IRC log of #maemo-ssu for Sunday, 2012-09-16

* kerio knows nothing about git00:01
*** dhbiker has quit IRC00:04
*** dhbiker has joined #maemo-ssu00:06
DocScrutinizer05lemme think about it, it's late on saturday00:10
DocScrutinizer05I'm not even sure right now if your statement is too terse for me00:10
DocScrutinizer05ooh, it's a A or B or C question, right?00:11
DocScrutinizer05meh, doesn't parse00:12
keriomerlin1991: it doesn't have to be the cssu-enabler package anyway00:12
kerioi mean, it's just repos00:12
kerioand i'd really like to not have to keep the enabler installed00:12
merlin1991kerio: we have to put the dependency on the repo somewhere00:13
keriono we don't00:13
merlin1991and the enabler makes most sense in that regard00:13
kerio...wait, what?00:13
keriothe enabler just installs them00:13
merlin1991why do you even want to uninstall the enabler?00:13
keriowhy do i have to keep it installed?00:14
merlin1991because it holds the domain info00:14
kerioalso, why do i have to keep busybox-symlinks-procps installed?00:14
keriomerlin1991: i don't care about the domain info, really00:14
keriomy HAM ignores that noise00:14
merlin1991but the domain has to work from installing cssu (which is add repo + install enabler) untill the millionth update00:16
merlin1991and without user interaction the enabler has to depend on the domain info in order to have ham not act up for the average user00:16
kerioidea: the .install installs community-ssu-enabler and the appropriate domain package00:17
merlin1991though since the enabler installs the mp after some point we could have the enabler depend on the domain package, the mp aswell which frees the enabler from the needed packges after mp got instlled00:18
merlin1991but on the other hand00:18
merlin1991the enabler contains the disabler aswell00:18
keriobtw00:24
kerioi hope you realize that downgrading to the nokia PR is something that will probably blow shit up00:25
merlin1991how?00:26
keriothere's bound to be something that goes horribly wrong00:26
kerioespecially when dealing with upstart/ke-recv00:26
merlin1991okay thumb is a special case, but everywhere else there's nothing changed that's going to go bad00:26
merlin1991the whole change to upstart is a 2liner that just adds a notice to /etc/fstab and ke-recv works in a backwards compatible fashion00:27
merlin1991the unly thing that "goes wrong" is the user experience due to missing features00:27
merlin1991s/unly/only/00:30
infobotmerlin1991 meant: the only thing that "goes wrong" is the user experience due to missing features00:30
*** NIN101 has quit IRC00:30
DocScrutinizer05as lil as I could follow, I think installing repos and enabler from same install script is it00:33
DocScrutinizer05cssu-s repo for Stable, cssu-t repo for Testing00:34
DocScrutinizer05and honestly, an unistalled that is neither documented not accessable, nor even does anything beyond what a reflash would do... (/me scratches head)00:35
DocScrutinizer05I'd rather prefer to tackle extensions to HAM to handle proper optional packages, finally00:36
merlin1991well it keeps the user data and settings at least compared to a reflash00:36
DocScrutinizer05errrrr00:36
DocScrutinizer05I once heard somebody mumbling a rootfs reflash will also retain user data00:37
*** andre__ has quit IRC00:37
merlin1991but not settings ;)00:38
DocScrutinizer05which settings?00:39
merlin1991everything that stores in the rootfs?00:39
merlin1991ie /etc/00:39
DocScrutinizer05which is what?00:39
merlin1991srly doc00:39
DocScrutinizer05led patterns?00:40
DocScrutinizer05transitions?00:40
merlin1991come on stop trolling00:40
DocScrutinizer05nah, i'm serious00:40
DocScrutinizer05I'll even open tickets for anything that not belongs to /etc00:41
DocScrutinizer05after all /etc is system wide, while $HOME is user data00:41
merlin1991then again maemo is single user00:42
DocScrutinizer05no it's definitely not00:42
DocScrutinizer05there's at very least user 2user" and user "root"00:42
DocScrutinizer05I heard there might even be devel00:42
DocScrutinizer05and the few locations with hardcoded "user" in pathnames etc are rather tickets than an argument not to check what's going on00:43
keriois maemo unix?00:44
DocScrutinizer05yes00:44
DocScrutinizer05definitely, *nix00:44
keriothen it's definetely multiuser00:44
DocScrutinizer05it evidently is00:44
DocScrutinizer05we maybe don't have a multiuser capable desktop00:45
keriocan't the ipad do that?00:45
DocScrutinizer05we definitely don't have a concurrent-multiuser desktop00:45
DocScrutinizer05otoh I think I heard madde can run apps on device and display is on PC, and that's not user "user", that's user "devel" or whatever00:46
DocScrutinizer05aiui00:46
DocScrutinizer05xephyr makes it00:47
DocScrutinizer05I'm stlll wondering what "userdata" or "settings data" might end up in /etc00:48
DocScrutinizer05NB not system installation/configuration data, like ssh keys and whatnot, no, *real* per-user data00:49
DocScrutinizer05LED patterns are arguable00:49
DocScrutinizer05in my book they should belong to user session and thus into $HOME00:50
DocScrutinizer05all modem stuff is system, usually, for most who comment on it, since there's only one modem on embedded devices, and it has only one SIM with one PIN00:51
DocScrutinizer05and one phone-number00:51
DocScrutinizer05evidence: PIN query dialog kicks in before even any H-D gets started00:52
DocScrutinizer05these are all highly intriguing architectural considerations00:53
DocScrutinizer05but I'd appreciate a lot one single instance of "user data" in /etc00:54
DocScrutinizer05since I just can't come up with anything right now00:54
DocScrutinizer05which was a shame if I'd tell BS to users asking for help regarding reflash00:55
*** andre__ has joined #maemo-ssu00:55
*** andre__ has joined #maemo-ssu00:55
DocScrutinizer05lemme put it simple: drwxr-xr-x 93 root root 10472 2012-09-15 20:10 /etc01:01
kerioDocScrutinizer05: well, i recently edited /etc/apt/preferences!01:06
DocScrutinizer05that's evidently not USER-data01:07
DocScrutinizer05that's clearly what I called "system config & installation"01:07
keriohrmpf01:07
DocScrutinizer05it's even *supposed* to get reset by reflashing, and also by "uninstalling" CSSU01:08
kerioheh01:08
DocScrutinizer05there's backup app for it, to re-install all apps and related system config01:09
DocScrutinizer05apps are not per-user, they are system global01:09
DocScrutinizer05also are system maintenance settings for stuff like apt etc01:09
kerioi suppose that the rootfs is to be considered reflashable without any effect on user data01:10
kerioDocScrutinizer05: procps is fukken huge01:10
DocScrutinizer05thought as much ;-P01:10
kerioalmost 1mb01:10
DocScrutinizer05not only huge but also greedy for CPU and other resources01:10
kerioon the other hand, busybox's ps sucks balls01:11
* kerio almost always uses pstree anyway01:11
DocScrutinizer05bb ps is outright useless01:11
DocScrutinizer05nevertheless i'm all for gnu-ifying and optifying procps01:13
* kerio seconds that01:13
keriootoh, procps is not GNU, is it?01:13
keriohahahahaha, ps' man page on os x lists this bug:01:15
kerioSince ps cannot run faster than the system and is run as any other scheduled process, the information it displays can never be exact.01:15
kerioDocScrutinizer05: are there any binaries in /usr, in procps?01:16
keriomeh, i'll just put them in /usr/bin/gnu anyway01:18
kerioaw, no, it has a bunch of crap in /usr01:20
kerioDocScrutinizer05: what do you think about the sysctl thing?01:22
kerioi'm not even sure that crap gets executed01:22
keriothere's no file in event.d for it01:22
keriohm, apparently it installs a symlink in /etc/rcS.d01:22
kerionot sure if that is executed01:22
DocScrutinizer05replace it by a dummy touching a flag file01:23
DocScrutinizer05you'll get to know, after next boot01:23
kerioanyway, as usual there's no reason to touch the system shell, so procps should be optified/gnuified01:24
DocScrutinizer05Xactly01:36
DocScrutinizer05ps aka procps might even one of the most suspicious cmds when we think about former(?) boot fsckup when replacing bb by proper shell with unixtools01:37
DocScrutinizer05eveb be* one...01:37
* DocScrutinizer05 sends a dsme -b to IroN90001:37
keriodsmetool -b, you mean01:38
DocScrutinizer05indeed01:39
DocScrutinizer05ok, still boots, despite procps01:41
keriohaha01:42
DocScrutinizer05now for gnu-ification at install time: is there any option to do that?01:42
kerioit's in the nokia repos!01:42
keriohm, what do you mean?01:43
keriowell01:43
DocScrutinizer05install to other --prefix dir, don't conflict with mp-fremantle-commuity-pr01:43
keriodpkg -u, move stuff around, change name and conflicts, dpkg -b01:43
DocScrutinizer05sounds good01:43
kerioi'm afraid apt doesn't work like that :)01:44
*** toxaris has joined #maemo-ssu01:52
*** tadzik has joined #maemo-ssu01:54
*** Vlad_on_the_road has quit IRC01:56
DocScrutinizer05I'm going to fsckng 'install' that shit with tar et al, instead dpkg/apt01:57
*** andre__ has quit IRC01:57
DocScrutinizer05or rather, simply copy away, uninstall, install mp-fremantle-community-pr01:58
merlin1991if you edit the .deb enough apt isn't going to complain at all01:58
*** pavi has quit IRC01:59
DocScrutinizer05that was my original plan :-)01:59
merlin1991what's actually the problem with procps02:00
DocScrutinizer05 but since it's already installed right now02:00
DocScrutinizer05...02:00
merlin1991where's the problem?02:00
DocScrutinizer05it conflicts with bb in mp-fremantle-community-pr02:00
merlin1991so the procps package has a conflict with busybox?02:01
DocScrutinizer05yep02:01
DocScrutinizer05wit busybox-symlinks-procps to be precise02:01
merlin1991hm so even if it is in extras people can't install it02:01
merlin1991(on stock maemo that is)02:02
DocScrutinizer05prolly not02:02
DocScrutinizer05it's in tools or whateber though02:02
merlin1991because the nokia mp depend on buybox-symlinks-procps aswell02:02
merlin1991so did you untar the fun or how did you du it now?02:03
DocScrutinizer05yeah, failwhale02:03
DocScrutinizer05apt-get insdtall procps02:03
merlin1991s/du /do /02:03
infobotmerlin1991 meant: so did you untar the fun or how did you do it now?02:03
merlin1991hm that removed the community mp then02:03
DocScrutinizer05yep02:03
merlin1991and what's your plan for now?02:03
DocScrutinizer05:shrug:02:03
merlin1991well without the community mp you won't have cssu updated via ham anymore02:04
DocScrutinizer05optify, uninstall, install mp-fremantle-community-pr02:04
DocScrutinizer05(update) tbh I'm not sure if that makes me any sad02:04
DocScrutinizer05right now I feel like cherrypicking02:05
merlin1991hm I'd rather have the "optified" binaries attached to a package so I can remove them again easily, but that's a convenience thing02:05
*** toxaris has quit IRC02:35
DocScrutinizer05sure, I'm almost temped to patch them for gnu-* and compile and package. but only almost02:50
merlin1991repacking that is a 5 min job for me :D02:50
DocScrutinizer05then I look at du -hs /scratchbox and compare that 6.5GB with the original 4.7GB on old PC, and reconsider02:51
merlin1991why do you even want to patch / recomile it, is there anything missing in the procps package from tools?02:51
merlin1991because otherwise you could simply repack the deb archive02:51
*** arcean has quit IRC02:52
DocScrutinizer05no idea, just thought I'd need to do some work to get it gnu-ified (and of course optified after that)02:54
merlin1991if the binaries in the original package are already what you want, then you only have to repack02:56
freemangordonmerlin1991: http://talk.maemo.org/showpost.php?p=1267043&postcount=67503:06
freemangordonsounds strange as I did a copy/paste from -testing03:07
freemangordonthough I will double-chack tomowwor03:07
freemangordon*tomorrow03:07
*** ChanServ has quit IRC03:23
*** ChanServ has joined #maemo-ssu03:28
*** calvino.freenode.net sets mode: +o ChanServ03:28
*** nox- has quit IRC03:34
gregoakerio, freemangordon: for just changing the version of a binary package: deb-reversion04:24
merlin1991nice there's a tool for that :D04:24
*** gri has quit IRC04:33
*** gri has joined #maemo-ssu04:33
gregoamerlin1991: it's a simple shell script, but it's indeed convenient for some situations. and might have saved freemangordon some time today :)04:46
*** LaoLang_cool has joined #maemo-ssu05:33
*** amiconn_ has joined #maemo-ssu05:33
*** amiconn has quit IRC05:33
*** amiconn_ is now known as amiconn05:33
*** dhbiker has quit IRC05:34
*** LaoLang_cool has quit IRC05:55
*** dreamer has quit IRC06:29
*** DocScrutinizer05 has quit IRC06:38
*** DocScrutinizer05 has joined #maemo-ssu06:38
*** dhbiker has joined #maemo-ssu07:01
*** fredrinLap has quit IRC07:27
*** polemo has quit IRC08:10
*** chainsawbike has quit IRC08:54
*** chainsawbike has joined #maemo-ssu08:55
*** ChanServ has quit IRC09:47
*** SpacedOut has quit IRC09:47
*** DocScrutinizer05 has quit IRC09:48
*** Sicelo has quit IRC09:48
*** trumee has quit IRC09:48
*** DocScrutinizer05 has joined #maemo-ssu10:10
*** Sicelo has joined #maemo-ssu10:10
*** SpacedOut has joined #maemo-ssu10:11
*** trumee has joined #maemo-ssu10:11
keriogregoa: nah, fmg had an excuse to recompile everything with gcc 4.710:50
*** NIN101 has joined #maemo-ssu10:59
*** luf has joined #maemo-ssu12:15
*** toxaris has joined #maemo-ssu12:21
*** arcean has joined #maemo-ssu12:23
freemangordonanyone with MfE account here?12:45
keriofreemangordon: ohai, you're awake12:46
kerioyour script forgets a couple of packages12:46
freemangordonkerio: why my script?12:46
keriothat commandline to downgrade12:46
freemangordonmy script installs only packages to be downgraded12:46
kerioyeah, but you forgot a couple12:46
kerio:)12:46
kerioa dude was here yesterday, i made him check12:47
freemangordonit relies on -mp dependencies for the others12:47
freemangordonwhich packages?12:47
keriohttp://mg.pov.lt/maemo-ssu-irclog/%23maemo-ssu.2012-09-15.log.html#t2012-09-15T22:05:0112:47
keriolibnspr4, libnss3, libnss3-certs12:47
freemangordonhmm, lelle check12:48
freemangordon*lemme12:48
freemangordonyeah, missing. fuck12:49
* kerio lols12:50
kerioapt would've never forgotten about those poor packages!12:50
freemangordonmerlin1991: "policy-settings-rx51(>= 0.4.10.8+0m5)"12:55
keriohm, is it intended that qtquickcompat and libqt4-test are automatically installed?12:57
freemangordonkerio: yes12:57
keriodo we have an actual need for qtquickcompat?12:59
freemangordonit is dummy12:59
*** ChanServ has joined #maemo-ssu13:08
*** calvino.freenode.net sets mode: +o ChanServ13:08
*** NIN101 has quit IRC13:09
*** NIN101 has joined #maemo-ssu13:15
keriodon't call me dummy!13:15
keriofreemangordon: btw, do you still have exact version dependencies in the metapackage?13:16
kerioneat, you don't :313:17
freemangordonkerio: i've never had any exact dependencies for anything but kernel and microb13:17
kerio...microb? why?13:17
freemangordonbecause we had screwed the versions13:18
freemangordon(me that is)13:18
keriooic13:18
freemangordonand I wanted to be sure wht is installed13:18
freemangordonhmm, strange reports about modes, WTF?13:19
freemangordon*modest13:19
kerioidk, i don't use modest or mfe13:19
*** Pali has joined #maemo-ssu13:26
lufpali: do you know about http://talk.maemo.org/showpost.php?p=1266655&postcount=33 ?13:38
freemangordonluf: that is supposed to be fixed13:39
freemangordonin kp50+13:39
freemangordonfixed means "same as stock kernel" in that context13:39
PaliI have Creative WP-350 (HFP+A2DP) and working fine with kp5113:40
freemangordonPali: it is bt_mice patch, do you remember in which version is fixed one included?13:40
Palifixed in v5013:42
freemangordonhmm, the guy says it is broken in 49 and 50, so it is broken in stock kernel too13:43
freemangordon(if anything is broken :) )13:43
lufI asked him if it works with stock kernel ...13:44
freemangordonand?13:44
lufFew minutes ago ;)13:44
freemangordonaah, ok :)13:44
freemangordonhmm, "because my last bt headset (se mw600) randomly disconnects under kp v50+ and that's why i am switching."13:45
freemangordonthere is where we reverted to stock13:46
freemangordonso I expect it is broken in stock too13:46
freemangordon*with13:46
freemangordonPali: maybe we should do some sysfs entry or whitelist to deal with that13:47
freemangordonwhitelist sounds better to me13:48
Paliok13:48
*** toxaris has quit IRC14:13
*** dhbiker has quit IRC14:16
freemangordonthe fuck, modest is missing a couple of build dependencies :(14:20
keriohm, to be fair the range on my LG HBS-700 sometimes disconnect for no apparent reason14:55
kerioand the range is really bad14:55
keriobut it could be the headphones14:55
kerioi can't exactly roll back to the stock kernel though :)14:59
merlin1991freemangordon: what was wrong with modest?14:59
keriomerlin1991: a couple of people complained about errors with mfe15:00
kerioor something like that15:00
freemangordonmissing build dependencies plus default control file15:00
freemangordonthat is for start15:00
merlin1991yeah modest and tinymail are slightly foobar with their control and rules! file that you have to move around manually15:01
freemangordonbut the problem was that THERE IS default control15:01
freemangordoni renamed it to control.linux :D15:01
freemangordonwill putt that in master later on15:02
freemangordonand there was a missing build dep to hildon-control-panel-dev15:02
lufkerio: have you tried to debug it? It should be bluez bug ...15:03
freemangordonWTF?15:04
freemangordonSep 16 15:04:41 Nokia-N900 modest[1941]: GLIB WARNING ** default - /usr/lib/modest/mail-plugins/libasmodestplugin.so: undefined symbol: modest_mail_operation_get_source15:04
kerioluf: nah, it's probably a range problem15:04
lufpali, free: sirpaul answer that the stock kernel has the same issu.15:04
freemangordonyeah, I knew it :(15:04
keriophew15:05
luffreemangordon: where can I take a look on the patch source?15:08
freemangordonhttps://garage.maemo.org/plugins/ggit/browse.php/?p=kernel-power;a=history;f=kernel-power-2.6.28/debian/patches/bt-mice.diff;h=5a19c40668110e9719b0f05954951ed2f5a7fb7c;hb=fab6e2bc66c231e4d4866d2939844fe1eeffc94d15:11
freemangordonluf: ^^^15:11
*** Gusse has joined #maemo-ssu15:29
luffreemangordon: The patch looks strange comparing bluez kernel part. Where did you get it?15:33
freemangordonget it? nowhere from15:33
freemangordoni did it15:33
freemangordonbut it works only for HID device15:33
lufAha :)15:33
freemangordonthere is a check as you may see, so it does not touch handsfree devices15:34
GusseHi freemangordon. Tested your latest modest version (modest_3.90.7-11.2+thumb2_armel.deb), but still no MfE working.15:34
freemangordonGusse: yeah, I know15:34
freemangordonfor some strange reason it does not load plugins15:34
freemangordoninvestigating at the moment15:34
GusseTook same time to get here. Something like ~15 years since I use IRC laste time :)15:35
freemangordonhehe15:35
GusseIs there anything that I can do or have you idea what is wrong?15:36
kerioGusse: to be fair, you deserve it for using modest15:36
freemangordonkerio: shut up15:36
freemangordonGusse: i am debugging right now, just hang here15:37
GusseNo problem15:37
keriofreemangordon: i jest, i jest15:37
*** NIN101 has quit IRC16:00
keriofreemangordon: btw, gregoa said that deb-reversion is a tiny script to change the version of a binary package in a simple way - you could use it for xorg (and maybe modest, if the old one worked)16:02
*** dhbiker has joined #maemo-ssu16:07
lufStrange. Some user reported error uploading files over bluetooth. He was able to do it using CSSU-S 3.0 and obexd 0.46. After upgrading to CSSU-S 4.1 obexd 0.46 doesn't work. Any idea? CSSU-S 4.1 + obexd 0.18 is working.16:07
*** toxaris has joined #maemo-ssu16:07
keriohuh, weird, i can't seem to send files via bluetooth to my computer16:10
keriook, it worked16:11
kerioi had to repair16:11
keriothat is, pair again16:11
lufkerio: from PC -> N90016:11
lufhttp://talk.maemo.org/showthread.php?t=59114&page=4116:12
keriophew, worksforme16:12
*** andre__ has joined #maemo-ssu16:12
lufWhat OS do you have on PC?16:13
kerioosx16:14
luf:) Ok.16:14
keriotry removing the pairing and remaking it16:14
keriofrom the n90016:14
lufHmmm. ok.16:14
*** dafox has joined #maemo-ssu16:25
lufkerio: pairing reset doesn't help.16:26
kerioluf: :(16:53
keriotried rebooting?16:54
* DocScrutinizer05 throws a few peanuts at kerio16:56
keriohm, apt wants to upgrade policy-settings-rx51 from 0.4.10.8+0m5 to 0.4.10.8+0m5, skipping 0.4.10.8+0m5 in the process16:59
keriothe fuck is going on?16:59
*** wmarone__ has quit IRC16:59
DocScrutinizer05o.O17:00
freemangordonkerio: yeah, saw that17:00
freemangordonbut isn't versions the same?17:00
kerioDocScrutinizer05: apt-cache policy gives me the same version three times17:00
*** wmarone__ has joined #maemo-ssu17:00
DocScrutinizer05lol17:00
keriohttp://fpaste.org/Z5aA/17:00
DocScrutinizer05I guess that's what you get with domains17:00
freemangordonGusse: you may want to upgrade modest now, it should be ok17:00
kerioDocScrutinizer05: yeah, i was worried about that17:01
kerio:s17:01
keriogoddammit pali17:01
kerioDocScrutinizer05: the thing is, it doesn't happen for other stuff17:01
keriofreemangordon: +thumb3, right?17:01
freemangordonyep17:01
keriohm17:01
kerioDocScrutinizer05: maybe it's a different thing, though17:02
kerioperhaps they're different17:02
keriowtf, they're three different sizes17:03
DocScrutinizer05LOL17:04
keriomeh, it happens with other packages too, but apt considers them the same version regardless17:04
kerio(stuff in cssu-testing that's the same as cssu-devel, but it was compiled on two different builders)17:04
DocScrutinizer05I guess apt is smart17:05
kerioyo Pali, are you around?17:05
Palihere17:05
DocScrutinizer05hi! :-)17:05
freemangordonkerio: i don't have -devel enabled17:05
keriohi!17:05
kerioPali: wtf is going on with policy-settings-rx51 and versions?17:06
Palicheck md5sum of versions17:06
keriomd5 of what?17:06
Palideb package17:06
DocScrutinizer05kerio: look at your own pastebin, it's obvious17:06
DocScrutinizer05that's what I meant with "smart"17:06
Paliif in devel, testing, ... has packages with same version string, but different md5sum, then apt will upgrade it17:06
kerioi see17:06
freemangordonkerio: yea, you installed it from -devel, now there is new one in -testing17:07
freemangordonsame here, so no problem at all17:07
DocScrutinizer05expected behaviour17:07
keriobut apt usually keeps those versions together17:07
kerioDocScrutinizer05: no it's not :c17:07
freemangordonkerio: it is17:07
keriohm, i need a package that's both in c-devel and c-testing17:08
freemangordoneven now hash in the same repo will cause apt to use what is in the repo17:08
Palikerio, and fpaste.org still not working Error 50017:08
kerioPali: why are you always the one with problems on fpaste?17:08
freemangordonGusse: ping17:08
DocScrutinizer05Pali: strange17:08
kerioPali: https://pastee.org/fnepg17:08
DocScrutinizer05Pali: still WFM17:08
keriohm, no, hold on17:09
Palikerio, this working :-)17:09
kerioPali: https://pastee.org/4srg2 better indentation17:09
DocScrutinizer05Pali: could you try a traceroute?17:09
kerioDocScrutinizer05: the domain stuff is unrelated to what apt does, it's only a ham thing17:10
DocScrutinizer05rrright17:10
Palithen in cssu-testing and devel must be differnet packages17:10
Palionly version string is same17:10
DocScrutinizer05[2012-09-16 16:03:48] <kerio> wtf, they're three different sizes17:11
kerioDocScrutinizer05: yeah but there are packages that are different, in c-testing and c-devel, but they have the same version string - and apt treats them the same17:11
DocScrutinizer05same md5sum17:12
kerioBinary files policy-settings-rx51_0.4.10.8+0m5_armel-devel/data/usr/share/policy/rules/rx51/policy.plc and policy-settings-rx51_0.4.10.8+0m5_armel-testing/data/usr/share/policy/rules/rx51/policy.plc differ17:13
kerioHRMPF17:13
keriowhat's the good one?17:13
kerio(everything else is the same17:13
DocScrutinizer05fun stuff the repo info files (whatever the name) probably have no md5sum of binary17:13
kerio)17:13
Paliwget fpaste.org working17:13
Palifpaste.org in rekonq not17:13
Palikerio, that plc file it compiled prolog code17:14
kerioPali: what's the md5 of your policy.plc?17:14
DocScrutinizer05Pali: use other browser17:14
merlin1991DocScrutinizer05: the packages file has md5sum and sha1sum of the whole deb17:14
Palikerio, why you need md5?17:14
DocScrutinizer05merlin1991: ooh17:15
kerioPali: to know what to install :)17:15
DocScrutinizer05then how can apt consider they are same17:15
Palikerio, prolog store for each source file full path and timestamp17:15
DocScrutinizer05when size differs17:15
DocScrutinizer05prolly kerio lied to us ;-P17:15
Paliso when you compile prolog file two times, it will be two time different17:15
keriomerlin1991: wait, so those packages that are the same in -testing and -devel actually compiled to the exact same files?17:15
kerioPali: oh, so they're actually the same? i see17:16
Paliyes17:16
Paliand no17:16
kerioheh17:16
kerioi meant they're functionally equivalent17:16
DocScrutinizer05[2012-09-16 16:11:39] <kerio> DocScrutinizer05: yeah but there are packages that are different, in c-testing and c-devel, but they have the same version string - and apt treats them the same17:16
DocScrutinizer05[citation needed]17:16
Paliyes, because it have all source predicates (same)17:16
kerioDocScrutinizer05: merlin1991 always said that those packages are probably different, because different compilers on different boxen17:16
Paliand no, because it has different additional predicates which return full path of file, timestamp of file, ...17:17
kerioPali: yeah, but developers that write stuff that depends on that should be shot :)17:17
DocScrutinizer05merlin1991: what would be proper URL to wget a package file?17:19
keriofreemangordon: is there some way i can test mfe works?17:19
DocScrutinizer05merlin1991: arbitrary one17:19
merlin1991DocScrutinizer05: depends on the repo sw, you usually go from the Releases file onwards because it has the main paths17:19
kerioDocScrutinizer05: sudo apt-get --print-uris install foo17:19
DocScrutinizer05thanks both17:19
merlin1991ie -thumb would have http://maemo.merlin1991.at/cssu/community-thumb/dists/fremantle/free/binary-armel/Packages17:20
kerioDocScrutinizer05: hrmpf, it doesn't work for stuff that's already installed :s17:21
merlin1991cssu-testing is http://repository.maemo.org/community-testing/dists/fremantle/free/binary-armel/Packages17:21
merlin1991kerio: add --reinstall and maybe -s ;)17:21
keriomerlin1991: nah, --print-uris is already a no-action17:21
DocScrutinizer05so it's always just the Packages file that holds *all* packages, and no further per-pkg info file?17:21
DocScrutinizer05outside the .deb itself17:22
merlin1991yes17:22
DocScrutinizer05:nod: and :-S17:22
*** NIN101 has joined #maemo-ssu17:22
DocScrutinizer05I guess often the Packages file is larger than the .deb17:22
merlin1991especially in extras-devel :D17:23
keriofor extras-devel, definetely17:23
DocScrutinizer05yeah17:23
keriobut repos on repository.maemo.org are stupid as hell17:23
keriothey keep *all* versions17:23
kerioand they have no pdiff support17:23
DocScrutinizer05well, cssu does as well afaik17:23
kerioDocScrutinizer05: and where are the community(-testing) repos? :)17:23
DocScrutinizer05I went W*T*F looking into mp-fremantle-community-pr17:23
merlin1991DocScrutinizer05: it's also on repository.mamo.org ;)17:24
DocScrutinizer05hehe, I see17:24
keriothus my request to merlin1991 to only apt-mirror the latest versions of packages17:25
kerioso we have a lightweight repo to use17:26
DocScrutinizer05so it seems I learnt enough already about apt, to have a walk outside when my morning coffe is over. apt-cache show xy gives me exactly the right segment for xy from Packages file17:27
kerioDocScrutinizer05: apt-cache show xy gives you the right segments for xy from all the Packages files17:27
keriotry doing that for hildon-desktop :)17:28
DocScrutinizer05:nod:  alarmclient has only one source ;-P17:28
DocScrutinizer05alarmd already much better17:29
kerioDocScrutinizer05: --no-all-versions17:30
DocScrutinizer05o.O17:30
* DocScrutinizer05 sips his coffe17:30
kerioi learn something new every day :D17:30
DocScrutinizer05FU apt, filename but no URL17:31
keriohahahaha17:32
DocScrutinizer05FU dpkg I guess17:32
DocScrutinizer05err nope17:32
DocScrutinizer05dpkg has no clue about repos and Packages file17:32
kerioanyway, i bet that removing the old versions would make HAM faster17:33
DocScrutinizer05you may bet on that17:33
DocScrutinizer05though freemangordon recently had a pretty sound point about gzip iirc17:34
keriohell, i don't know: is it possible to have the repo actually point to a different webserver for the uris?17:34
DocScrutinizer05dunno if it was only for apt-worker17:34
merlin1991well apt worker is going to be a billion times faster with less stuff to build dependency trees :D17:34
DocScrutinizer05yes, obvious17:35
DocScrutinizer05it's exponential17:35
DocScrutinizer05almost like n items take apt n^2 timeunits to process17:36
DocScrutinizer05might turn out to actually be n^317:36
kerioor n^4!17:37
kerioor n^5!17:37
kerioor n^2 logn17:37
merlin1991kerio relax you talked enough about it so I'm setting it up17:37
kerio^proof that being really annoying works17:38
DocScrutinizer05lol17:38
DocScrutinizer05will we still keep the "archive"?17:38
merlin1991just need to get my crontab right17:38
merlin1991DocScrutinizer05: sure17:38
DocScrutinizer05I.E. will I still find KP47 on merlin's repo?17:39
keriothat's apt-mirror, this would have to be a different system17:39
kerioi suggest maemo.merlin1991.at/light/<reponame>17:39
merlin1991DocScrutinizer05: I'm going to waste even more space on my box, the original archive stays in place17:39
DocScrutinizer05:-)))17:39
keriomerlin1991: you could just hardlink the files17:39
DocScrutinizer05hmm, seems I suggested that ;-P17:40
DocScrutinizer05[2012-09-14 18:29:41] <DocScrutinizer05> shouldn't bee too hard and too space consuming, when you use hardlinks?17:40
merlin1991the way apt-mirror and reprepro work I'd have really bad problems setting this up17:41
DocScrutinizer05possibly17:41
merlin1991(with hardlinks)17:41
merlin1991and since I don't have a tool filter the dists file I'll just waste some space :P17:41
keriomerlin1991: make reprepro make a lightweight repo out of your apt-mirror, and make it use "ln" as the downloader :D17:41
merlin1991rofl17:41
* DocScrutinizer05 wonders where went his awesome script that replaced all dup files to other folder in a folder hierarchy by hardlinks to other folder17:42
DocScrutinizer05oooh it was just one carefully crafted commandline, admittely a tad lenghty17:43
merlin1991hm, I possibly could even make it debs only17:43
keriowc -l opername.log =>   24  144 1152 opername.log17:44
DocScrutinizer05WUT?17:44
kerio24 lines17:44
DocScrutinizer05aah17:44
kerio:)17:44
* kerio symlinks /dev/null there17:45
DocScrutinizer05I gather all 24 are same?17:45
DocScrutinizer05uniq opername.log17:46
keriothe time is different though17:46
keriobut yeah, service provider name match17:46
DocScrutinizer05oooh timestamp, yeah X-P17:47
kerioabout 19 centuries ago, the service provider name matched17:47
DocScrutinizer05hrrhrrhrr17:47
* merlin1991 configured reprepro and crosses fingers17:48
DocScrutinizer05that's so embarrassing17:48
merlin1991eeew I have wrong checksums in my apt-mirror17:49
keriomerlin1991: to be fair we probably don't need the multiple versions in a repo structure17:49
kerio...although i suppose that it's historical data17:49
merlin1991technically repository.maemo.org should only contain up to 3 versions of a package17:49
merlin1991but the system is flawed17:49
DocScrutinizer05kerio: aiui we even got the.deb in original merlin repo17:49
DocScrutinizer05and I appreciate that much17:50
DocScrutinizer05e.g for getting KP4717:50
kerioDocScrutinizer05: i meant that we probably only need the debs, not the whole structure17:51
DocScrutinizer05a shame you never know what old version of which pkg you might need eventually, so keeping them all is the only nice alternative to discarding all old stuff17:51
merlin1991shit extras-devel reached 100G already17:52
DocScrutinizer05kerio: for apt-get install KP=47 we need the structure17:52
keriooh, k17:52
keriowhy do you need KP47?17:53
DocScrutinizer05why you don't?17:53
keriobecause i'm always on the bleeding edge B-)17:53
kerio(those are sunglasses, not boobs)17:53
merlin1991damn 4 files have missmatching checksums17:54
merlin1991ffs17:54
DocScrutinizer05better wear protective goggles ;-P17:54
DocScrutinizer05merlin1991: how come?17:54
merlin1991repo madness wget problems, whatnot?17:54
DocScrutinizer05:-S17:55
merlin1991but it's only 4 so I can go and check manually17:55
DocScrutinizer05:nod:17:55
DocScrutinizer05or simply rm ?17:55
DocScrutinizer05if those are shit pkgs17:55
merlin1991I'd ruin my apt-mirror db that way17:55
DocScrutinizer05hmm17:56
DocScrutinizer05k17:56
DocScrutinizer05nfc bout that17:56
DocScrutinizer05I should get that friggin 50bucks 3TB server4you stuff before I am too broke to dare17:56
*** kerio has left #maemo-ssu17:57
*** kerio has joined #maemo-ssu17:57
kerioshit17:57
merlin1991oh ffs the md5sum in the repo on maemo.org is wrong!17:57
kerioi should really find a way to enable stickychan without having it floodkill me when reconnecting17:57
keriomerlin1991: LOL17:57
DocScrutinizer05merlin1991: thought as much17:58
DocScrutinizer05kerio: stickychan? floodkill?17:58
merlin1991I just hope reprepro has flag to ignore packages with wrong checksums17:59
DocScrutinizer05merlin1991: well, how the fsck you created that existing mirror?17:59
kerioDocScrutinizer05: a znc plugin that tries really hard to forcibly keep you in the channels you give it17:59
DocScrutinizer05kerio: :nod:18:00
merlin1991DocScrutinizer05: apt-mirror is dumb, it only gets Realese and Packages file and after wgets18:00
merlin1991afaik it does no hashcheck18:00
DocScrutinizer05I already considered augmenting znc to accept a paramenter <timestamp> to replay command18:00
kerioDocScrutinizer05: but a side effect is that it makes you join *all* of your channels when you connect to the irc server18:00
DocScrutinizer05I know18:01
keriofuck yeah znc o/18:01
DocScrutinizer05the real problem isn't stickychan but playback/buffer18:02
*** macmaN has quit IRC18:03
DocScrutinizer05of course you want a big buffer, but you don't want it to playback all the friggin content since last new-year on reconnecting your client18:03
DocScrutinizer05so client should do manual playback <timestamp>, where <timestamp> is last line's timestamp in local log of that chan18:04
merlin1991hm I'll have to reach X-Fade18:04
*** macmaN has joined #maemo-ssu18:04
DocScrutinizer05ouch, he's 'not available'18:04
DocScrutinizer05are you still not master of nokia repos?18:04
DocScrutinizer05:-/18:04
keriofreemangordon: wtf, why is gcc 4.7 in community-thumb not thumbified?18:05
keriogcc4.7 and libstdc++18:05
merlin1991DocScrutinizer05: the amount of access I would get is by far not enough to fix a problem like this18:05
DocScrutinizer05be happy about that18:05
DocScrutinizer05I wouldn't want to take the responsibility18:05
keriomerlin1991: which packages are these?18:06
DocScrutinizer05indeed, and what's the supposed fix you're not able to apply as admin?18:07
merlin1991travelapp (non-free) and from free: libqvd-config-perl, python-pygoocanvas, qutim-dev and webapprunner18:08
merlin1991DocScrutinizer05: all the access I'll get would be delte / promote single packages18:09
merlin1991but this problem here either needs a reindex or some manual file moving18:09
DocScrutinizer05:-S18:09
DocScrutinizer05X-Fade: ping ^^^^18:09
*** macmaN has quit IRC18:10
keriomerlin1991: won't apt/ham complain about those packages, anyway?18:10
DocScrutinizer05kerio: as soon as you touch tem, prolly yes18:11
DocScrutinizer05them*18:11
kerioso... just ignore them18:11
keriothey're borked anyway18:11
DocScrutinizer05:nod:18:11
DocScrutinizer05hard to do in a script like reprepo18:11
DocScrutinizer05even harder if it's not a script18:12
*** macmaN has joined #maemo-ssu18:12
DocScrutinizer05apt-cache search webapprunner  --> nada18:13
kerioFailed to fetch http://repository.maemo.org/extras-devel/pool/fremantle/free/w/webapprunner/webapprunner_1.0-2_armel.deb  Size mismatch18:13
DocScrutinizer05apt-cache refuses to even find any of those packages18:14
DocScrutinizer05seems they are orphaned?18:15
DocScrutinizer05stale?18:15
DocScrutinizer05hmmm http://maemo.org/packages/view/python-pygoocanvas/18:16
*** myuu__ has joined #maemo-ssu18:16
DocScrutinizer05looks kinda strange18:17
DocScrutinizer05missing version18:17
myuu__got a random n900 question: whats the biggest single battery killer? anyone know18:17
DocScrutinizer05modem18:18
DocScrutinizer05by far18:18
freemangordonwhen on 3G18:18
myuu__yeah ever since i upgraded18:18
myuu__well its been on for damn near two days with no sim18:18
DocScrutinizer05and in large distance from BTS, and actually transmitting18:18
myuu__i love this phone but it wasnt useable18:18
myuu__going to clean it up and use it as a portable box though18:19
DocScrutinizer05hmm?18:19
freemangordonmyuu__: wasn't usable what for?18:19
DocScrutinizer05and why?18:19
freemangordonGusse: ping18:20
DocScrutinizer05you say your modem broke with last update?18:20
myuu__well after two years of use18:20
myuu__between the battery and the random crashes18:20
myuu__i coulndt use it as a phone anymore18:20
freemangordon"all telephony functions disabled"?18:20
DocScrutinizer05crossed out SIM?18:21
myuu__who me?18:21
DocScrutinizer05sure18:21
myuu__no im not having any problems i just upgraded my phone18:21
myuu__but i noticed a massive increase in battery18:21
DocScrutinizer05you totally lost me18:21
freemangordonme too :)18:21
freemangordon<myuu__> i love this phone but it wasnt useable18:22
freemangordon<freemangordon> myuu__: wasn't usable what for?18:22
freemangordon<DocScrutinizer05> and why?18:22
myuu__the battery18:22
freemangordonaah18:22
freemangordonstrange18:22
DocScrutinizer05not strange, sounds just like PEBCAK18:22
freemangordonhehe18:23
keriohuh?18:23
DocScrutinizer05aka "rogue p2p and widgets"18:23
merlin1991hm remotesms has a wrong checksum aswell18:23
kerioyou only use your n900 while sitting down?18:23
freemangordonkerio: mine does more than 24hours with heavy usage18:23
myuu__really?18:24
DocScrutinizer05merlin1991: same18:24
merlin1991kerio: now where do I dump the repo url? :D18:24
freemangordoni.e: online all day long, 4 IM accounts, BT handsfree, 2-3 hours of browsing, etc18:24
DocScrutinizer05merlin1991: version missing in last entries in history on http://maemo.org/packages/view/remotesms/18:24
keriofreemangordon: yeah, that's about right18:24
merlin1991I added a ignore list for the offending packages for now18:24
kerioi charge it once per day18:24
DocScrutinizer05I might be numb, but it looks borked to me18:24
freemangordonkerio: but myuu__ said it was unusable because of the battery, ehatever that means18:25
* DocScrutinizer05 suspects aborted/failed build on buildserver, due to messed up versioning18:25
* merlin1991 is not sure where to put this new and awesome repo18:25
merlin1991url wise18:25
keriomerlin1991: maemo.merlin1991.at/light/extras-devel18:26
keriofreemangordon: some people actually talk on the phone18:26
kerioi know, it's weird18:26
merlin1991it's only 11 gigs compared 100 gigs real devel18:26
freemangordonkerio: this is only device I use18:26
keriome too18:26
keriobut i almost never talk on the phone18:26
freemangordonand i use it as a phone too18:26
myuu__well the last straw was the usb18:26
freemangordonmy phone number is both corporate and personal18:27
myuu__but mine wouldnt last overnight from full charge18:27
freemangordonmyuu__: connected to internet?18:27
DocScrutinizer05merlin1991: (11 vs 100) sounds about right18:27
myuu__was running 0 widgets, 3 homescreens18:27
keriomyuu__: you had a busted battery18:27
freemangordonyeah18:27
kerioor some weird shit going on in the background18:27
myuu__connected to 3g but idle18:27
myuu__well whatever was going on should still be going on now, unless it was directly related to 3g18:28
freemangordonthat should easily make it through 3-4 days18:28
merlin1991kerio: http://maemo.merlin1991.at/light/extras-devel/18:28
merlin1991:D18:28
DocScrutinizer05myuu__: there are decent tools to investigate and usually exactly spot the culprit18:28
kerio^_^18:28
merlin1991now I need to test if it works18:28
DocScrutinizer05\o/18:28
myuu__well the battery performance is actually good now couldnt have been the actual battery18:28
* DocScrutinizer05 hands merlin1991 the crown18:28
merlin1991it's as reduced as possible, only the latest debs, not even sources18:28
freemangordonmyuu__: old sim?18:28
kerio:D18:28
myuu__im guessing18:29
freemangordonmyuu__: btw you don;t really need 3G18:29
keriofreemangordon: unless your operator is 3g-only18:29
DocScrutinizer053G is a known battery hog18:29
freemangordonkerio: hmm, never seen such operator, but I guess it is not valid worldwide18:30
DocScrutinizer05especially when doing actual data over it18:30
freemangordonDocScrutinizer05: if you don't make a data connection it is ok18:30
keriofreemangordon: H3G is 3g-only, at least here in italy18:30
DocScrutinizer05and even worse when your operator is a moron and allows inbound connections18:30
keriohehe18:30
myuu__do they make a microsim to old sim adapter? id like to actually test aftre cleaning up some packages18:30
keriooh god, microsims :(18:30
myuu__yeah =\18:30
DocScrutinizer05yes18:31
merlin1991hm crontab syntax, wtf is dom mon and dow18:31
freemangordonmyuu__: why microsim?18:31
kerioday of month, something and day of week18:31
DocScrutinizer05there are such adapters. NEVER EVER use them in slot-in SIM sockets18:31
DocScrutinizer05on N900 it's safe18:31
freemangordonnormal SIMs not produced anymore?18:31
merlin1991can it be that in /etc/cron.d/ the syntax is slightly different?18:31
myuu__of course they are but i went with a contract phone this time, seems like the ones being pushed out now are all or mostly micro18:32
DocScrutinizer05no, except user18:32
freemangordonmyuu__: and how is your new phone battery life?18:33
keriomerlin1991: the usual fremantle free non-free, right?18:33
merlin1991yep18:33
myuu__10 hours with some heavy use18:33
merlin1991okay crontab is in place, it should update by itself18:34
keriomerlin1991: how quick are you updating it?18:34
merlin1991daily18:34
kerio*how often18:34
* DocScrutinizer05 found it hard to get less than that on normal 'heavy' N900 usage18:34
merlin1991because I source my local apt-mirror mirror and that one runs daily18:34
freemangordonmyuu__: but,but.. its worse18:34
kerioi see18:34
myuu__well i havent actually seen this battery run out18:35
merlin1991I guess you prefer every 5 minutes ;)18:35
myuu__my n900 wouldnt last the day18:35
DocScrutinizer05myuu__: pretty please investigate what's the problem18:35
myuu__i dont doubt it was something frequently checking the 3g18:35
myuu__because like i said18:35
myuu__the n900 has been on for two days18:36
freemangordonmyuu__: an advice: go and find what's wrong with your n90018:36
DocScrutinizer05myuu__: I don't like unqualified badmouthing the N900 battery standby time18:36
myuu__yeah im going to buy an adapter18:36
myuu__well im not badmouthing18:36
myuu__i was just asking what the biggest batteryhog was18:36
freemangordonit can easily do 2-3 days with light usage18:36
myuu__i literally just said the phone has been on for two days18:36
freemangordonor even more18:36
merlin1991btw n900 whitout any connection idles for over a week :D18:36
myuu__while ive been cleaning packages18:36
myuu__yeah its not even at 70%18:36
myuu__dont take this the wrong way lol18:37
myuu__n900 was my favorite phone easily18:37
DocScrutinizer05even with WLAN it still idles 5...8 days18:37
freemangordonyep, if router has proper PM18:37
kerioi don't have any idea of the idle battery life18:37
keriomy phone is almost never idle :318:37
DocScrutinizer05kerio: get more N90018:38
DocScrutinizer05;-P18:38
freemangordonwhile still possible :D18:38
kerioi wonder what's the HAM catalog for that extras-devel18:38
DocScrutinizer05as long as you can't afford one N900 idling for a week, you got too few of them18:39
DocScrutinizer05kerio: "merlin light-e-devel18:39
keriono, the catalog level or whatever18:39
kerioi mean, as far as i understood it's just a matter of gpg keys18:40
freemangordonmerlin1991: why not put .install file in that catalog?18:40
merlin1991what catalog?18:40
* DocScrutinizer05 just thought we're a pretty cute and elitist bunch here ;-P18:40
freemangordonthe repo, sorry18:40
merlin1991freemangordon: that's the next step :D18:40
keriobut then why did Pali add the same gpg keys to four different catalog domains?18:40
DocScrutinizer05own repos, own channel, own "distro"18:40
merlin1991kerio: it goes repo name -> repo-id -> gpg key18:41
kerioDocScrutinizer05: we're just better than that lowly bunch in #maemo18:41
keriohm18:41
kerioi see18:41
keriomerlin1991: are the packages in light-extras-devel signed with your key?18:41
merlin1991yes18:41
merlin1991obviously I can't resign with the maemo.org key .D18:41
merlin1991but since you have cssu-testing the key is added anyway18:42
kerioalright, so with the new community-ssu-enabler those packages end up being higher than the nokia packages18:42
kerio:D18:42
merlin1991nope they end up below18:42
freemangordonDocScrutinizer05: aah, now I see why you're against thumbified -stable. You want to keep the good stuff only for you ;P18:42
merlin1991because the catalogue they come from has no associated id and priority rating18:42
DocScrutinizer05hehehe18:42
keriofreemangordon: he doesn't even use thumb :(18:42
freemangordondoes not matter18:42
freemangordonkerio: there is a joke here for a dog which neither eats the bone nor gives it to anyone :D18:43
kerioDocScrutinizer05: try it!18:44
keriopleeeeeeeeeeeeeeeease :318:44
merlin1991apt-worker seems to be faster now :)18:44
merlin1991btw I've added a .install file18:44
freemangordongood18:44
freemangordongoing to try it18:45
keriomerlin1991: bzipped Packages maybe?18:45
* DocScrutinizer05 sighs18:45
merlin1991kerio: gzipped18:45
kerioit's still 7mb18:45
kerioD:18:45
* DocScrutinizer05 ponders to grab one of the virgin N900 sitting in shelf, to run thumb on it18:45
freemangordongzip is in extras, ain't?18:45
DocScrutinizer05could somebody help me out with a 5 words of instruction? (sorry I'm actually not used to that stuff anymore)18:46
freemangordongoto maemo.merlin1991.at18:47
DocScrutinizer05assume device has WLAN18:47
kerioDocScrutinizer05: starting from pr1.3?18:47
DocScrutinizer05yep18:47
freemangordonopen cssu/thumb18:47
kerioDocScrutinizer05: http://maemo.merlin1991.at/cssu/community-thumb/community-thumb-fremantle.install18:47
freemangordonclick on .install file18:47
keriothen run the enabler18:47
keriothen ham upgrade18:47
DocScrutinizer05ok18:47
keriohmm18:47
freemangordonDocScrutinizer05: the same as for other CSSU flavours18:48
keriohttp://maemo.merlin1991.at/cssu/community-thumb/community-thumb-fremantle.install, run enabler, ham upgrade18:48
DocScrutinizer05seems feasible, even during having a shower18:48
kerio^ 5 words18:48
DocScrutinizer05fine :-)18:48
DocScrutinizer05duh, battery missing18:48
freemangordonhehe18:48
merlin1991DocScrutinizer05: it's enough if you have it attached to a charger, err the whole battery?18:49
DocScrutinizer05found one18:49
DocScrutinizer05date, region, language? GNAH!18:50
DocScrutinizer05eeeek that theme is sooooo ugly18:51
merlin1991which one are you ususally using?18:51
DocScrutinizer05err, not THIS one18:52
DocScrutinizer05and not those idiotic widgets ;-P18:52
freemangordonWTF: gzip (1.3.5-10) unstable; urgency=medium18:52
freemangordon-- Bdale Garbee <bdale@gag.com>  Fri, 20 May 2005 22:34:49 -060018:52
freemangordonmerlin1991: consider updating gzip in CSSU18:53
DocScrutinizer05W*T*F?! :-o kbd clicks are like 10% 100% 10% 10% 100% volume on this device18:54
jon_yit's telling you to go big or go home :)18:55
merlin1991hm who's upstream for gzip?18:55
merlin1991well the speedup with the smaller repo isn't that intense but it's at least something18:58
DocScrutinizer05ok, installation running, i'm off to have a shower18:58
merlin1991I guess I should try thumb and compare :D18:58
* kerio pictures DocScrutinizer05 returning from the shower and having to tap "accept" on the kernel installation :)18:59
freemangordonmerlin1991: it is way faster. btw thumb-compiled gzip is in the repo, lets see if there will be any difference18:59
freemangordonhmm, something's wong here19:01
freemangordon*wrong19:01
freemangordondoes apt use its own gzip?19:01
merlin1991I have no idea19:03
keriofreemangordon: it uses bzip2 too19:03
kerioa lot of bzip219:04
keriomerlin1991: there's no guarantee people have gzip installed19:04
freemangordonI don;t have it19:04
freemangordonso my apt gzip is alreay thumb19:04
freemangordon*already19:04
kerioyour apt gzip is busybox19:05
DocScrutinizer05hmm, i'm going to "update all"?19:05
keriobut apt doesn't use gzip anywhere, with the maemo repos19:05
keriobzip2 is used19:05
freemangordonDocScrutinizer05: only maemo19:05
keriofreemangordon: thumb-compile bzip219:05
DocScrutinizer05k19:05
freemangordonkerio: i will compile bb19:05
keriofreemangordon: yeah but it won't help apt at all19:05
kerioapt uses gzip if there's something to gunzip19:06
keriobut there's not, because pretty much everything is bzipped19:06
freemangordonkerio: I don't have gzip installed on my device19:06
keriofreemangordon: no one has19:06
kerio/bin/gzip is owned by busybox19:06
freemangordonbut apt-get update says "gzip"19:06
freemangordonnot bzip219:06
kerioPackage gzip is a virtual package provided by:19:06
kerio  busybox 3:1.10.2.legal-1osso31+0cssu019:06
kerioand it says gzip because merlin is silly19:07
freemangordonyes, that is why i will compile bb19:07
freemangordongot it?19:07
* DocScrutinizer05 sticks a stickytape patch on frontcam and takes N900 to bathroom19:07
* kerio points freemangordon to http://repository.maemo.org/extras-devel/dists/fremantle/free/binary-armel/Packages.bz219:08
freemangordonDocScrutinizer05: you need a charger if your battery is depleted19:08
kerionot that bzip2 saves much space, mind you19:08
DocScrutinizer05my bathroom has power outlets ;-P19:08
freemangordonkerio: and bzip2 is part of bb or not?19:08
keriofreemangordon: it's not19:08
DocScrutinizer05though my spare batteries usually are ~80% charge19:09
keriobzip2 is part of bzip219:09
myuu__mine too19:09
myuu__whats with that19:09
kerioi wonder if there's a way to make apt prefer gzip19:09
myuu__i know this is an old building but it was restored, there obv a need for bathroom outlets19:10
freemangordondpkg -l | grep bzip19:10
freemangordonrc  libzip1                                             0.8-119:10
DocScrutinizer05ohhh you!!!! *****  ctlr-bs doesn't allow me to escape from progress bar to task switcher19:10
keriofreemangordon: oh, you haven't installed it? :(19:10
kerioDocScrutinizer05: it's probably a system popup or something like that19:10
freemangordonkerio: no, and apt still works19:11
DocScrutinizer05yeah with delay. Now it looks thoroughly fsckdup19:11
keriofreemangordon: then i suppose that you fall back on gzip19:11
kerio\_o_/19:11
DocScrutinizer05blank top bar, HAM scroll area, bottom progress bar19:11
merlin1991kerio: nah we should use lzma ;)19:11
freemangordon"/usr/lib/apt/methods"19:11
keriomerlin1991: wtf, your http server is kinda busted19:12
keriono nvm, it's just that i changed the preferred compression method19:12
DocScrutinizer05seems this will take a few minutes now19:12
freemangordonyeah 10 or so19:12
DocScrutinizer05bbl19:13
DocScrutinizer05eeeek it boots19:15
DocScrutinizer05;-P19:15
freemangordonyou look suprpised :P19:15
DocScrutinizer05shocked, by nokia tune19:16
freemangordonBTW I think we should ship transitions.ini with reduced transitions time, 80% of the "lag" comes from it19:16
DocScrutinizer05~lart useless transitions19:17
* infobot moos at useless transitions19:17
keriofreemangordon: while you're there, can you recompile busybox-power too? :319:17
freemangordonkerio: it is recompiled19:17
freemangordonand is in the repo19:17
DocScrutinizer05ouch!!! OVI docsToGo update19:17
kerioDocScrutinizer05: hahaha19:18
merlin1991DocScrutinizer05: you should always have a hand "remove all the cruft" script :D19:18
* DocScrutinizer05 tags a postit note to screen19:19
DocScrutinizer05documents to go installer KILL KILL19:20
freemangordonDocScrutinizer05: try microb ;)19:20
merlin1991I have a nice shell script I execute whenever I reflashed a device for new testing of $randomthing19:20
DocScrutinizer05you share that, you.... :-)19:21
DocScrutinizer05hmm, microB works. Can't say if better19:23
DocScrutinizer05no idea yet how to compare19:23
freemangordonopen some heavy site19:23
freemangordonFB if you have an account19:23
freemangordonyoutube should do the job too19:24
DocScrutinizer05http://some_heavy_site  50019:24
*** abufaisal has joined #maemo-ssu19:24
DocScrutinizer05hey 6 times 100% one 10% for youtube19:24
DocScrutinizer05indeed seems to render like mad, subjectively19:25
freemangordonyep19:25
kerio:D19:26
DocScrutinizer05ok, bought19:26
kerioyay ^_^19:27
freemangordonrty apt-get update19:27
keriohave you done a side-by-side comparison?19:27
freemangordon*try19:27
freemangordonin the terminal19:27
freemangordonkerio: noone has done such comparisons so far19:27
DocScrutinizer05freemangordon: apt-get update is worthless since completely different set of repos19:27
freemangordonhmm, try it, nokia, ovi and maemo-extras are there19:28
DocScrutinizer05kerio: I'll do in an hour19:28
DocScrutinizer05and might video it19:28
freemangordonDocScrutinizer05: though make sure you use the same conditions, i.e no sim cards and such19:29
DocScrutinizer05sure thing19:29
DocScrutinizer05will switching to tablet mode do as wellß19:29
DocScrutinizer05?19:29
freemangordonhmm, but what you will comare then?19:29
freemangordon*compare19:30
DocScrutinizer05flash still renders like a pil'o'shit19:30
kerioflash is not thumb :)19:30
freemangordonwell, othing to be done there :(19:30
DocScrutinizer05:nod:19:30
freemangordonDocScrutinizer05: and have in mind only about 10-15 packages are recompiled19:31
DocScrutinizer05I'll just compare microB (unless I go mad)19:31
freemangordonwhile essential ones like clutter,gtt,qt,xserver, it is still far from what can be achieved19:31
DocScrutinizer05some tardhead made a gitmo video with "FZ:The torture never stops"19:32
DocScrutinizer05quite obviously ignored or didn't get the sexual connotation19:33
keriohaha19:33
DocScrutinizer05or i'm missing the sexual sub-lines in gitmo pics19:33
DocScrutinizer05OHNOES, NAM19:34
freemangordoniirc " The torture never stops" as WASP's song19:34
kerioOHNOES19:34
keriowhat happened?19:34
keriofreemangordon: apt-worker is recompiled too, right?19:34
freemangordonyep19:34
DocScrutinizer05FZ with well known 'nam little children19:34
kerioDocScrutinizer05: somehow i think the speedups are more related to gcc 4.6/4.7 than thumb :)19:34
DocScrutinizer05instead gitmo19:34
freemangordonkerio: it is combined19:34
freemangordonless RAM usage/better compiler19:35
DocScrutinizer05search tube for zappa, click on first "torture never stops" from dropdown19:35
freemangordonthough I agree newer gcc plays major role there19:35
DocScrutinizer05incredibly stupid picture material to that nice song19:36
* freemangordon wonders where is zeq19:36
DocScrutinizer05took his donated N900 and ran? ;-P19:36
freemangordonhehe19:36
freemangordonhe said he has some job right now19:36
freemangordonwhich is good, having in mind he is a freelancer19:37
freemangordonI hope you are wrong19:37
DocScrutinizer05freemangordon: HEY, I actually can compare to virgin plain vanilla 1.319:38
DocScrutinizer05got another N900 ;-D19:38
freemangordonhmm, not fair19:38
freemangordonfor PR device that is :P19:38
DocScrutinizer05what now?`19:38
freemangordoncompare with CSSU-T19:38
freemangordonbecause of Qt19:38
DocScrutinizer05that's my daily phone then19:39
freemangordonyeah, should work. but you will have higher memory usage on your daily device because of the contacts, etc19:39
freemangordonyou'd better install CSSU-T on that PR13 device19:40
merlin1991I call for IMPERIAL testing19:40
freemangordonand use that, otherwise a comparison between daily device with contacts,music, whatnot will result of -thumb winning with several laps19:41
freemangordonDocScrutinizer05:also I suspect you have swap on uSD19:42
DocScrutinizer05are you mad?19:43
freemangordonme? why is that?19:43
DocScrutinizer05me neither ;-P19:43
freemangordonswap on uSD is the next best thing after the hot water :P19:43
DocScrutinizer05nah, swap on uSD is too dangerous for me19:43
freemangordonhmm, strange19:43
DocScrutinizer05ya'know, my pouch has a mgnetic clip19:44
freemangordonanyway, you seem experienced enough to know how to do a correct comparison19:44
DocScrutinizer05and once every two years I hotswap uSD19:44
DocScrutinizer05I hope so ;-D19:45
freemangordonswapoff/swapon?19:45
DocScrutinizer05nanannannaaaa van't hear you (/me puts fingers in ears)19:45
kerioDocScrutinizer05: just go the estel route and glue a magnet on the sensor XD19:45
DocScrutinizer05*COUGH*19:45
freemangordonkerio: quiet, don't say that word to doc19:46
merlin1991and then hotswap the µSD :D19:46
keriowhy, does he have a problem with glue?19:46
freemangordonhehehe19:46
freemangordonkerio: not with the glue, but with the word :P19:47
DocScrutinizer05es<tab>  ... <tab><tab><tab>  :-P19:48
DocScrutinizer05~seen estel_19:49
keriohe'll always be in the userlist of our heart19:49
infobotestel_ <~Estel@Maemo/Community/council/Estel-> was last seen on IRC in channel #maemo, 16d 5h 18m 37s ago, saying: 'happy trolling, i'm out.'.19:49
DocScrutinizer05incredible he really meant it this time19:49
DocScrutinizer05though, I guess he's probably already here under another nick not banned19:50
keriohaha19:51
merlin1991well unless he wants to voice his opinion he doesn't even have to be in here to read what's going on, there's such a thing as logs19:51
DocScrutinizer05sure19:51
DocScrutinizer05I don't mind even ants and marsians reading logs19:51
DocScrutinizer05I'm however quite thankful for teotwaki taking action19:52
DocScrutinizer05and I'm not allowed to do any administrative action against estel_, so I won't relieve the +b19:54
lufCan anybody who has installed CSSU-T 5+ send me result of:19:55
lufcat /usr/share/dbus-1/services/rtcomeventlogger.events.EventsBook.service19:55
* kerio doesn't have that file19:56
freemangordon-sh: /usr/share/dbus-1/services/rtcomeventlogger.events.EventsBook.service: not found19:56
lufdpkg -L rtcom-eventlogger-async19:57
*** fredrinLap has joined #maemo-ssu19:57
luf?19:58
DocScrutinizer05Package `rtcom-eventlogger-async' is not installed.19:58
lufFuck.19:58
merlin1991wtf I checked if there is a dependency on eventlogger-async and there was one iirc19:58
kerionot installed here19:59
merlin1991hm rdepends is empty19:59
merlin1991wtf19:59
keriomerlin1991: no reverse depends \_o_/19:59
lufOk so I have another question (longer one). I have obexd and rtcom-eventlogger-async. rtcom isn't needed by obexd unless you want call history. So I didn't put it into obexd dependency. But nobody knows that rtcom -asyn is needed.20:00
keriocall history where?20:00
merlin1991car kit20:00
lufPBAP call history.20:00
kerioi see20:00
kerioi would've put it as a dependency20:00
kerioalthough...20:00
keriohm20:01
merlin1991obexd depends on librtcom-eventlogger1-async20:01
*** abufaisal has quit IRC20:01
* kerio thinks that's just the library, and not the daemon that logs20:01
lufmerlin1991: yes. it's the client part.20:01
lufBut not the dbus deamon.20:01
kerioluf: how big of a log does it keep?20:02
lufWhole ;)20:02
keriois it the same as the call list in Phone?20:02
merlin1991err why the fuck does the client library not depend on the server dbus deamon20:02
DocScrutinizer05err is this a discrete log? isn't it just records in genuine call log?20:02
DocScrutinizer05or just what kerio asked20:03
merlin1991it makes nfc to have client stub library that has no dependency on the "same system" dbus service20:03
lufkerio: yes. It gets the data from the same source ;)20:03
kerioi see, so it doesn't really keep a log20:03
kerioit uses the log that's already there20:03
merlin1991luf: you should have had a dependency from the client lib to the server20:03
lufkerio: sure.20:04
keriothen i'm cool for its addition20:04
keriohm, does it add additional resource usage if you're *not* using carkit?20:04
DocScrutinizer05seems kerio adopts my standard catalog of questions :-)20:05
lufkerio: no. As it's dbus daemon which is started when it's called for the first time.20:05
kerioperfect, then20:05
kerioand i assume it just waits for requests from obexd?20:05
lufOk. I don't like it but I add the dependecy to the client part.20:05
merlin1991luf: why don't you like it?20:05
merlin1991you have a dbus client helper lib, ofc it should depend on the respective service20:06
lufmerlin1991: because it's needed only in one specific case.20:08
lufIt leads to situation you have to install a lot of packages even you don't use them at all.20:09
merlin1991so you want users of PBAP to manually install the dbus service when they need it?20:09
lufmerlin1991: I understand. That was my question if there is something less restrictive than depend on.20:10
DocScrutinizer05just optify that stuff, so it doesn't eat up space on NAND20:11
merlin1991well I don't know what obexd does if PBAP is up and the dbus service is missing20:11
DocScrutinizer05(assuming you can't do a call during early boot anyway)20:11
merlin1991if it works over that gracefully you can ofc keep the dbus service optional20:11
lufmerlin1991: Just returns empty call history list ;)20:11
lufI'll add the dependency into helper package.20:12
DocScrutinizer05optify only for files >1k obviously20:13
DocScrutinizer05maybe even only >5k20:13
DocScrutinizer05luf: PBAP-basic (comes with CSSU), BPAB-full (includes history and depends on -basic)20:15
DocScrutinizer05I guess you meant that when you said "hekper package"20:15
merlin1991DocScrutinizer05: only that almost everything is part of the obexd binary ;)20:16
DocScrutinizer05no, aiui the dbus service isn't20:16
DocScrutinizer05so the daemon isn't ?20:16
merlin1991it's a tiny tiny service to govern access to the eventlogger db, it's the smallest part :D20:17
lufDocScrutinizer05: the daemon has 14 kB :)20:17
DocScrutinizer05yeah, my thought on that: optify and forget20:17
DocScrutinizer05nobody gives a fsck about some (100) kB on /opt20:18
DocScrutinizer05NAND though is another game20:18
merlin1991freemangordon: when do you think you'll push a new mp20:18
* merlin1991 feels like running grep sort and friends to gather statistics .D20:19
freemangordonmerlin1991: don't know maybe in 3-4 days, why?20:19
merlin1991see above ;)20:19
freemangordonI saw it, but i'll wait for new -T first20:19
freemangordon:P20:20
freemangordon-thumb follows -testing ;)20:20
merlin1991new -T, hm I suppose 1 to 3 months from now? :D20:20
* DocScrutinizer05 waits for new S20:20
merlin1991WTF happened to maemo.org20:20
* freemangordon is going afk for a while, bbl20:20
merlin1991I'm missg all css on the wiki20:20
DocScrutinizer05weekend madness20:20
merlin1991DocScrutinizer05: what's currently a candidate for new S ?20:21
DocScrutinizer05errr, dunno20:21
DocScrutinizer05ke-recv?20:21
DocScrutinizer05I thought there was quite some pkgs to finally include to S20:22
merlin1991see http://wiki.maemo.org/Community_SSU/ChangelogStable#21.2011.38-1Smaemo4_.2805.09.2012.29 apart from the stuff in the latest -T everything that is good enough is in S20:22
DocScrutinizer05ooh 5.9.20:23
DocScrutinizer05nm20:23
DocScrutinizer05I should finally get some offtime and head out to RL20:24
Gussefreemangordon: Modest working OK again with thumb3.  Thanks! Unbelievable great support :)20:35
freemangordonGusse: np, after all that's why we are here :)21:03
freemangordonmerlin1991: do you plan to push new -T with fixed obexd sevrice?21:04
merlin1991if luf optifies the rtcom-eventlogger-async, then yes21:05
freemangordonwell then, new -thumb will follow the new -T21:05
freemangordonin the meanwhile i'll thumb-comile more packages21:05
freemangordon(and willl try to find why xsevers is broken when gcc 4.7,2 copiled)21:08
freemangordonthu fuc :D21:08
freemangordoncompiled, compiled21:08
freemangordonI suspect the same issue as with modest21:08
merlin1991what was the issue with modest?21:09
freemangordonfor some reason exports were not exported :(21:09
freemangordonMfE plugin imports functions from modest, but they were missing21:10
freemangordonI added -shared -rdynamic (along with -pie -mword-relocations to make maemo-invoker happy) to compile flags and it worked21:11
*** nox- has joined #maemo-ssu21:19
keriofreemangordon: gcc-4.7 isn't thumb21:19
*** fredrinLap has quit IRC21:39
*** dhbiker has quit IRC22:11
*** dhbiker has joined #maemo-ssu22:14
*** cryptonite has joined #maemo-ssu22:23
cryptonitehello22:24
cryptonitemay i ask if cssu and cssu-thumb are differemt things?22:27
*** toxaris has quit IRC22:29
freemangordonkerio: what do you mean?22:31
freemangordoncryptonite: they are different and the same ;)22:31
freemangordon-thumb is cssu-testing recompiled with new toolchain and different instructions set22:32
cryptoniteok22:32
freemangordoncryptonite: http://talk.maemo.org/showpost.php?p=1220597&postcount=122:32
cryptoniteso can i move from cssu-stable to cssu-thumb?22:32
freemangordonin theory yes22:33
cryptonitebut in practice?22:33
freemangordonDocScrutinizer05 moved one of his spare devices from stock PR13 to -thumb with no problems just a few hours ago22:34
cryptonitei dont want to have to reflash my phone22:34
freemangordonin practice too, though I don;t know if anyone have ever done it22:34
cryptonitehow is that?22:34
cryptonitei mean what do most people install?22:35
freemangordonwell, i said the I don't know22:35
cryptoniteok22:35
cryptonite:)22:35
freemangordonmost move from -testing to -thumb AFAIK22:35
freemangordoncryptonite: see, it should be safe22:36
freemangordonbut better do a backup before trying it22:36
freemangordon~bm22:36
infobotbackupmenu is, like, http://talk.maemo.org/showthread.php?t=6397522:36
freemangordon^^^22:36
freemangordonjust to be on the safe side22:36
*** NIN101 has quit IRC22:36
freemangordoncryptonite: and make sure you've read the THUMB thread on tmo22:37
DocScrutinizer05sane advice22:37
*** BCMM has joined #maemo-ssu22:37
cryptonitei am reading it at the moment22:37
freemangordonok22:37
cryptoniteso i will do a backup in the first place22:38
*** Pali has quit IRC22:38
freemangordonyou'd better do22:39
freemangordoncryptonite: which kernel do you use?22:39
keriocryptonite: it should be the same as a move between cssu-stable and cssu-testing22:39
cryptoniteright now the power5122:39
freemangordonwell, you should be safe then, there won;t be a kernel flashed22:40
cryptoniteso then i will just click the link on the thumb thread refering to community-thumb-freemantle and wait for the transition?22:43
cryptonitefingers crossed22:43
*** freemangordon has quit IRC22:44
*** freemangordon has joined #maemo-ssu22:48
freemangordonthe fuck, a power outage for a fraction of a second here :(22:49
freemangordonbut enough to restart all of the machinery around22:49
freemangordoncryptonite: yes22:50
cryptoniteu no ups?22:50
cryptonite:)22:50
freemangordonno, why should I need UPS at home22:50
cryptonitedo i have to do the manual editing as suggested in #66622:53
freemangordoncryptonite: link please22:55
cryptonitehttp://talk.maemo.org/showpost.php?p=1267008&postcount=66622:57
freemangordoncryptonite: no, since you are a new user22:57
freemangordonthis is if you already have older version installed, which you don't22:57
cryptoniteok thanx a lot for your help22:58
cryptonitelogging out to test things22:58
cryptonitei hope i will return with good news22:59
cryptonitecya22:59
*** cryptonite has quit IRC22:59
*** Gusse has left #maemo-ssu23:04
*** dhbiker has quit IRC23:18
kerioshould CSSU ship a proper "su"?23:26
lufWhat about optifying whole obex?23:56
freemangordonluf: when is obexd started?23:57
freemangordonas /opt is mounted relatively late23:57
luffreemangordon: to be honest I have no idea :)23:57
kerio/home is mounted relatively late, and /opt is bindmounted even later23:58
lufBut it's running under user so I think it's late enough.23:58

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