IRC log of #maemo-ssu for Tuesday, 2013-03-26

*** sunny_s has quit IRC00:03
*** NIN101 has quit IRC00:04
*** sunny_s has joined #maemo-ssu00:06
*** nox- has joined #maemo-ssu00:09
*** freemangordon has quit IRC00:10
*** freemangordon has joined #maemo-ssu00:12
*** arcean has quit IRC00:27
*** xes has quit IRC00:39
*** dhbiker has quit IRC01:09
*** Martix has quit IRC01:48
*** unclouded has joined #maemo-ssu02:12
*** kolp has quit IRC02:50
*** LaoLang_cool has joined #maemo-ssu03:12
*** LaoLang_cool has quit IRC03:18
*** amiconn has quit IRC03:24
*** amiconn has joined #maemo-ssu03:29
*** M4rtinK2 has quit IRC04:18
*** nox- has quit IRC04:37
*** M4rtinK has joined #maemo-ssu05:01
*** amiconn has quit IRC05:10
*** amiconn_ has joined #maemo-ssu05:10
*** amiconn_ is now known as amiconn05:10
*** M4rtinK has quit IRC05:12
*** keesj has quit IRC05:33
*** keesj has joined #maemo-ssu05:34
*** DocScrutinizer05 has quit IRC06:04
*** DocScrutinizer05 has joined #maemo-ssu06:04
*** sunny_s has quit IRC06:15
*** sunny_s has joined #maemo-ssu06:18
*** M13 has joined #maemo-ssu06:29
*** LauRoman has joined #maemo-ssu07:17
*** LauRoman has quit IRC08:06
*** Milhouse has quit IRC08:14
*** Milhouse has joined #maemo-ssu08:18
*** dhbiker has joined #maemo-ssu08:31
*** sunny_s has quit IRC09:14
*** sunny_s has joined #maemo-ssu09:17
*** PaulFertser has quit IRC10:01
*** PaulFertser has joined #maemo-ssu10:03
*** Natch_i has joined #maemo-ssu10:58
*** ZogG has quit IRC11:01
*** ZogG has joined #maemo-ssu11:02
*** ZogG has quit IRC11:02
*** ZogG has joined #maemo-ssu11:02
*** keesj has left #maemo-ssu11:03
*** Natch has quit IRC11:06
*** Natch_i is now known as Natch11:06
*** M4rtinK has joined #maemo-ssu11:08
*** kolp has joined #maemo-ssu11:10
*** M4rtinK has quit IRC11:29
*** Milhouse has quit IRC11:34
*** futpib has joined #maemo-ssu11:38
*** andre__ has joined #maemo-ssu11:43
*** andre__ has joined #maemo-ssu11:43
*** M13 has quit IRC12:02
merlin1991freemangordon: pong12:25
*** Martix has joined #maemo-ssu12:31
*** lizardo has joined #maemo-ssu12:33
*** unclouded has quit IRC12:58
*** amiconn has quit IRC12:59
*** amiconn_ has joined #maemo-ssu12:59
*** amiconn_ is now known as amiconn12:59
*** sunny_s has quit IRC13:14
*** sunny_s has joined #maemo-ssu13:16
freemangordonmerlin1991: pong13:56
merlin1991you pinged me ;)13:57
freemangordonmerlin1991: yep. what are the problems of CSSU repo and who is aware of them?13:57
merlin1991problem is that upload directory does not has proper permissions13:58
merlin1991aware are the important people13:58
freemangordonmerlin1991: do they have names?13:58
freemangordon:)13:58
merlin1991we're basically waiting for a mail from X-Fade to get it up again13:58
*** sunny_s has quit IRC13:58
freemangordonX-Fade?13:58
merlin1991yes X-Fade13:58
freemangordonWhy X-Fade, afaik repos are under community control.13:59
freemangordonthat guy warfare is sysop13:59
freemangordon(iirc)13:59
merlin1991freemangordon: I even have an account there14:00
merlin1991but there is no trace of the garage group -- repoaccess sync on the servers14:00
merlin1991I can upload with my maintainance account for now14:00
merlin1991but only when I get back to vienna :)14:00
merlin1991sitting in HEL on my way back14:01
merlin19915 hours of airport14:01
merlin1991AGAIN14:01
merlin1991yay14:01
*** sunny_s has joined #maemo-ssu14:01
freemangordonmerlin1991: ok, there are some problem. What i don;t understand is why you expect X-Fade ti fix that?14:01
freemangordons/ti/to/14:01
infobotfreemangordon meant: merlin1991: ok, there are some problem. What i don;t understand is why you expect X-Fade to fix that?14:01
merlin1991I expect X-Fade to tell us how it used to work14:01
merlin1991so we can get it run the same again14:01
merlin1991because otherwise chemist has no access whatsoever14:01
freemangordonmerlin1991: for how long do you wait for that ansewr?14:01
merlin1991I'll get -testing up to speed regardless of that answer14:02
freemangordonmerlin1991: see, my point is that now we have control over the repos, I guess it is a god idea problems to be discussed with sysop, whoever he is14:03
merlin1991freemangordon: we did discuss with him14:03
merlin1991but he has no clue aswell14:03
freemangordonooh, so he waits for X-Fade's answer too?14:04
freemangordonmerlin1991: please, answer, for how long?14:04
*** andre__ has quit IRC14:04
freemangordonapproximately :)14:04
merlin1991I will upload to -testing when I get back home14:05
merlin1991only -stable will have to wait for the answer14:05
freemangordonthats great, but this is not answering my question ;)14:05
merlin1991well without X-Fades answer we can#t fix it14:06
merlin1991only squash in some temporary solution14:06
freemangordonmerlin1991: we don;t have expertise to assign some group/user rights? come on14:06
merlin1991freemangordon: we don't have expertise to assign some group/user rights based on some group somewhere in garage14:07
merlin1991we can by hand add people, but the whole idea was to have the webinterface to do this seamless without sysop interference14:08
merlin1991there's a reason https://garage.maemo.org/projects/cssu-stable/ and https://garage.maemo.org/projects/cssu-testing/ exist14:08
freemangordonmerlin1991: sure, UI is a good thing, but afaik we have 2 maintainers. I guess giving rights to them would't take that much time, even if done by hand14:09
merlin1991yea14:09
merlin1991anways I have enough acess to upload to -testing for now14:09
freemangordonmerlin1991: and what about -stable? I know you know how fast was X-Fade when Nemein were paid, what do you thing, where in the priority list is Maemo now?14:11
freemangordon*think14:11
*** andre__ has joined #maemo-ssu14:11
merlin1991I can upload to -stable aswell14:11
freemangordoncan chem|st upload too?14:11
merlin1991nope14:11
freemangordon:(14:12
freemangordoncould a "temporary" solution be made for him?14:12
freemangordonmerlin1991: or you will upload (this time only) to stable too?14:13
merlin1991probably what you just said14:13
freemangordonok14:14
freemangordonkerio: ping14:14
freemangordonkerio: I know you have a lot of free time :P. Do you wanna test fmrx-package? To check if we can push a new version, which will get installed on top of Nokia's?14:15
RaimuI can be aa tester too.15:28
Raimu-a15:28
*** sunny_s has quit IRC15:30
*** sunny_s has joined #maemo-ssu15:32
*** LauRoman has joined #maemo-ssu15:38
DocScrutinizer05[2013-03-26 13:50:47] <mashiara> Basic unix filesystem permissions and group mapping. Garage groups are15:40
DocScrutinizer05[2013-03-26 13:50:48] <mashiara> created from the garage database by a cron job every 5 minutes or so.15:40
DocScrutinizer05[2013-03-26 13:50:48] <mashiara> - Niels15:40
DocScrutinizer05[2013-03-26 13:53:53] <merlin1991> well we do need that cronjob :D15:40
DocScrutinizer05[2013-03-26 13:54:51] <mashiara> I suppose se script is found on garage15:40
DocScrutinizer05[2013-03-26 13:55:16] <merlin1991> I'll have a look when I'm back in vienna15:40
DocScrutinizer05FYI15:41
DocScrutinizer05<<duck and cover! spam heading in>>15:48
DocScrutinizer05# User and group export for midgard15:48
DocScrutinizer05#15:48
DocScrutinizer0535 * * * *      root    /root/garage_maintenance/garage_executer.sh midgard/create_users_csv.php /var/tmp/gforge_users.csv15:48
DocScrutinizer0535 * * * *      root    /root/garage_maintenance/garage_executer.sh midgard/create_groups_csv.php /var/tmp/gforge_groups.csv15:48
DocScrutinizer0535 * * * *      root    /root/garage_maintenance/garage_executer.sh midgard/create_members_csv.php /var/tmp/gforge_members.csv15:48
DocScrutinizer05# script for autobuilder. Updates email addresses from csv file to passwd15:48
DocScrutinizer0550 *    * * *   root    /var/tmp/uui /var/tmp/gforge_users.csv > /dev/null 2>&115:48
DocScrutinizer05#create system accounts from garage and sync to drop15:48
DocScrutinizer05*/10 *   * * *   root    /usr/local/bin/add_groups_users_git_ssh.sh > /tmp/add_groups_users_git_ssh.log dev/null 2>&115:48
*** arcean has joined #maemo-ssu15:55
DocScrutinizer05merlin1991: http://privatepaste.com/fe7f61897116:09
DocScrutinizer05merlin1991: no group cssu-testing or cssu-devel16:12
DocScrutinizer05err cssu-stable16:12
merlin1991DocScrutinizer05: what exactly is that privatepaste?16:16
DocScrutinizer05tmp/add_groups_users_git_ssh.log16:17
DocScrutinizer05excerp16:17
merlin1991hm probably generated from https://garage.maemo.org/projects/cssu/16:17
DocScrutinizer05obviously16:17
merlin1991but that's not what we want we want the cssu-testing and -stable bits16:17
merlin1991arf16:17
DocScrutinizer05yes, and those are not even mentioned in that log16:18
merlin1991perfect :/16:18
DocScrutinizer05which is a bit strange, since i'd think they are garage groups like all the other 544 groups processed in that log16:18
merlin1991possibly the reason why the upload never worked for chemis?16:18
DocScrutinizer05yep, definitely16:19
merlin1991simply the wrong group at work?16:19
DocScrutinizer05nfc what /usr/local/bin/add_groups_users_git_ssh.sh does exactly16:19
merlin1991I'll look into that16:19
merlin1991at least now we have a proper starting poing :)16:20
merlin1991s/poing/point/16:20
infobotmerlin1991 meant: at least now we have a proper starting point :)16:20
DocScrutinizer05but to me it seems it parses *all* garage groups and creates user accounts for each such group16:20
DocScrutinizer05but it's missing out on cssu-s and cssu-t16:20
DocScrutinizer05real n*f*c what makes those two groups different16:21
*** M4rtinK has joined #maemo-ssu16:21
merlin1991I'll have a look16:21
merlin1991a little bit of digging should bring up everything :)16:21
DocScrutinizer05yep, sure16:22
DocScrutinizer05I'd appreciate when you take over here16:22
DocScrutinizer05ping me when you need help with permissions16:22
merlin1991will probably need help for midgard/create_groups_csv.php which seems to sit inside /root/*, but can't tell yet because I'll start digging in vienna16:23
DocScrutinizer05ooooh, groups are from groups.csv??16:23
DocScrutinizer05then it's highly likely that cssu-t|s are missing in there16:24
merlin1991probably16:24
DocScrutinizer05gforge_groups.csv                                                                                                                                                122075/119K              100%16:26
DocScrutinizer05"cssu-testing";"cssu-testing uploaders";"cssu-testing.garage.maemo.org";"1";16:26
DocScrutinizer05"cssu-stable";"cssu-stable uploaders";"cssu-stable.garage.maemo.org";"1";16:26
DocScrutinizer05"16:26
DocScrutinizer05(for reference:)16:27
DocScrutinizer05"sleepscreen";"Sleep Screen Changer";"sleepscreen.garage.maemo.org";"1";16:27
DocScrutinizer05"ubi";"Ubi";"ubi.garage.maemo.org";"1";16:27
DocScrutinizer05"pengupop-maemo5";"Pengupop";"pengupop-maemo5.garage.maemo.org";"1";16:27
DocScrutinizer05...16:29
DocScrutinizer05gforge_members.csv                                                                                                                                                20282/138K               14%16:29
DocScrutinizer05"chemist";"cssu";"A";"Admin";16:29
DocScrutinizer05"chemist";"cssu-stable";"A";"Admin";16:29
DocScrutinizer05"merlin1991";"cssu";"A";"Default";16:30
DocScrutinizer05"merlin1991";"cssu-stable";"A";"Default";16:30
DocScrutinizer05"merlin1991";"cssu-testing";"A";"Default";16:30
DocScrutinizer05"merlin1991";"gforge";"A";"Admin";16:30
DocScrutinizer05...16:32
DocScrutinizer05gforge_users.csv                                                                                                                                                 692350/3562K              18%16:33
DocScrutinizer05"chemist";"chemist@dostortuXXX.oXX";"RĂ¼diger";"Schiller";"";16:33
*** Martix has quit IRC17:08
*** Martix has joined #maemo-ssu17:09
*** freemangordon has quit IRC17:13
*** freemangordon has joined #maemo-ssu17:14
*** Pali has joined #maemo-ssu17:32
*** sunny_s has quit IRC17:57
*** sunny_s has joined #maemo-ssu18:00
Paliping Estel_, kerio18:24
kerioyou just did18:24
kerioPali: battery.reporting.design is different from battery.reporting.last_full18:25
keriothe battery applet should use battery.reporting.last_full18:25
keriodesign... meh, it can come from both bq27k and rx51 i think18:25
keriodoesn't bq27k also have a way to read the resistor?18:25
Palikerio, no18:25
kerioanyway, the battery applet should use last_full, not design18:26
kerioand even without calibrating, last_full from bq27k is more accurate than rx5118:26
Palikerio, so that bool flag to do not use rx51_battery design value will be for applet and when disabled applet will use last_full instead design18:26
PaliOK?18:27
keriok18:27
keriothere's still the issue of not having a way to read the sysfs nodes when bq27k isn't calibrated18:27
Palikerio, I do not agree that bq is always better then rx51 when is not calibrated18:27
Paliif you swap batteries it can be totally bad18:28
Paliso rather default value is to use rx51_batttery when battery is not calibrated18:28
Palikerio, temp solution is to use register files18:29
keriothis is a battery applet issue though18:29
keriothat's just UI :)18:29
keriodoes hald-addon-bme do its own calculations for percentages and stuff like that, or does it use the percentage from bq27k?18:29
kerioPali: when the gconf flag is set to 0, the battery applet should show "no data, please calibrate" or something like that18:30
kerio(if bq27k isn't calibrated)18:30
kerioactually, no18:30
keriojust use last full, and show <current charge>/<last full charge> (CI)18:31
Palikerio, when battery is not calibrated it trying to guess bars (value 0-8) from voltage18:31
Paliand it report percentage (calculated from bars)18:32
kerioPali: i think that there should be a flag in hald-addon-bme that reports if the battery is calibrated or not18:32
kerioand there should be a way for the user to tell the battery applet to always consider bq27k data as valid18:33
keriolike DocScrutinizer05 would probably do18:33
kerio(except maybe show a small "(CI)" when not calibrated, so you also know when to calibrate again)18:33
PaliI can add CI flag to applet18:34
kerionormal users might prefer the guesstimate and the data from rx51-battery18:34
keriobecause they just don't want to bother with calibration18:34
Palikerio, write to wiki page what applet should shown when battery is calibrated and when is not calibrated18:34
kerioPali: will you also add some kind of UI for this, maybe turning the battery entry in the status menu into a button to access it?18:35
PaliI'm not planning to rewrite UI18:36
Palino time for it18:36
keriobut is it technically possible?18:36
Paliadding/changing text is easy18:36
kerioit could be merged with usbmode too18:36
Palikerio, it is possible, but I do not know who will do it...18:37
kerioi heard freemangordon is AWESOME at gtk :D18:37
freemangordonhehe18:37
freemangordonwho told you that?18:37
Palikerio, rather thinking about features, fix bugs18:37
kerioPali: added my proposal18:42
DocScrutinizer05some basic stuff about shutdown: there's a level for graceful shutdown (maybe EDV1 is a good voltage for that), that needs the be tripped constantly for a rather long guard time (60s) before *anything* happens. To ensure it's constantly tripped you need set a flag "shutdown-threshold-tripped" when you detect it got tripped, then probe frequently (like 1/s) , and you should reset the flag and stop probing frequently and reset that18:57
DocScrutinizer05flag when you detected 2 or 3 high values during your guardtime. When guard time of e.g. 60s expired and flag is still set, you "suggest" system to shut down and you start some power saving maybe, I.E. stop extreme powerhogs like phonecalls and camera and video playback.18:57
DocScrutinizer05Then there's another lower threshold that's hardcoded into GAIA TPS65950 which triggers when dead end for system is near from voltage level, and *that* one is supposed to trigger a power failure interrupt on kernel level and shuts down system the fastest way a linux system can do18:59
kerioDocScrutinizer05: yeah but EDVF causes brownouts19:00
kerioand similar voltages, i assume19:00
DocScrutinizer05userland gets notification about such power-failure event from kernel, but kernel will not wait for anything and go umount hard devices and then shutdown hard19:00
kerioDocScrutinizer05: the same as an alt+sysrq+U S B?19:06
kerioi mean, does the Sync happen?19:06
DocScrutinizer05on a power-failure? nope, takes too long19:06
*** NIN101 has joined #maemo-ssu19:06
DocScrutinizer05kernel immediately stops writing to all storage19:07
DocScrutinizer05to avoid aborting a write with power outage which would probably ruin the HDD19:07
DocScrutinizer05or the uSD/eMMC for that matter19:07
DocScrutinizer05power-failure shutdown shall be clearly <1s to complete19:09
DocScrutinizer05that's btw what I also suspect bme does when BSI gets disconnected. That's the reason why BSI contact blade is shorter then the +VB and GND blades of bat contact, so when you unplug the battery the kernel gets some milliseconds to stop everything before battery power gets disconected19:12
keriowat, is it really that well-designed?19:24
*** Woody14619a has joined #maemo-ssu19:31
*** Woody14619a has quit IRC19:31
*** Woody14619a has joined #maemo-ssu19:31
DocScrutinizer05you got a device to look at it, no?19:31
DocScrutinizer05and you already know that with active bme the device doesn't work as soon as battery gets removed19:32
DocScrutinizer05so evidently bme cares about BSI19:33
DocScrutinizer05and you're free to check how fast it shuts down, by simply removing battery while it's charging and thus supposed to be powered from USB19:34
*** Woody14619 has quit IRC19:34
DocScrutinizer05the only (res primary) reason device shuts down when you remove battery in that situation is bme doing that, when it detects BSI going ..00..19:35
DocScrutinizer05see my exlanations in tmo thread "hot swapping of battery, now possible"19:36
kerioi didn't mean the behaviour, i meant the "contact blade for BSI is shorter than + and GND so that when removing the battery the device has time to shut down"19:44
kerioi thought it was different just because19:45
keriowell, """time""" to """shut down"""19:45
*** Woody14619a is now known as Woody1461919:48
*** sunny_s has quit IRC19:57
*** sunny_s has joined #maemo-ssu20:00
DocScrutinizer05I'm rather sure it's >1ms just for the shorter blade of BSI. And then we also got a buffer C (even a few of them) so the system has some 10s of ms for emergency shutdown20:00
DocScrutinizer05particularly since cmt has a direct line to BSI and thus the biggest power hog has a way to stop consuming power even before APE had a chance to signal power-failure to it via interface20:02
kerioi still don't get what can be done in 10ms to help with shutdown20:04
DocScrutinizer05what can be done? I guess 3..8 million opcodes can get executed20:09
kerioand one seek? :)20:11
DocScrutinizer05no seeks20:18
DocScrutinizer05power-down20:18
DocScrutinizer05abort or finish writes20:19
*** sunny_s has quit IRC20:55
Palikerio, I updated wiki bme, look20:57
kerioPali: "$charge/$last_full_charge (CI)"20:58
kerioi just realized that i haven't been thinking about the hybrid situation of no bme but no full bme replacement20:58
keriobut meh, that situation can go fsck off20:59
*** sunny_s has joined #maemo-ssu20:59
kerioPali: hold on, why bother with "shutdown at edv1" if you're also going to "shutdown at 3150"?21:03
keriooh nvm, edv1 is higher than 315021:03
Palikerio, because default value is shutdown at edv121:17
Palinot at specified voltage21:17
keriooh so it doesn't shutdown at the voltage if it's told to shutdown at edv1?21:24
*** dhbiker has quit IRC21:24
*** xes has joined #maemo-ssu21:32
Palikerio, both21:32
kerioi see21:32
Palikerio, but if voltage to too low, edv1 will happen first21:33
kerioyeah, yeah :)21:33
*** dhbiker has joined #maemo-ssu22:18
*** futpib has quit IRC22:59
*** sunny_s has quit IRC23:04
*** lizardo has quit IRC23:04
*** xes has quit IRC23:09
*** Martix has quit IRC23:11
*** Vlad_on_the_road has joined #maemo-ssu23:23
*** NIN101 has quit IRC23:43

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!