IRC log of #maemo-ssu for Wednesday, 2013-01-09

*** MrPingu has quit IRC00:00
*** MrPingu has joined #maemo-ssu00:01
*** ekze has quit IRC00:01
* amiconn is somewhat confused00:03
amiconnI have cssu thumb installed (updated from cssu-t), but now ham offers a (plain?) cssu upgrade. Should I remove/ disable the plain cssu repository?00:04
freemangordonamiconn: refresh00:04
amiconnI didn't migrate today, but a few days ago00:05
freemangordonstill, refresh, there is newer -thumb00:05
amiconnAh, today it's not the same offer, it's -thumb again00:05
amiconnBut anyway, should plain cssu-t repo be disabled or not?00:05
MrPingubasically thumb = testing00:06
freemangordonamiconn: no00:06
amiconnMrPingu: Yeah, but if testing gets an update (usually a bit earlier than thub) I suspect it would migrate a buch of packages back to non-thumb00:07
MrPinguamiconn: yes, that's right. Because thumb doesn't version number00:07
amiconnAls ham got even slower with the addition of the cssu-thumb repo00:07
amiconn*Also00:07
MrPingu*change00:07
MrPinguthat's also true00:08
MrPinguwhenever theres a new version of testing version numbers increase00:08
MrPinguso it pulls the latest but doesn't care if it's thumb or non-thumb00:09
amiconnYeah, hence my question00:09
amiconnThe only reason that would speak against disabling plain cssu-t repo is if the thumb repo is incomplete00:09
kerioyep00:09
amiconnBut this is what I don't know00:10
MrPinguEnjoy this update, you won't be getting a new update soon :P :P :P00:11
*** joshgillies has joined #maemo-ssu00:17
*** dafox has quit IRC00:19
*** MrPingu has quit IRC00:34
DocScrutinizer05amiconn: basically there's no proper upgrade strategy for cssu-thumb. You'll face same "issues" each time cssu-t gets a new upgrade. Since cssu-thumb is just a "patch" on top of cssu-t00:36
DocScrutinizer05for cssu-thumb to become a proper "distro" it would need to port all cssu-t packages to thumb repo and disable cssu-t00:37
freemangordonwhich makes no sense :)00:37
DocScrutinizer05for now it only has those pkgs that are actually thombified00:38
freemangordon:nod:00:38
DocScrutinizer05and that will cause conflicts with the pkgs in those repos it's basing on00:38
freemangordoncould some who has stock tklock installed test something for me?00:39
DocScrutinizer05as soon as there are upgrades that are not in thumb repo yet, but thumb has an older thumbified version of those pkgs00:39
freemangordonset display backlight timeout to 10 seconds, wait for screen to lock and unlock with slider. Now check after what time the screen will get locked again. is it 10s?00:40
DocScrutinizer05actually I think CSSU-T does same mess with original Nokia repos on which it is basing, but we rarely see issues there since upgrades on Nokia repos are highly unlikely00:40
freemangordonanyone? ^^^00:41
DocScrutinizer05wait 20 s ;-D00:41
freemangordonDocScrutinizer05: could you test that for me?00:41
freemangordonok00:41
*** BCMM has joined #maemo-ssu00:41
DocScrutinizer05dim after 10, off after another 4 orso00:43
freemangordondid you unlock with visual slider?00:43
DocScrutinizer05nah, with slider switch00:43
freemangordonno, use visual00:44
DocScrutinizer05err, how?00:44
freemangordonpress power button00:44
freemangordonand slide-to-unlock00:44
freemangordonyou may need to press power button teice00:44
freemangordon*twice00:44
DocScrutinizer05meh 30 and still on00:45
freemangordonyep00:45
DocScrutinizer05still on00:45
freemangordonI thought it is bug in tklock replacement :D00:45
freemangordonbut it seems it is by design00:45
DocScrutinizer05noooow00:45
freemangordonor bug in mce00:45
DocScrutinizer05unlocked00:46
freemangordonsome 60s?00:46
DocScrutinizer052000:46
DocScrutinizer053000:46
DocScrutinizer055000:46
DocScrutinizer056000:47
DocScrutinizer05dim after 7000:47
freemangordonnice :D00:47
DocScrutinizer05o/&00:47
freemangordonwell, not much that we can do, mce is closed :(00:47
DocScrutinizer05I never use that method to unlock00:48
freemangordonyou will ruin your slider swith ;)00:48
DocScrutinizer05lockswitch works as supposed00:48
freemangordonsure00:48
DocScrutinizer05I rather uin my powerbuttons usually, and scratch my screen00:49
freemangordonhowever, thanks00:49
DocScrutinizer05my lockswitch is fine00:49
DocScrutinizer05o/00:49
DocScrutinizer05afk00:49
DocScrutinizer05btw my timeout been 60 before I set it to 10 for you. 70 = 60 + 10 FWIW00:50
DocScrutinizer05no, I won't reboot with timeout set to 1000:50
freemangordonhmm, yeah, could be that00:50
freemangordonDocScrutinizer05: have 1 more minute?00:51
freemangordonfor another test00:51
DocScrutinizer05sorry, nope00:51
freemangordonok00:51
DocScrutinizer05post instructions, maybe later00:51
freemangordonDocScrutinizer05: while on desktop, wait for screen to dim(not lock) and press it. do you see h-d settings button to appear?00:52
DocScrutinizer05never00:52
freemangordonhmm, try it00:53
freemangordoni remember i've seen that, but I am not sure00:53
DocScrutinizer05I "try" that 20 to 100 times a day00:53
DocScrutinizer05first touch even get swallowed by something, during dim00:53
DocScrutinizer05I think there's a setting for that in mce.ini00:54
freemangordonbut not when on desktop00:54
freemangordonit is called "event eater" :D00:54
DocScrutinizer05ok, I'll try it00:54
DocScrutinizer05:-/00:54
DocScrutinizer05answer stays: nope00:56
freemangordonok, thanks00:56
DocScrutinizer05check mce.ini!00:56
freemangordonnah, it is tklock00:57
DocScrutinizer05*sigh*00:57
DocScrutinizer05[TKLock]00:58
DocScrutinizer05# Blank immediately instead of dim before blank when tklock is enabled00:58
DocScrutinizer05..00:58
freemangordonDocScrutinizer05: it is some bug in my replacement00:59
DocScrutinizer05# Policy for touchscreen interrupts00:59
DocScrutinizer05#00:59
DocScrutinizer05# 1 - disable immediately00:59
DocScrutinizer05# 0 - wait until display is blanked00:59
DocScrutinizer05DisableTSImmediately=100:59
DocScrutinizer05fool!00:59
freemangordon:D00:59
DocScrutinizer05when I suggest to check mce.ini, you should consider it00:59
freemangordonDocScrutinizer05: i've just reinstalled stock01:00
freemangordonand tested01:00
freemangordonsettings button does not appear01:00
freemangordonthis pretty enough for me01:00
DocScrutinizer05since stock checks DisableTSImmediately=101:00
freemangordonno, this tells mce what to send to tklock01:01
DocScrutinizer05if you "fix" yours, you're swapping one bug for another01:01
freemangordonand it is tklock who does the work01:01
DocScrutinizer05mhm01:01
DocScrutinizer05maybe01:01
freemangordonand I fail to "eat" the event when on the desktop01:02
freemangordonfor some reason I am going to check01:02
DocScrutinizer05:nod:01:02
*** BCMM has quit IRC01:03
DocScrutinizer05you're right, DisableTSImmediately=1 works globally, not only on desktop01:05
freemangordonyep01:05
DocScrutinizer05I never had a doubt on this, but it occurred to me just now01:05
DocScrutinizer05btw qt is fecked, it takes pen-down events for button-activate, instead of the worldwide usual pen-up01:07
DocScrutinizer05I noticed that in calculator01:07
DocScrutinizer05actually the stnadard correct GUi behaviour is to button-press *only* on pen-down followed by pen-up in same widget01:10
freemangordonoh, the fuck, got it. h-d reacts on release event :(01:11
DocScrutinizer05tadaa we're on topic01:11
DocScrutinizer05;-P01:11
freemangordonyep01:11
freemangordonso I should play it smart there :D01:11
DocScrutinizer05you can neither slide out of a button nor slide in, to press it01:12
DocScrutinizer05you need to press and release on same button to make button-push01:12
DocScrutinizer05except for *very* special cases like stopwatch (which usually fails to do so ;-P)01:13
DocScrutinizer05instead, fekkin calculator does01:14
*** wmarone has quit IRC01:59
*** wmarone has joined #maemo-ssu02:00
amiconnDoes anyone actually use slide-to-unlock?02:09
* amiconn only learned that this even exists a year or so after he got his N90002:09
freemangordonamiconn: me02:10
freemangordonit is useful for one-hand operation02:10
freemangordonesp in portrait mode02:10
*** kolp has quit IRC02:15
*** Martix has quit IRC02:22
*** Martix has joined #maemo-ssu02:23
*** jonwil has joined #maemo-ssu02:46
Palifreemangordon, tklock working with autorotation02:58
Palibut I have only background for beta theme02:58
PaliI pushed changes to gitorious03:09
freemangordonPali: damn, I can't push latest commits now :D03:19
freemangordonphew, pushed :D03:20
freemangordonPali: pull from gitorious03:20
freemangordonPali: I am not sure rotation in tklock is a good idea. actually it is bad03:21
freemangordonfor various reasons03:21
freemangordonbut i'll try it03:21
freemangordonPali: did you try it with forced rotation enabled?03:23
Palino, without forced rotation03:23
freemangordonhmm, nothing rotates here03:25
DocScrutinizer05autorotation on lockscreen sounds silly03:26
freemangordon:nod:03:26
freemangordonPali: what is supposed to rotate?03:26
Palifreemangordon, full screen03:27
Palibut only for beta theme03:27
freemangordonit does not03:27
freemangordonaah, which one is that?03:27
freemangordonorange?03:27
freemangordondigital nature?03:28
Palitry digital nature03:28
freemangordonok03:28
freemangordonok, rotates03:29
Palia bad, that should be alpha theme, not beta03:29
Palineed to change Makefile and background image in ./share/03:29
freemangordonso, it rotates only if theme has portrait background?03:30
freemangordonPali: ^^^03:30
Paliyes03:30
freemangordonhmm, well, ok03:31
freemangordonwe can provide backgrounds for alpha and beta03:31
freemangordonPali: though I'd rather disable autorotation03:32
freemangordoni.e. if started in landscape->show landscape lock, the same for portrait03:32
freemangordonbut do not rotate03:32
freemangordonno idea if that is possible03:33
freemangordonPali: provide a gconf key to disable that please :)03:33
Paliok03:34
freemangordonbesides wrong background that doesn't look bad. did you test it with missed events icons?03:35
freemangordonlooks ok with icons too03:43
*** dhbiker_ has quit IRC03:53
*** Pali has quit IRC03:54
*** dhbiker has joined #maemo-ssu03:55
*** arcean has quit IRC04:00
*** dhbiker has quit IRC04:10
*** dhbiker has joined #maemo-ssu04:12
*** M4rtinK has quit IRC04:17
*** nox- has quit IRC05:04
*** amiconn_ has joined #maemo-ssu05:44
*** amiconn has quit IRC05:44
*** amiconn_ is now known as amiconn05:44
*** DocScrutinizer05 has quit IRC06:03
*** DocScrutinizer05 has joined #maemo-ssu06:03
*** Milhouse has joined #maemo-ssu06:07
*** Milhouse has quit IRC07:07
*** Milhouse has joined #maemo-ssu07:08
*** Milhouse has quit IRC07:23
*** M13 has joined #maemo-ssu07:35
*** joshgillies has quit IRC08:56
*** jonwil_ has joined #maemo-ssu09:26
*** jonwil has quit IRC09:27
*** jonwil_ is now known as jonwil09:27
*** Pali has joined #maemo-ssu09:49
freemangordonPali: the more I think, the more I dislike the idea of having tklock autorotate functionality :) . For some reason it just does not feel "right". try the following: lock while in landskape, rotate to portrait, press the power button10:18
freemangordonthe same transition happens if the device was locked in portrait and power button is pressed in landscape10:20
freemangordonthere is a fair amount of time you just can't unlock as transition takes place10:20
keriodon't do the transition then10:21
keriojust draw the new widgets in portrait instantly10:21
kerio(and keep the slider in the middle of the screen as if it's in landscape)10:21
freemangordonkerio: what is the poit of doing that if the device in in landscape?10:21
freemangordon*is in10:22
keriono, i mean, only rotate the clock and the icons, leave the slider as it is, so if the rotation was botched you can still unlock10:22
freemangordonkerio: current behaviour (landscape only) is useful even if you hold the device one-handed in portrait. you can still unlock and/or read the clock10:24
keriothat's what i usually do10:24
keriowell, not really, i just rotate it in my hand, hold it against my chest and use the lockslider10:24
keriowhere's double-power-to-unlock dammit10:24
freemangordonwell, because most of the time i use visual unlock, I can assure you the portrait mode/autorotation breaks my user experience, at least :)10:25
freemangordonkerio: will come. some day :P10:25
keriothe portrait/landscape autorotation breaks my user experience a lot more often10:25
freemangordonkerio: in tklock?10:26
kerioi reckon it's only useful in OMP and possibly Phone10:26
keriono, in general10:26
freemangordonkerio: that will be fixed by adding "lock to current orientation"10:26
freemangordon(someday)10:26
Palilock to current location is simple10:29
Palijust need to read configuration options (like this) from somehere...10:29
freemangordonPali: even worse: lock while there is a landscape-only app on top and wait tklock to rotate to portrait10:29
freemangordonnow you have to wait for 2 transitions10:29
kerioheh10:29
freemangordonPali: not to say there are terrible visual artifacts while rotating10:33
keriofreemangordon: when do you use portrait mode?10:34
freemangordonkerio: when needed :)10:37
kerioyeah but when do you need it?10:37
freemangordonreading a book for example10:37
keriohuh, really? :o10:38
freemangordonyep10:38
freemangordonopera is perfect10:38
kerioi would've guessed that landscape lets you read better10:38
freemangordonno10:39
tadzikportrait is useful if you have only one hand available (in the public transport for example)]10:39
freemangordonyep10:39
keriofor me, that's basically just Phone and OMP10:39
* freemangordon is afk10:39
PaliI hate busybox in maemo !!!!! it does not have utf8 support10:40
Paliand if I write some utf8 chars (like slovak á č ť ž) and I want delete it I must press backspace two times!10:41
Paliwhich generate negative space...10:42
Pali(becuase that utf8 chars are 2 bytes)10:42
Palikerio, do you know how to fix that?10:42
keriomove to england and change every non-ascii character in your name10:43
Paliit is fixed in upstream busybox?10:43
kerioPali: from ssh, it WORKSFORME (busybox-power)10:43
kerioand it doesn't work in busybox stock10:44
Paliit is really bad that we still not have utf8 support in terminal...10:44
kerioso... install busybox-power i suppose10:44
kerioor bash10:44
kerio*and bash10:44
Paliand I found another bug in maemo bash10:44
Paliwhen I press DELETE key on normal keyboard via ssh bash generate '~' char10:45
Palifunny that maemo busybox not doing it10:45
Palikerio, ok then I will report bug that on maemo it is not possible to use localized Xterm10:45
Paliand some upstream busybox is needed10:45
Palifor cssu10:45
kerioPali: still haven't checked on the real terminal though10:46
Palissh busybox doing it too10:46
Paliso problem is in busybox10:46
Pali(not in osso-xterm)10:46
*** Pali has quit IRC10:59
*** Pali has joined #maemo-ssu11:13
*** Pali has quit IRC11:31
*** amiconn_ has joined #maemo-ssu11:34
*** amiconn has quit IRC11:34
*** amiconn_ is now known as amiconn11:34
*** dhbiker has quit IRC11:42
*** Pali has joined #maemo-ssu11:44
*** dhbiker has joined #maemo-ssu11:50
*** joshgillies has joined #maemo-ssu12:37
*** M13 has quit IRC12:40
*** Martix has quit IRC12:51
*** LaoLang_cool has joined #maemo-ssu12:57
*** M4rtinK has joined #maemo-ssu13:05
*** LaoLang_cool has quit IRC13:07
*** arcean has joined #maemo-ssu13:12
*** kolp has joined #maemo-ssu13:35
*** anthony__ has joined #maemo-ssu13:35
*** lizardo has joined #maemo-ssu13:36
*** dhbiker has quit IRC13:58
*** arcean_ has joined #maemo-ssu14:16
*** arcean has quit IRC14:18
*** joshgillies has quit IRC14:25
Palikerio, I updated hald-addon-bme, package is here: https://atrey.karlin.mff.cuni.cz/~pali/rx51-bme-replacement/14:32
Palinow it report both active and idle time to empty14:32
Paliand updated version of status area battery plugin is in cssu-devel14:42
Palinow applet show also idle time to empty14:43
Palimerlin1991, need to testing if ^^^ it working14:43
Paliand if yes then it is candidate for cssu-testing14:43
*** ruskie has quit IRC14:44
merlin1991Pali: put it on tmo?14:45
Paliok14:53
Palideb package is here: http://maemo.merlin1991.at/cssu/community-devel/pool/free/s/status-area-applet-battery/status-area-applet-battery_1.0-2_armel.deb14:54
Palimerlin1991, do you know if clock-ui on gitoruous cssu working?14:55
merlin1991it is semi working, it has no packaging yet14:56
*** ruskie has joined #maemo-ssu14:58
*** luf has joined #maemo-ssu15:32
luffreemangordon: are you crazy? location->fixed_title = g_strdup (dgettext("osso-connectivity-ui", "conn_ti_bluetooth_cpa"));15:42
luffreemangordon: I think it's not good patch (also mixing code change + some white space rework).15:53
luffreemangordon: I'll update the -l10n packages (it's better than this patch).15:54
*** Pali has quit IRC15:57
*** Pali has joined #maemo-ssu16:48
*** jonwil has quit IRC17:40
*** LinuxCode has joined #maemo-ssu17:41
freemangordonluf: what is wrong with that patch? (excluding white-spaces)18:04
freemangordonluf: did you look at the original _()18:04
freemangordonluf: it makes no sense to duplicate localisation strings, it is just wasting of RAM18:06
freemangordonluf: check here https://gitorious.org/community-ssu/libhildonfm/blobs/master/hildon-fm/hildon-file-common-private.h#line4018:11
freemangordonI really don'18:11
freemangordont see why adding one more localisation domain should be a problem18:12
freemangordonPali: do you mind to make branch (next or portrait-support) of tklock and to revert your portrait changes in master18:16
DocScrutinizer05Pali: DEL key is an age old problem in *every* shell, it's not a bug but basically PEBKAC18:17
freemangordonPali: I wan't tklock in next testing ;)18:17
Palialso HOME key is problem18:17
freemangordonthe point of REing it was that huuge memleak18:17
DocScrutinizer05rtfm ;-)18:17
DocScrutinizer05keybindings probably is the key18:18
Palifreemangordon, ok I will move portrait support to separate branch18:18
freemangordonPali: thanks a lot18:18
DocScrutinizer05freemangordon: we should publish memleak issue in tklock to a wider public, explain the implications (swap fragmentation etc), and announce CSSU is going to fix it18:20
DocScrutinizer05I pretty much would like you doing that, since it's been your finding18:21
freemangordonDocScrutinizer05: I think i've already done that18:21
DocScrutinizer05ooh then I missed that18:21
freemangordonwell, it is on alarm-ui thread, but still http://talk.maemo.org/showpost.php?p=1295060&postcount=6118:23
freemangordonand it was widely discussed on #maemo18:23
freemangordonalso I think I made a post when I put it in cssu-devel, lemme try to find it18:24
freemangordonhmm, well, no, there is just an announce that tklock is in -devel, but no reason why it is there18:27
DocScrutinizer05I'd like to see cssu-devel not getting referred to in posts to general public, too many trigger-happy lusers out there18:27
freemangordonyeah18:27
freemangordonDocScrutinizer05: BTW how is that different to powerkeymenu?18:27
DocScrutinizer05hm?18:27
freemangordonI mean it gous through the same release cycle18:28
freemangordon*goes18:28
freemangordonthe fuck18:28
*** Pali has quit IRC18:28
DocScrutinizer05so what's the question?18:28
freemangordonreason for powerkeymenu replacement was annonced on alarm-ui thread too (i.e. memleak)18:29
DocScrutinizer05missed that too18:29
freemangordonit was sitting in -devel, bugs got smashed and now it is in -testing18:29
freemangordonwhy tklock should be any different18:29
freemangordon?18:29
DocScrutinizer05the reason for my request been to convince more users to get CSSU18:29
freemangordonhmm, I wan't 30% raise then. For my extra work as PR :P18:30
DocScrutinizer05thus profiting from your awesome work18:30
freemangordonDocScrutinizer05: I will start promoting CSSU when -stable catches with -testing18:31
freemangordonIf you get what I mean18:31
DocScrutinizer05swap fragmentation due to (assumed) memleaks been THE PITA of maemo fremantle since beginning18:31
freemangordondeffinitely18:31
*** Milhouse has joined #maemo-ssu18:32
DocScrutinizer05and I always wondered why my device works great after 15, 30, 60 days uptime18:32
freemangordonBut I am afraid now to actively promote CSSU, as -stable is waaay behind18:32
DocScrutinizer05simple: I never use powermenu or tklock18:33
luffreemangordon: don't you see codedump for browserd when you're using gnomevfs-extra?18:33
freemangordonDocScrutinizer05: And I don't want to actively promote -testing, esp in the light of recent h-d and obexd problems18:33
freemangordonluf: hmm, didnt checked18:33
freemangordonlemme see18:34
DocScrutinizer05freemangordon: so please help on geting things to pace regarding all that18:34
DocScrutinizer05my plate is recently filled with other stuff :-/18:34
freemangordonDocScrutinizer05: why do you think mine is not? :)18:34
DocScrutinizer05I'm sure it is, but you should think about others who want to benefit from your work as well18:35
freemangordonluf: any instructions I need to follow18:35
DocScrutinizer05freemangordon: and you're talking to merlin1991 and chem|st in here more often than I do, lately18:36
lufReboot phono start browsing obexFTP using filemanger.18:37
freemangordonDocScrutinizer05: lets put it that way - I wait for HiFo to become in charge of the servers. Then we'll find a way to "sneak" a package to every n900, informing its user for CSSU existence18:37
DocScrutinizer05while I'm way more disconnected form CSSU recently than I'd like to be, due to friggin council, migration, whatnot else18:37
freemangordonluf: do I really need to reboot18:37
freemangordon?18:38
lufbrowserd usually crashed for me just first time I enter the bluetooth device.18:38
freemangordonluf: ok. will do it18:38
lufNo idea. I ever reboot when uploaded new libhildofm18:38
freemangordonluf: aah, men, comeon, ofc you'll have segfault18:38
lufLet's share what I need to kill to use new libhildonfm library.18:38
freemangordonthe library is either mmapped, or evicted18:39
freemangordonnext time a page is needed, it is read from the new image, thus the segfault18:39
lufSo it needs to recompile also browserd?18:40
freemangordonluf: no, you got me wrong18:40
lufI freshly reboot device.18:40
freemangordon<luf> No idea. I ever reboot when uploaded new libhildofm18:40
freemangordonI am lost :D18:41
freemangordonluf: do I need to have a webpage opened?18:41
lufOk. I did dpkg -i <new libhildon>; powermenu -> shutdown; power on; start osso-filemanager -> browse Bluetooth -> my device => browserd crash (core dump). After that it's ok till next reboot.18:42
lufNo, it's browserd not browser.18:43
DocScrutinizer05freemangordon: S will always behind of T, and version number will get out of sync pretty much anyway. Hell they already did, there were from very beginning18:45
freemangordonluf: I did the same, no crash here18:45
freemangordonDocScrutinizer05: it is not about ersioning, you know?18:45
freemangordon*versioning18:45
DocScrutinizer05so pester chem|st to get your memleak fixes into a S build, tomorrow!18:46
freemangordon-stable just don't follow the release cycle we've agreed upon ;)18:46
lufbrowserd -s 1862 -n RTComMessagingServer18:46
freemangordonluf: elaborate please18:47
DocScrutinizer05freemangordon: last S is from xmas18:47
lufThis process crashed.18:47
luffreemangordon: ^^^18:47
freemangordonDocScrutinizer05: sure, but there are packages in -testing as old as 6 months, which are not in -stable18:48
luffrom: gdb /usr/sbin/browserd coredump18:48
DocScrutinizer05yes, like camera-ui18:48
freemangordonno18:48
DocScrutinizer05as elaborated above, the S versions won't follow T18:48
freemangordonsure18:49
DocScrutinizer05there will be different mix of new stuff in S than been in last version of T18:49
freemangordonfor so-called "replacement" packages18:49
luffreemangordon: bt from gdb is very helpful: #0  0x41a8a548 in raise () from /lib/libc.so.618:49
lufCannot access memory at address 0xbea506218:49
DocScrutinizer05if there's some pkg not going to S for 6 months, then usually there's a reasn for that18:49
DocScrutinizer05if there's not, please holler18:50
freemangordonluf: you need debug symbols of browserd. Which are nowhere to be found :D:D:D18:50
freemangordonas browserd is closed source18:50
lufStrange it's only once per boot :( I have no idea where to find it. And it's related to osso-filemanager (access BT device -> gnomevfs).18:50
freemangordonluf: I don;t have such crash here. On my primary device18:50
*** arcean_ is now known as arcean18:51
luffreemangordon: You're brave man to use it on primary device :D18:51
freemangordonluf: I trust your code :P18:51
lufStrange also is that browserd loadk gvfs same as gnomevfs ...18:51
freemangordon(you should trust mine too :P )18:51
freemangordon:D:D:D18:51
luffreemangordon: I give you once very very good advise. Never ever trust my code :D18:52
freemangordonluf: well, I looked at the commit ;)18:52
freemangordonluf: I don't see browserd crashing here, it is something on your device or build system18:53
freemangordoni tried obexftp on 2 devices18:53
freemangordonprimary/develop18:54
lufOk I have to live with it then :D18:54
lufBTW do you also see wrong/default icons when browsing obexFTP? I think it's libhildonfm related :(18:54
freemangordonluf: though mine libhildonfm is 4.7.2 thumb-compiled18:54
luffreemangordon: hmm who knows.18:55
freemangordonluf: icons (where exist) are ok. but as I already told you , you should used BT dbus api to get both icons and names18:55
freemangordonnot gconf18:55
freemangordonit is not correct, for some reason18:56
luffreemangordon: no icons for devices. But for e.g. images in the devices ... First it shows picture but after some few amount of time it's changed for the default one.18:56
*** NIN101 has joined #maemo-ssu18:56
freemangordon2 of 4 devices supporting FTP are missing both name and icon in gconf18:56
freemangordonthe others are shown ok18:56
freemangordon(PC and PDA)18:56
freemangordonone of devices which is missing name and icon is n95018:57
lufHowever I hate the "directory" preloading. When you click on Bluetooth it starts connecting all BT devices in the list.18:57
freemangordonluf: this is the other problem I saw18:57
lufBut I'm unable to disable such behaviour :(18:57
freemangordonget rid of all the caching18:57
freemangordon(if possinle)18:58
freemangordon*possible18:58
freemangordonas there is no way to "refresh"18:58
luffreemangordon: there is. cache is cleared after some amount of time.18:58
freemangordonthat does not look battery-friendly :(18:59
freemangordonluf: when is a device re-browsed?18:59
freemangordonis it on some constant timeout?18:59
lufNo when no activity of browsing. I don't know it exactly as I think it's not BT based.19:00
freemangordonaah, maybe in obexgw code19:00
lufI think in gnomevfs code.19:01
freemangordonluf: is it possible to force the upper layers to refresh? from obexftp code?19:01
freemangordonI bet it is19:01
lufcaches are good. The only problem is prereading directories. I'm unable to disable it for top level directory :(19:01
freemangordonand more important - does obex gets notified if remote directory has changed?19:02
lufI don't understand how it is done :(19:02
luffreemangordon: no idea there are some monitors but I don't know if it's for remote side changes ...19:03
freemangordonluf: well, lets first fix the missing names/icons19:03
lufIt's no problem for me :D I have all devices in gconf :D19:04
freemangordonand please, "like" my patch. obexftp won;t be the first library in fremantle to use localisation strings from unrelated domains19:04
freemangordonactually it is like spaghetti in that regard19:05
luffreemangordon: I don't like but ok. I stop the work on localization patching.19:05
freemangordonluf: we already have those translations in the system. no pint to duplicate it just for the sake of it19:05
freemangordon*point19:06
freemangordonIMO19:06
luffreemangordon: as I wrote. I don't like but I accept it.19:06
freemangordonluf: remember, 256MB of RAM should suffuce for everything :D:D:D19:06
freemangordonok19:06
lufCome on I started linux on 16 or 32 MB of memory ... one FD was enough ...19:07
lufMaybe it was even less ...19:07
freemangordonhehe. I wrote my first x86 assembly code on a machine with 512KB of RAM. not even 640 ;)19:08
freemangordonno HDD, nothing, only FD19:08
lufYour first was x86? :D Not mine ...19:08
freemangordonfor ASM yes19:08
lufI wrote I first used linux ... not x86.19:09
freemangordonwell, I did a little on 650219:09
freemangordonluf: those are CPUs, not oses19:09
freemangordon;)19:09
lufSharp MZ800 - my first true love :D19:09
freemangordonWTF is that? Z80?19:10
lufIt was 8-bit computer from Sharp ;)19:10
freemangordonaah19:10
lufhttp://www.old-computers.com/museum/computer.asp?c=208&st=119:11
freemangordonwell, here we had 8bit PCs, stolen from Apple II. i.e. Apple II REed :D19:11
freemangordonand made with russian TTLs19:12
lufAfter few year we connected FD to that machine. How incredibly fast loading after that tweak :D19:12
freemangordontweak? :D:D:D19:12
luf:D19:12
freemangordonwhat from was it loading before that, punchcards?19:13
luf#define OBEX_TIMEOUT 2000019:13
freemangordonluf: hmm, not that bad19:13
luftapes (same as music tapes) ;) You can see it in the picture ...19:13
freemangordonthough... if we can interrupt instead of poll it will be perfect19:13
freemangordonaah, yes, tapes19:13
freemangordonnow I remember19:14
freemangordonthough I never used one19:14
freemangordon(for data storage)19:14
lufI don't understand "if we can interrupt instead of poll ..."19:14
freemangordonluf: aiui now obex is polling the remote devices every 20s, ain't?19:15
lufNo. It disconnected after such time of no activity.19:15
freemangordonaah, ok19:16
lufAnd invalidating caches are done during reset (connecting).19:16
freemangordonI see19:16
freemangordonhmm, not that bad19:16
lufom_connection_invalidate_cache19:17
freemangordonluf: keep it as it is for now, just fix the names/icons and lets put it in the wild19:17
freemangordonwill see how it will behave19:17
freemangordonluf: btw I am under the impression obexd is very slow. could be because my dongle is 1.1 ofc19:19
lufI need to disable preconnecting devices.19:19
freemangordonyes19:19
freemangordonyou are right19:19
freemangordonthat makes a lot of sense19:19
luffreemangordon: It's fast for me here. But depends on BT range ...19:19
lufHonestly I have no idea how to do that :(19:20
freemangordondon;t think it is a range issue, most probably BT1.1 is too slow19:20
freemangordonluf: will check if I can do something for you19:20
luffreemangordon: I tried fix some problem and I saw in logs: device not found several time (because too far). But it sometimes connects and sometimes fail.19:21
freemangordonluf: the distance between the dongle and n900 is about 50cm ;)19:21
lufIt should be close enough :D19:22
freemangordonyep19:22
lufI'm thinking about enabling connect after change directory to the device. However it should break some behaviour :(19:23
*** Pali has joined #maemo-ssu19:25
merlin1991freemangordon: what's the official name of thumb (1) and thumb2? I'm reading a spec sheet of an arm atm and trying to find out if they have thumb1 or thumb2 in the arm :D19:26
freemangordonafaik thumb1 is just thumb19:27
merlin1991Thumb High Code Density 16-bit Instruction Set .... which one xD19:29
Palifreemangordon, I have simple patch which enable autorotation in tklock only if was enabled by user via gconf19:29
freemangordonPali: ok, should do the job too. though I wonder if the backgrounds should be distributed as a part of tklock package19:31
freemangordonmerlin1991: check for ldrex/strex instructions19:31
Paliwhy not to distribute it?19:32
freemangordonif they exist, it is thumb219:32
freemangordonPali: not saying we should not, just wondering if a separate package would be better19:32
merlin1991hm nope they don't exist19:33
* merlin1991 headdesks19:33
PaliHAM has background also in HAM package19:33
Pali*portrait background19:33
freemangordonPali: hmm, yep, you are right19:33
Palibut we can also update theme package19:33
merlin1991$company is buying an arm chip with native a native java pipeline, thumb1 to run $generic linux on arm19:33
Paliand move also background from HAM to theme...19:33
* merlin1991 headdesks again19:33
freemangordonPali: that will be the best approach IMO19:34
Palibut theme packages are non free19:34
freemangordonmerlin1991: hmm that could be jazelle19:34
merlin1991yep it features jazelle19:35
freemangordonPali: that should not be a probem afaik19:35
merlin1991thrown away money if the chip in the targeted appliance is going to run arm only19:35
freemangordonmerlin1991: aah, this is a different kind of beer ;)19:35
merlin1991hence me headdesking19:35
PaliI have shell script which convert binary deb package to source dsc19:35
Paliand theme is ideal for that (no compilation needed, only copy files)19:36
freemangordonmerlin1991: thumb1 is baaad, baaad idea performance wise AFAIK19:36
freemangordonPali: sure, lets ask merlin1991 about it :P19:36
merlin1991so I want to run linux on arm, so I'll get a nice and shiny chip that has jazelle but no thumb2, and will run it in arm mode, doesn't that sound good?19:36
PaliAFAIK, jazelle is closed19:36
Palino public documentation19:36
freemangordonmerlin1991: sounds very clever to me :D19:37
freemangordonPali: yes19:37
freemangordonmerlin1991: you could use that to steal jazelle docs and to publish them :D19:37
merlin1991management: but it can run java (TM) developer: BUT WE RUN C CODE ANYWAY19:37
freemangordon:D:D:D19:37
freemangordonmerlin1991: no, it CANNOT run java19:38
freemangordonit executes some of the java opcodes in CPU19:38
freemangordonbut you should have handlers for not-supportd opcodes19:38
freemangordonmerlin1991: what about adding alpha and beta themes in CSSU?19:39
freemangordonand moving all the various portrait images to where they belong?19:39
merlin1991sounds okish?19:40
freemangordonPali: seems merlin1991 is fine with that19:40
freemangordonwe can provide portrait background for adeclock too19:41
freemangordonand for OMP19:41
freemangordon(at some point)19:41
PaliOMP using own design in portrait mode19:41
freemangordonis it? ok, no OMP then19:42
Palifreemangordon, I fixed memory leak in tklock19:46
Paliyou forgot to unref gconfclient19:46
freemangordonPali: are you sure it should be unreffed? ok, thanks19:47
Palihttp://developer.gnome.org/gconf/stable/gconf-gconf-client.html#gconf-client-get-default19:47
kerioPali: yay, i'll update ASAP19:47
freemangordonyep19:47
freemangordonthanks a lot19:47
freemangordonkerio: don;t it is still not in -devel19:47
freemangordonPali: push on gitorious, please, so i'll be able to build a new .deb and upload it to -devel19:48
kerioheh19:49
Palipushed19:49
keriofreemangordon: the bme replacement? :o19:49
keriomadness19:49
freemangordonkerio: no, tklock19:49
keriooh19:49
freemangordonPali: thanks19:49
Palikerio, battery status menu plugin is in cssu-devel19:49
kerioneat!19:49
Paliand updated hal addon bme is on posted link19:49
keriodoes it *need* the newer bme replacement?19:49
Paliyou can use stock applet or that new19:50
Paliold version has some problems, so I pushed fixed to cssu-devel19:50
PaliALL: test new battery status menu plugin from cssu-devel19:50
Paliit should work with bme, bme replacement or only with kernel modules19:51
freemangordonPali: will do. do I need anything else?19:51
Palialso it should handle stopping bme and loading kernel modules...19:51
Palifreemangordon, no19:51
kerioPali: how and when? :o19:51
freemangordoni.e. like hald-addon replacement?19:51
Palikerio, sudo stop bme19:51
freemangordonPali: ok19:51
Paliit should report 0 (instead bad values)19:52
freemangordonjust lemme upload tklock and i'll install it19:52
Palialso modprobe bq...19:52
Paliaplet should handle it and se values19:52
Palicheck if charger attached notification working19:52
kerioPali: your /etc/event.d/bme loads the bq modules though19:53
Palialso check if battery full/empty is correctly reported19:53
Palikerio, try also rmmod what happen19:53
Paliand then modprobe19:53
Paliaplet should still work19:53
Pali(without crash)19:54
keriobtw, does it still check dpkg versions on startup?19:54
keriobecause the delay caused by it *is* noticeable and annoying19:54
Paliyes19:54
kerioand it's replaceable by a check of existance of /usr/sbin/bme_RX-5119:55
Palinow there could be also delay with initializing dbus19:55
Paliit using dbus for idle time to empty19:55
Paliwhich sending hald-addon-bme19:55
freemangordonPali: do you mind me to remove portrait background PNG from the package?19:55
freemangordontklock that is19:56
Paliyou should see two time values in aplet first is avg and second is idle19:56
Palifreemangordon, I will delete portrait background when we have theme packages on gitorious and in cssu19:56
freemangordonok19:56
freemangordonwill leave it there for now19:56
freemangordonPali: hmm, debian/changelog not chaged19:57
freemangordon*changed19:57
PaliI did not changed it19:57
freemangordonok, i'll do19:57
PaliI pushed changes very quicky :-)19:57
freemangordonbut it will be FMG there :)19:58
freemangordonosso-systemui-tklock (0.3.1.0+0cssu0) is in cssu-devel20:04
freemangordonmerlin1991: WTF is "deb http://happens:now@maemo.merlin1991.at/cssu/future/ fremantle free non-free"20:07
kerioPali: your ssl cert expired20:07
keriofreemangordon: dammit, now you went and displayed the password to everyone20:08
merlin1991freemangordon: testing-testing20:08
Palikerio, what?20:08
freemangordonaah, ok :D20:08
kerioPali: https://atrey.karlin.mff.cuni.cz/20:08
keriooh, and it's also self-signed, wtf20:08
merlin1991kerio: mine is also self signed ;)20:09
freemangordonPali: what is package name for battery plugin?20:09
Palikerio, its server on our faculty20:09
keriofreemangordon: status-area-applet-battery20:09
kerioPali: i figured20:09
kerioPali: it's still a botched job, considering that there's a way to get a real SSL cert for free20:09
kerioPali: dammit, why don't you increase the versions when you release new packages? :c20:10
freemangordonPali: hmm, looks nice :D20:13
merlin1991kerio: ssl cert for free? o_O20:18
keriomerlin1991: http://www.startssl.com/20:19
kerioa class 1 cert, ofc20:19
keriosingle-host20:19
merlin1991isn't it a time limited offer?20:19
kerioapparently not20:19
keriomerlin1991: http://www.startssl.com/?app=25#9020:20
kerio(faq: why are class 1 certificates free?)20:20
kerio(answer: because the effort for those is JACK SHIT)20:21
*** Martix has joined #maemo-ssu20:29
freemangordonPali: on my devel device I have a time estimation for the battery life, on the pimary I don;t have it. I guess this is by design, but what is the idea?20:36
Palithat data are reported by BME and hald-addon-bme20:37
*** kolp has quit IRC20:37
freemangordonhmm?20:37
freemangordonboth are on BME20:38
Paliif you do not see time then BME reporting 020:38
freemangordonok20:38
merlin1991kerio: I pondered about getting a proper certificate for *.merlin1991.at but it's so friggin expensive20:39
DocScrutinizer05who cares about certs?20:45
*** MrPingu has joined #maemo-ssu20:45
DocScrutinizer05honestly a cert confirming that merlin1991.at is in fact merlin1991.at does help me how exactly?20:46
DocScrutinizer05I'm not expecting any virus injecting a spoofed IP for merlin1991.at into my etc/hosts20:47
keriohowever, you could expect a MITM20:47
DocScrutinizer05neither20:48
DocScrutinizer05what for?20:48
*** ivgalvez has joined #maemo-ssu20:48
Paliplease store cert hash to DNSSEC :-)20:48
DocScrutinizer05particularly i'm happy with accepting that self-assigned cert once20:49
Palicurrent SSL implementations are bad in all browsers...20:49
Paliwhy browsers show me warnings when I opening HTTPS pages with self-signed certificates? And why *not* for HTTP pages?20:50
DocScrutinizer05[2013-01-06 20:53:52] <wpwrak> DocScrutinizer05: that's bee on heise news a few days ago. certificate-based "trust" must die :)20:50
DocScrutinizer05(^^^ re TRUKTRUST)20:51
ivgalvezDocScrutinizer05 hi20:51
DocScrutinizer05hi ivgalvez :-D20:51
DocScrutinizer05welcome back!20:51
ivgalvezalmost fully recovered20:51
DocScrutinizer05hope you're well again!20:51
DocScrutinizer05good!20:51
ivgalvezstill digesting tons of emails20:52
freemangordonivgalvez: hi20:52
DocScrutinizer05I feel with you20:52
freemangordonglad you're back20:52
ivgalvezthanks20:52
ivgalvezI've been asking Nokia about those DNS problems mentioned by Nemein20:52
ivgalvezdo you know anything new about it?20:52
DocScrutinizer05I guess they are caused by negotiations between HiFo and Nokia not finished yet20:53
ivgalvezmaybe, the domain transfer won't be made until everything is finished with their legal department20:54
DocScrutinizer05Nokia not touching those DNS anymore, since they already *should* be under HiFo control20:54
ivgalvezyep, but not until everything signed20:54
DocScrutinizer05sure20:54
*** kolp has joined #maemo-ssu20:55
DocScrutinizer05btw full ACK to your "we should take what we can get now..." re those negotiations20:55
ivgalvezwe are gonna need volunteers to help on scratchbox.org maintenance and migration20:55
DocScrutinizer05*sigh*20:56
PaliDocScrutinizer05, when will HiFo migrate TMO?20:56
DocScrutinizer05we need so many volunteers, and have so few20:56
ivgalvezyep20:56
Palithere was some deadline in januar...20:56
DocScrutinizer05Pali: it's moving20:56
DocScrutinizer05ivgalvez: please read chanlog of last council meeting, last friday20:58
DocScrutinizer05ivgalvez: we got some misunderstandings sorted there, and basically sketched what to do with migration during next 2 months - not further than just the point where HiFo owns the servers20:59
ivgalvezdoing it now20:59
DocScrutinizer05ivgalvez: I'd very much like to ask HiFo to try and re-negotiate with Nokia if Nokia could re-dedicate the money for those RX300 to dedicated hosting maybe21:00
DocScrutinizer05we honestly can't handle those RX300, too expensive21:00
ivgalvezthat was previously asked and not just once21:00
DocScrutinizer05too "big" for us21:00
ivgalvezI can't even imagine why they decided to go that way21:01
lufmerlin1991: it's timeo for your favorite game ...21:01
merlin1991..?21:01
DocScrutinizer05if that's out of scope completely, then the basic idea now is to have nemein move al vservers on those two machines21:01
*** kolp has quit IRC21:01
DocScrutinizer05(that way) ... is how big business is doing that stuff21:02
lufmerlin1991: Jan, 9 2013 20:00 CET ... :D21:02
lufmerlin1991: I think you want to play ping-pong game ...21:02
ivgalvezDocScrutinizer05: yes, unfortunately the question was left out of discussion by Nokia some time ago, for some reason it would be easier for them to donate machines (tax refund?)21:02
DocScrutinizer05probably21:02
ivgalvezbut at least we will have room to move our stff21:03
DocScrutinizer05yes, we need to evaluate the specs for the RX300 again - Eero allowed us to have it re-specified to our needs21:03
DocScrutinizer05since they aren't ordered yet21:04
ivgalvezwhy is it told that the autobuilder will go to physical instead of VM?21:04
DocScrutinizer05it's not exactly simple, especially when we assume we can't pay CoLo for long anyway21:04
DocScrutinizer05mere performance considerations I guess21:05
DocScrutinizer05ivgalvez: for CoLo we'd need a hot (or at least cold) spare, otherwise maemo.org is doomed when the server(s) emit blue magic smoke21:06
DocScrutinizer05but all that is future plans, which pretty much are without any HiFo own hw in my book21:07
freemangordonmerlin1991: aren't we supposed to have a meeting?21:07
DocScrutinizer05considering we can get dedi server with almost same power for the fee we need to pay for rackspace alone21:08
ivgalvezDocScrutinizer05: but considering that there is no more activity on Maemo 4 and previous autobuilders and Fremantle autobuilder is not so crowded, it could go virtual, no matter if it takes a bit longer to build applications21:08
DocScrutinizer05yes21:08
DocScrutinizer05exactly21:08
ivgalvezDocScrutinizer05: we could sell servers on eBay and pay for hosting :D21:08
DocScrutinizer05that's the plan21:08
DocScrutinizer05alas it will burn value21:09
DocScrutinizer05but hey21:09
*** kolp has joined #maemo-ssu21:09
DocScrutinizer05not burning as much value as keeping those big irons will21:09
ivgalvezsure21:09
DocScrutinizer05so the both machines are ~7000EUR total21:09
ivgalvezlet's sell them21:10
luffreemangordon: did you remove Conflicts: lbxml2 from zlib? Or you just don't include it into thumb?21:14
luffreemangordon: there are several reports with old libxml2 and new zlib :(21:14
freemangordonluf: I wonder where did the get those from21:15
freemangordonno, Conflicts is there21:15
freemangordonalong with Depends in libxml221:15
luffreemangordon: hmmm, strange.21:16
lufmerlin1991: ping-pong meeting?21:16
freemangordonluf: BTW are you sure for the hour?21:17
keriowill HAM correctly handle Conflicts and/or Breaks?21:17
luffreemangordon: sure: Latest version: Testing(2013-01-07): 21.2011.38-1Tmaemo7.2; Stable(2012-12-24): 21.2011.38-1Smaemo5 | Meeting 9.1.2013 20:00 CET21:17
lufkerio: reportes used FAM or apt-get ...21:18
freemangordonmerlin1991: ping21:33
freemangordonmerlin1991: please, next time you appoint a meeting, setup an alarm on your n900 :(21:34
luffreemangordon: it seems he has another favorite game :D21:34
freemangordonyeah :D21:34
*** kolp has quit IRC21:37
luffreemangordon: I'm unable to find gnomevfs documentation :(21:37
*** kolp has joined #maemo-ssu21:37
freemangordonluf: what about libhildonfm2-doc_2.28.22+0cssu3_all.deb?21:38
luffreemangordon: and I see no way to don't device preconnection ...21:38
lufin osso-gnomevfs-extra21:38
freemangordonluf: not much we can do21:40
freemangordonI guess it reads only root dir21:40
luffreemangordon: it reads all sudirs ...21:41
freemangordonand caches them?21:41
freemangordonthe fuck, that is stupid21:41
lufSo when you see Bluetooth - it reads /*; when you see / it reads /<device>/*21:41
lufSo when you go to the Bluetooth it tries to connect to all devices in the list :(21:42
freemangordonI know that21:42
luffreemangordon: yes and cache them.21:42
freemangordonluf: leave that as it is now. lets finish the other stuf first21:43
lufI can maybe postpone it till cd into it. But it's againist ls (I'm affraid).21:43
luficon and name is quiet easy ... :)21:43
*** ivgalvez has quit IRC21:43
freemangordonwell, that should do the kob. against ls? what this means21:43
freemangordons/kob/job/21:44
infobotfreemangordon meant: well, that should do the job. against ls? what this means21:44
lufgnomevfs-ls21:44
*** dafox has joined #maemo-ssu21:44
freemangordonnever heard of it21:44
freemangordongnomevfs-ls obexd:///21:45
freemangordonError opening: Invalid URI21:45
lufIt's working for me ;)21:45
lufAre you trying it on N900?21:46
freemangordonyes21:46
*** kolp has quit IRC21:47
freemangordonluf: maybe playing with gnomevfs flags will to the job21:47
lufroot:N900dev# gnomevfs-ls obex:///21:47
luf[BT MAC1] [link: obex://[BT MAC1] ](Symbolic Link, x-directory/normal)size 0mode 044421:47
luf[BTMAC2] [link: obex://[BT MAC2] ](Symbolic Link, x-directory/normal)size 0mode 044421:47
freemangordonaah,wait21:47
freemangordonit is obex, not obexd :D21:47
kerioPali: applet+bme replacement: charger is detected, unplugging is detected, it should be charging fine now21:48
freemangordonluf: gnomevfs-ls connects on demand21:49
*** kolp has joined #maemo-ssu21:50
freemangordonso it os something in libhildonfm21:50
freemangordon*is21:50
freemangordonor in filemanager21:50
luffreemangordon: exactly21:54
freemangordonluf: maybe you should tweak om_chdir_to_uri21:55
freemangordonmore precisely om_utils_get_path_list_from_uri21:55
freemangordonto not recurse21:55
freemangordons/recurse/traverse/21:55
infobotfreemangordon meant: to not traverse21:56
*** kolp_ has joined #maemo-ssu21:56
freemangordonhmm, no, scratch that21:56
*** kolp has quit IRC21:56
freemangordonluf: BTW I don;t think child directories are traversed21:58
freemangordonit takes time to read a directory21:58
freemangordoni.e. to change to it21:59
*** kolp_ has quit IRC22:00
lufCall sequence is do_get_file_info (I disable connecting there), monitor_add, do_open_directory22:00
*** kolp has joined #maemo-ssu22:00
*** dhbiker has joined #maemo-ssu22:00
freemangordonlooks fine to me22:00
*** dhbiker has quit IRC22:00
luffreemangordon: It's not ok as it's for folder childs ...22:00
freemangordonmaybe open_directory should not read first level dirs. but it might be better that way22:01
lufWhen see Bluetooth the sequence is called for obex:///; when see obex:/// it's called for all devices.22:01
lufI see no correct way to disable this behaviour so I'm giving it up.22:02
lufmerlin1991: meeting?22:02
freemangordonluf: om_chdir_to_uri22:03
freemangordonfor (l = path_list; l; l = l->next) { ...22:04
freemangordonpath_list is build in om_utils_get_path_list_from_uri22:04
lufThere is no chdir called :(22:06
freemangordonthere is22:06
freemangordongw_obex_chdir22:06
lufBut om_chdir_to_uri isn't called.22:06
freemangordonhttp://pastebin.com/0ZtUDhcW22:06
freemangordonseems like calling gw_obex_chdir results in some monitor get called22:07
freemangordon;)22:07
freemangordonluf: comment for (l = path_list; l; l = l->next) { loop and check what will happen22:07
lufNo the problem is in filemanager or libhildonfm gnomevfs-ls don't do this.22:08
lufAnd of course in logs I see om_chdir_to_uri below bluetooth connection. So it's too late ;)22:08
freemangordonwell, try it22:09
luffreemangordon: please trust me it's loose of time. chdir is also called from gnomevfs-ls22:12
freemangordonok22:12
freemangordon22:08 <luf> freemangordon: please trust me22:13
freemangordon:D:D:D22:13
freemangordonluf: btw did you put function loggin everywhere?22:14
luffreemangordon: Not everywhere but in most functions ...22:19
freemangordonluf: maybe after all it is better that way.22:20
freemangordonas changing dirs happens instantly22:20
lufIt'll be better to not do it just in obex:///22:20
lufI'm curious how long it takes to you with 4 devices ...22:20
freemangordontakes long?22:21
freemangordonit is syinc22:21
freemangordon*async22:21
lufI'm not sure about the async ...22:21
freemangordondevices appear instantly22:22
freemangordonluf: now I am convinced it is better that way22:22
freemangordonas soon as you click on "bluetooth", you can select the device22:23
freemangordonand wait after chdir (of course)22:23
luffreemangordon: you're waiting to loop all devices ...22:24
freemangordonno22:24
freemangordonthere is no delay22:24
freemangordonstart filemanager->wait for BT to appear->click it->immediately click on a device22:25
lufHmmm, ok.22:28
freemangordondo you agree?22:29
luffreemangordon: I can't now. I don't have working osso-gnomevfs-extra :D22:29
freemangordonohh. git clean -df; git reset --hard HEAD22:30
freemangordonbackup your changes first :P22:30
luffreemangordon: you're right22:35
lufconfirmed22:35
freemangordon:)22:37
luffreemangordon: can you confirm also icon behaviour?22:37
freemangordonwhat do you mean?22:37
lufGo into Images folder. and feel free to scroll.22:37
lufI see picture icon and for some of them it's changed into icon with exclamation mark.22:37
freemangordonconfirmed22:38
freemangordonsome bad mime handling22:38
lufI think it's libhidlonfm behaviour :(22:38
freemangordon*mime type22:38
freemangordoncould be22:38
lufI have to dig into after change icon/name from gconf to dbus.22:38
freemangordonbut pdfs are shown ok22:38
freemangordontext files too22:39
freemangordonit is only pictures that are bad22:39
luffreemangordon: it depends on some timeout here ...22:39
freemangordonvides are ok too22:40
freemangordonhmm, seems my second disk is failing :(22:41
lufpackages are also ok ...22:41
lufcatch the disk! :D22:42
keriofreemangordon: buy moar SSDs22:42
keriomake a RAID101022:43
kerioan array of mirrors of array of mirrors22:43
*** freemangordon has quit IRC22:48
lufIt seems freemangordon also lost his primary disk :D22:52
DocScrutinizer51what's the cmd e.g. in IRSSI to list chan users?22:55
tadziktry /names maybe?22:56
DocScrutinizer51yeah22:56
DocScrutinizer51thanks22:56
tadziknp22:56
*** lizardo has quit IRC22:58
*** freemangordon has joined #maemo-ssu23:00
freemangordon"An error was detected on device \Device\Harddisk1\D during a paging operation." :(23:02
luffreemangordon: why you escape D and H :D :D23:02
luf* why do you23:02
*** anthony__ has quit IRC23:03
freemangordon:D:D:D23:13
*** dhbiker has joined #maemo-ssu23:18
freemangordonPali: no low battery warning here :)23:21
freemangordonbattery on 4% according to the applet23:21
*** NIN101 has quit IRC23:29
luffreemangordon: It seems I see bug in osso-gnomevfs-extra :D23:35
*** kolp has quit IRC23:38
*** kolp has joined #maemo-ssu23:40
luffreemangordon: fix pushed into git23:40
freemangordonhmm :D23:53
freemangordonluf: BTW when browsing BT2.0 (or whatever n950 is) obexftp is lightning fast23:54

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