: bme replacement working fine on my n900 without problems Jul 10 11:03:02 but it is still experimental Jul 10 11:03:11 yeah i know, but it already has come quite a way since before Jul 10 11:03:33 the replacement BME does report the right values Jul 10 11:03:38 couldnt get it done with the stock one Jul 10 11:03:53 maybe it is time to announce bme replacement on TMO :-) Jul 10 11:04:05 ~bme-replacement Jul 10 11:04:05 [bme-replacement] http://atrey.karlin.mff.cuni.cz/~pali/rx51-bme-replacement http://atrey.karlin.mff.cuni.cz/~pali/projects/maemo/bme-replacement.html See also: http://wiki.maemo.org/Bme_replacement . Please, use wiki page to report bugs/problems and/or solutions to them! Jul 10 11:04:06 there is a thread about it already, but a proper thread might not be bad Jul 10 11:09:34 DocScrutinizer05, any plans concerning www? Jul 10 11:19:35 not really Jul 10 11:19:38 yet Jul 10 11:20:15 [2013-07-10 13:01:49] mashiara: this is very first line of that kind that shows up in www VM maemo_http_error.log: Jul 10 11:20:17 [2013-07-10 13:01:53] [Fri Dec 28 09:31:26 2012] [error] [client 10.0.0.1] PHP Warning: (pid:20625): Can not write property. Object in read only mode in /usr/share/php/midcom/lib/midcom/core/dbaobject.php Jul 10 11:20:38 [2013-07-10 13:02:33] since then it never stopped Jul 10 11:20:40 [2013-07-10 13:02:49] could you give any clue what has been done that is causing this? Jul 10 11:21:11 (FYI: it seem 2012-12-28 been date of migration#1) Jul 10 11:23:18 not really as the install package of Midgard is incomplete. As such I can only look at the code and pray that I can find a culprit. Yesterday I also had a late night conversation with Mashiara and he/she promised to get back to me today with more info. Apparently the missing channel package can be copied from the current maemo install, but I will receive more detailed instructions today. Hence also the question, should I conti Jul 10 11:23:18 nue trying to get Midgard to run locally or is it a waste of time and energy? Jul 10 11:25:19 root@www1:/var/log/apache2# cat */*/*/maemo_http_error.log |grep -c "Can not write property" Jul 10 11:25:21 1537712 Jul 10 11:25:52 not really what? Jul 10 11:27:38 anthonie_: I think we can't gain much by making another local install of midgrad1 run anywhere Jul 10 11:27:48 Normally when an object turns into read only mode, it's either a configuration problem of the software, or a I/O problem hardware wise... Glancing at yesterdays logs, there seems to be enough diskspace, so I say it's either a CPU lockup (which would be consistent with the overload) or the Midgard code. Something makes me think it's the latter but it's too early to say anything before it actually runs. Jul 10 11:28:01 if you like to do it I won't stop you nevertheless Jul 10 11:29:01 cpu lockup is not possible a reason Jul 10 11:29:29 >>Object in read only mode in /usr/share/php/midcom/lib/midcom/core/dbaobject.php<< Jul 10 11:29:59 It's not a atter of "liking it", i am afraid, but it's a question of necessity. Jul 10 11:30:54 then let me put it this way: there is no obvious necessity to have another local midgard1 instance running Jul 10 11:31:37 particularly since you hardly can reproduce the errors we encounter on maemo.org Jul 10 11:32:00 so the or better, my question is, what's the plan? Is it usefull putting more hours into Midgard1 (I am willing to do that, I just need to be sure we want to continue down this road) or do we plan on a forced upgrade and get it over with? In the end, Midgard1 needs to be upgraded soon enough anyway. Jul 10 11:32:28 so unless you need such installation for your personal educational puposes, I don't know why and how we could benefit from it Jul 10 11:33:06 No, hopefully the errors wouldn't be reproduced, but the idea was to get a hint of what's going on with Midgard from the perspective of a working install, rather than jumping into a pool full of trouble with a system I never used before. Jul 10 11:33:30 there is no sesnible migration path from midgard1 incl all the custom maemo addons, to any other CMS incl midgard2 Jul 10 11:33:52 I was already afraid and expecting to hear that... Jul 10 11:34:21 if there were, probably nemein (inventors of midgard) whould have done it on maemo (where nemein been the primary/sole maintainer) quite a long time ago Jul 10 11:34:25 even more reason to gather and discuss what needs to be done, if anything... Jul 10 11:35:56 look, all that stuff worked at a certain point (or span) in time, some 12..24 months ago Jul 10 11:36:50 yes? Jul 10 11:36:55 we are still trying to recover from damage done during (and immediately prior to) migration#1 Jul 10 11:37:22 but you're juggling god knows how many balls here... Jul 10 11:37:59 there's still an undelivered package in contract items between Nemein and Nokia (and community) that's titled "knowledge transfer" Jul 10 11:38:16 anthonie_: (balls) yes. So what? Jul 10 11:38:31 what's the alternative? Jul 10 11:39:39 it's going to be very hard to keep sane with all this mess... There is no real alternative besides either going in and try and fix error for error, or trying to get a clean and working system up and running again and fill it with content from there. Jul 10 11:40:13 there are exactly 3 guys who have a clue about maemo.org infra nad how stuff is supposed to work: x-fade/Niels mashiara/Rambo/Eero Ferenc Jul 10 11:40:26 the situation right now is: A clean install can not be made without all kinds of tricks and trouble Jul 10 11:40:47 anthonie_: the second suggestion is tempting but infeasible Jul 10 11:41:06 of those, only Mashiara is avaible on Midgard channel Jul 10 11:41:14 it would take man-years of work Jul 10 11:41:30 yes Jul 10 11:41:36 I know, it is tempting, but I did not know two days ago it was THIS urgent... Jul 10 11:42:03 x-fade been around here, but it was his phantom only Jul 10 11:42:38 dilemmas... Jul 10 11:42:41 wauw Jul 10 11:42:52 replacement bme -> first time in nearly 4 years that i see "Fully Charged" Jul 10 11:42:56 again, it's NOT primarily midgard that gives us headache, it's the custom addons Jul 10 11:43:55 which ones are that? Jul 10 11:44:10 anthonie_: if you're willing to learn about how midgard *basically* works, you'd be the ideal candidate for the aforementioned knowledge-transfer Jul 10 11:44:47 the transfer hasn't taken place because there is no one to "receive" it? Jul 10 11:45:35 (which ones) basically they are known as "autobuilder", "repo-management", "karma", a lot of small others like "ask for maintainership" function and whatnot else Jul 10 11:45:58 account syncing from garage to www to repo to drop to .... Jul 10 11:46:23 12 dozen nasty php and python scripts Jul 10 11:46:32 :) Jul 10 11:46:36 jay! Jul 10 11:46:51 ...that do "promote package", "import package to repo", etc pp Jul 10 11:47:16 midgard itself is only "the GUI" Jul 10 11:48:00 we even have another system complementary to midgard, on garage Jul 10 11:48:07 I forgot the name Jul 10 11:48:21 Huh? up and running? Jul 10 11:48:27 yes, sure Jul 10 11:48:42 garage db is postgresql Jul 10 11:49:00 you know what? I'll talk to mashiara and discuss with him to see what would be the wisest path from here Jul 10 11:49:19 if you think that's a good move Jul 10 11:50:00 he most likely will tell you "please discuss this with the new maintainers of maemo" - ergo in the end me :-S Jul 10 11:50:03 the thing is, if we start troubleshooting remotely straight away, chances are we need a lot more help from him than we may be able to get. Jul 10 11:50:25 uhuh Jul 10 11:50:26 lol, hello Doc, I would talk to you about our patient... Jul 10 11:53:16 you seen my initial post of today? answer to your initial question? Jul 10 11:53:34 no, on tmo? Jul 10 11:54:17 [2013-07-10 13:09:34] DocScrutinizer05, any plans concerning www? Jul 10 11:54:19 [2013-07-10 13:19:35] not really Jul 10 11:54:24 [2013-07-10 13:20:15] [2013-07-10 13:01:49] mashiara: this is very first line of that kind that shows up in www VM maemo_http_error.log: Jul 10 11:54:56 also please varefully re-read what I posted regarding "knowledge-transfer" in last maybe 20 lines here Jul 10 11:55:01 carefully* Jul 10 11:55:10 will do Jul 10 11:55:39 [2013-07-10 13:44:10] anthonie_: if you're willing to learn about how midgard *basically* works, you'd be the ideal candidate for the aforementioned knowledge-transfer Jul 10 11:56:10 I'm supposed to coordinate said knowledge transfer sessions Jul 10 11:56:27 I know, I did read that, bloody scary man, considering the amount of work but if it needs to be done it needs to be done Jul 10 11:58:34 those knowledge transfer sessions are not meant to teach knowledge about basic midgard principles, or about basic operation and system design of maemo autobuilder (the latter maybe a bit more than the former). We have some wiki pages that for example summarize what jacekowski found out about how cauldron autobuilder in maemo flavour is set up and working Jul 10 11:59:36 think of it like "you got 30 questions plus follow up clarification free to ask, mashiara will answer" Jul 10 12:00:59 it's basically a support contract? Jul 10 12:01:27 no, it's knowledge transfer Jul 10 12:01:45 fair enough Jul 10 12:01:55 ok, so what would need to be done in order to achieve that transfer and how do we do that? Jul 10 12:08:25 learn about maemo.org infra internals, get a guest/user account in the relevant VMs so you can inspect scripts etc, chat with others like you who also want to participate, discuss the questions you have, write them down and and them to me, then I coordinate a session with mashiara/ferenc/niels,probably on IRC in a dedicated channel Jul 10 12:09:09 and hand* them to me... Jul 10 12:09:29 clear, let's do that Jul 10 12:11:00 http://wiki.maemo.org/Maemo.org_Infrastructure/Autobuilder Jul 10 12:11:05 is a start Jul 10 12:11:43 bookmarked as well as http://wiki.maemo.org/Migrating_to_Community-driven_Infrastructure Jul 10 12:12:30 http://wiki.maemo.org/Migrating_to_Community-driven_Infrastructure and http://wiki.maemo.org/Maemo.org_Infrastructure further entry points Jul 10 12:13:06 and of course google for cauldron Jul 10 12:15:07 http://extras-cauldron.garage.maemo.org/HOWTO.html Jul 10 12:16:25 all right, clear Jul 10 12:29:29 I will be reading those and get back to you Jul 10 12:47:44 mornin Jul 10 12:51:15 Drathir, morning there and afternoon here :D Jul 10 12:58:04 ~ugt Jul 10 12:58:05 from memory, ugt is Universal Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. http://www.total-knowledge.com/~ilya/mips/ugt.html Jul 10 13:01:13 DocScrutinizer05: nice i didnt know that O.o Jul 10 13:01:26 lol like that one :D Jul 10 13:03:45 Happy Halloween then! Jul 10 13:05:28 freemangordon: I sent new version of ARM errata 430973 workaround patches to LKML (CC to you) Jul 10 13:07:44 btw builider fixed? Jul 10 13:24:42 could some unix savvy check `man ps` and `ps -o start=anfang,user,pid -U root --sort=start;` and tell me if for them it sorts on start time, if yes why and if not then also why? Jul 10 13:24:53 for me it seems ps has a bug Jul 10 13:25:05 I get sort on PID always Jul 10 13:25:34 just like if --sort wasn't given at all Jul 10 13:28:12 sort on both PID and time Jul 10 13:32:21 hm? Jul 10 13:33:34 http://privatepaste.com/a7fd70ac82 doesn't look to me like it's sorting on start Jul 10 13:33:57 rather it's strctly sorting on pid only Jul 10 13:36:02 check, double check -> nope, http://privatepaste.com/e404030c03 Jul 10 13:40:25 well, that's pretty insignificant Jul 10 13:40:38 it's ordered on both starttime as well as pid Jul 10 13:41:44 yours even doesn't obey -U root though Jul 10 13:41:56 see "user: daemon" Jul 10 13:42:53 might be a glicht due to renaming/re-owning the process during ps execution time Jul 10 13:43:38 ps jaxkuid,-ppid,+pid - as according to man ps, works as expected here Jul 10 13:44:02 I wonder if that issue is related to "personality" for me as well as your result Jul 10 13:44:11 I am having a look Jul 10 13:45:02 ps -o start=anfang,user,pid -U root --sort=-start; Jul 10 13:45:11 *same* result Jul 10 13:45:38 ps -o start=anfang,user,pid -U root --sort=-pid; works as expected Jul 10 13:48:28 ps -o etimes=anfang,user,pid -U root --sort=-etimes; also borked Jul 10 13:49:00 without "=anfang" same Jul 10 13:49:07 :-/ Jul 10 13:49:34 no matter if root or regular user Jul 10 13:52:58 dunno Jul 10 13:54:09 hey is something wrong with nickserv? Jul 10 13:54:31 proppably services dont respond... Jul 10 13:54:41 ok Jul 10 13:54:42 but need to confirm... Jul 10 13:55:23 yes confirmed. **** BEGIN LOGGING AT Wed Jul 10 14:08:55 2013 Jul 10 14:16:55 ~ping Jul 10 14:16:56 ~pong Jul 10 14:17:19 wtf?? Jul 10 14:17:46 dat freenode Jul 10 14:17:53 FU nickserv! Jul 10 14:18:36 wow Jul 10 14:18:41 at least Jul 10 14:19:37 * DocScrutinizer05 headdesks Jul 10 14:25:10 what a shitty day to be infra-admin... I pity you, Doc Jul 10 14:38:05 add to that the fact I'm not even infra-admin, I'm infra-admin-coordinator/manager Jul 10 14:38:52 I feel with you Jul 10 14:38:54 [2013-07-10 16:34:20] [440] DocScrutinizer05 NickServ Services are currently unavailable Jul 10 15:58:10 a question to the php / apache cracks: would a ini_set('max_execution_time', 0); in a .php affect the apache process even after the particular .php finished? Jul 10 16:03:03 DocScrutinizer05, should end after the script ends Jul 10 16:04:35 DocScrutinizer05, there is, however, also a seperate time_out setting for apache, normally it can be found in httpd.conf Jul 10 16:05:34 what would happen to a running .php script that a apache process started that gets "killed" by such apache timeout? Jul 10 16:05:59 good question Jul 10 16:08:48 we got all clear signal for the "object in read-only mode" error-log, more or less Jul 10 16:09:55 but we still suffer form apache2 processes freezing and getting stuck one after the other, so somebody would need to kill them hard (-9!) every few hours, or www comes to a grinding halt Jul 10 16:10:34 I'm not willing to re-enable apache on www.m.o before we solved this issue Jul 10 16:10:37 all clear as in "it's a non-critical error message" or all clear as in "solved"? Jul 10 16:11:33 as in "there's a good explanations why it happens, and it probably doesn't cause any havoc (except spamming logfiles)" Jul 10 16:11:46 good enough for now Jul 10 16:13:49 since migration#1 we got some optimizations that include anonymous access to db, which results in objects r/o, but some midgard templates use direct modification of db-objects as in obj.timestamp=strtoint(obj.timestamp) Jul 10 16:33:56 anthonie_: feel free to make required "knowledge" pages in "wiki.maemo.org/Maemo.org_Infrastructure/*" I will eventually make that page link to all the pages within. Jul 10 16:35:02 FWIW, garage is gforge afaik Jul 10 16:42:52 yep, gforge Jul 10 16:43:08 that been the name I wasn't able to recall Jul 10 16:45:43 ~dict forge Jul 10 16:45:46 Dictionary 'forge' (1 of 14): \Forge\, v. i. [See {Forge}, v. t., and for sense 2, cf. {Forge} compel.] 1. To commit forgery. [1913 Webster] 2. (Naut.) To move heavily and slowly, as a ship after the sails are furled; to work one's way, as one ship in outsailing another; -- used especially in the phrase to forge ahead. --Totten. [1913 Webster] And off she [a ship] forged without a shock. --De Quincey. [1913 Webster]. Jul 10 16:47:10 ~dict forgery Jul 10 16:47:13 Dictionary 'forgery' (1 of 5): \For"ger*y\, n.; pl. {Forgeries}. [Cf. F. forgerie.] 1. The act of forging metal into shape. [Obs.] [1913 Webster] Useless the forgery Of brazen shield and spear. --Milton. [1913 Webster] 2. The act of forging, fabricating, or producing falsely; esp., the crime of fraudulently making or altering a writing or signature purporting to be made by another; the false making or material alteration of or addition to ... Jul 10 17:14:20 anthonie_: do you think the apache2 process could freeze completely when it tries to terminate due to timeout, while it runs a php script? Jul 10 17:16:42 yes, I think that's pretty well possible Jul 10 17:20:31 maemo.org???? Jul 10 17:21:33 what happened?? Jul 10 17:23:07 SAiF: the software, it no worky Jul 10 17:23:58 I mean, I cant acces maemo.org. Can you access it? Jul 10 17:37:17 /topic Jul 10 17:40:05 anthonie_: yes, that's what I think as well. If I had to guess I'd say apache sends a ^C to php which php ignores. Then apache waits for child php process to terminate which never happens since php ignored ^C and now is blocking on stdout or whatever to apache process Jul 10 17:42:17 ooops... Jul 10 18:09:11 Up up up Jul 10 18:09:15 Hey... Jul 10 18:10:50 www. is up Jul 10 18:17:31 Thanks to you guys.. good night Jul 10 18:21:42 we're just testing Jul 10 18:22:58 if my idea why apache freezes (resp the intermediate solution) doesn't pan out, it will go down again without prior notice, and I won't restart it until we have another idea how to fix this issue Jul 10 19:18:34 hello, I have a new strange bug. When I type to fast, applications close with 'internal error' Jul 10 19:18:55 ony me? Its hardware issue? Jul 10 19:36:39 hw error *might* be possible. Jul 10 19:37:02 though that wouldn't depend on fast but rather on "hard" typing Jul 10 19:37:30 never heard sth like that before Jul 10 20:40:56 intermediate solution didn't pan out? Jul 10 20:43:32 thank you DocScrutinizer05 Jul 10 20:48:14 GRRR Jul 10 20:50:48 GRRR like GRRRUNT? Jul 10 21:26:48 (*.net *.split *.hell) Jul 10 22:04:44 Good Night (UGT) Jul 10 22:52:08 22:38 -DocScrutinizer05:#maemo- is offline Jul 10 22:52:23 how you do that in color? Jul 10 22:52:28 freenode Jul 10 22:52:42 hmm? Jul 10 22:52:58 ooh Jul 10 22:53:10 never thought it made it to this chan Jul 10 22:53:46 yes yes that Jul 10 22:53:56 hah, I don't see it Jul 10 22:53:57 what it is? Jul 10 22:54:08 i see... Jul 10 22:54:16 sth you shouldn't use ;-) Jul 10 22:54:30 reserved for bots Jul 10 22:54:52 it's called a notice Jul 10 22:54:56 but looks nice... Jul 10 22:55:06 for those who see it Jul 10 22:55:40 red however 04is no problem Jul 10 22:55:59 yes red see also... Jul 10 22:57:12 also largely deprecated but not forbidden in this channel Jul 10 22:57:28 i also see lot time ago rainbow words... Jul 10 22:57:50 yes, that's when I will forbid it Jul 10 22:57:59 *that* somthings is usefuf one... Jul 10 22:58:09 sometimes* Jul 10 22:59:28 RED: Unknown command Jul 10 22:59:31 * DocScrutinizer05 headdesks a little Jul 10 23:00:03 http://talk.maemo.org/showthread.php?p=1358180#post1358180 Jul 10 23:19:10 suuure, I I'm bored to death, so I found sth exciting: checking MD5SUM of /var/tabletsdev/nokia_N810/downloads7* Jul 11 01:37:26 DocScrutinizer05: how about checksums matching? Jul 11 02:50:25 hi Jul 11 02:55:15 having trouble uploading to autobuilder.... does the web interface work? I can't get scp to work either but maybe my ssh key has not propagated to the system yet? **** ENDING LOGGING AT Thu Jul 11 02:59:58 2013