*** Free-MG has quit IRC | 00:52 | |
*** Raimu-Z has quit IRC | 01:16 | |
*** Raimu-Z has joined #maemo-ssu | 01:18 | |
*** Pali has quit IRC | 02:07 | |
*** arcean has quit IRC | 02:19 | |
*** gri has quit IRC | 03:14 | |
*** gri has joined #maemo-ssu | 03:27 | |
*** M4rtinK has quit IRC | 03:53 | |
*** nox- has quit IRC | 04:38 | |
*** amiconn has quit IRC | 05:39 | |
*** amiconn_ has joined #maemo-ssu | 05:39 | |
*** amiconn_ is now known as amiconn | 05:39 | |
*** DocScrutinizer05 has quit IRC | 06:35 | |
*** DocScrutinizer05 has joined #maemo-ssu | 06:35 | |
*** Mephistopheles has joined #maemo-ssu | 10:01 | |
*** Pali has joined #maemo-ssu | 10:31 | |
*** LaoLang_cool has joined #maemo-ssu | 10:36 | |
*** andre__ has joined #maemo-ssu | 10:46 | |
*** andre__ has joined #maemo-ssu | 10:46 | |
*** LaoLang_cool has quit IRC | 10:50 | |
*** andre__ has quit IRC | 10:51 | |
*** andre__ has joined #maemo-ssu | 11:08 | |
*** Skry has joined #maemo-ssu | 11:14 | |
*** Pali has quit IRC | 11:41 | |
*** arcean has joined #maemo-ssu | 12:06 | |
*** BCMM has joined #maemo-ssu | 12:24 | |
*** Mephistopheles has quit IRC | 12:27 | |
*** M4rtinK has joined #maemo-ssu | 13:26 | |
*** lizardo has joined #maemo-ssu | 13:28 | |
* merlin1991 uploads 569Mb of stuff | 13:49 | |
merlin1991 | or in other words the -testing update is incoming | 13:49 |
---|---|---|
*** Pali has joined #maemo-ssu | 13:51 | |
kerio | :D | 13:51 |
kerio | freemangordon: start compiling 280Mb of stuff | 13:51 |
merlin1991 | fsckd changelog | 14:01 |
merlin1991 | ETOOMANYCHANGES | 14:02 |
*** arcean has quit IRC | 14:07 | |
*** arcean has joined #maemo-ssu | 14:08 | |
*** lizardo has quit IRC | 14:09 | |
*** lizardo has joined #maemo-ssu | 14:22 | |
merlin1991 | Pali: ping | 14:26 |
Pali | merlin1991, pong | 14:26 |
merlin1991 | for the osso-wlan eap fix, did we have a bugreport? | 14:26 |
Pali | I do not know but, patch is correct for me | 14:26 |
Pali | last param for exec functions must be NULL | 14:28 |
merlin1991 | yeah I trust you, I was only asking because I'm currently writing the changelog and I'd like to reference the bug # if there is one | 14:28 |
Pali | ok | 14:30 |
Pali | what about operator name widget fixes? | 14:30 |
merlin1991 | I built it from git | 14:30 |
merlin1991 | will be included | 14:30 |
Pali | ok | 14:32 |
Pali | merlin1991, do you have permission for extras repository? | 14:33 |
merlin1991 | still not working | 14:34 |
merlin1991 | hm 3rd attempt to upload the qt4 sources, why does my connection have to drop all the time today? WHY? | 14:51 |
kerio | Pali: btw, i still haven't found any reference to mmcblk0p2 in ke-recv | 14:59 |
kerio | so i think it's something else that enables the mass storage mode | 15:00 |
Pali | kerio, mass storage mode is enabled by script osso-usb-mass-storage-enable.sh | 15:05 |
Pali | ke-recv call it for mmc device | 15:06 |
Pali | now I'm trying to find where is hardcoded mmc device partittion | 15:06 |
kerio | to be fair i didn't look too much into it, i just grepped the ke-recv source for mmcblk | 15:07 |
*** chainsawbike has quit IRC | 15:08 | |
*** chainsawbike has joined #maemo-ssu | 15:11 | |
Pali | seems that ke-recv export all mmc devices when mass storage mode is enabled | 15:12 |
Pali | and if some partition cannot be umounted then is not exported | 15:12 |
kerio | oic | 15:12 |
Pali | so is this enought? | 15:12 |
kerio | wait, no | 15:12 |
kerio | does it try to umount the whole of mmcblk1? | 15:12 |
Pali | see syslog | 15:14 |
kerio | i don't have it installed D: | 15:14 |
Pali | so install it | 15:14 |
kerio | what do i have to install? | 15:14 |
Pali | apt-cache search syslog | 15:15 |
Pali | some ksyslogd or syslogd or similar package | 15:15 |
kerio | sysklogd? | 15:15 |
Pali | should be | 15:15 |
merlin1991 | http://wiki.maemo.org/Community_SSU/Changelog#T-maemo5 | 15:19 |
merlin1991 | mp should arrive soonish | 15:19 |
*** arcean has quit IRC | 15:23 | |
merlin1991 | freemangordon: ^^ | 15:24 |
kerio | oh god, sysklogd fills /var/log with crap | 15:24 |
kerio | D: | 15:24 |
merlin1991 | hm | 15:31 |
merlin1991 | I copied the busybox depends from the pr mp but there's stuff missing | 15:32 |
merlin1991 | arf typo, ofc | 15:39 |
DocScrutinizer05 | kerio: fool, that's what /var/log/ is meant for | 15:46 |
DocScrutinizer05 | and it's not 'crap' | 15:46 |
kerio | i don't need to log every time i open and close the slide | 15:46 |
DocScrutinizer05 | then learn to config syslog | 15:47 |
DocScrutinizer05 | tbh I don't care what you *think* *you | 15:47 |
DocScrutinizer05 | ** neeed | 15:47 |
kerio | i don't wanna D: | 15:47 |
Skry | put /var/log to tmpfs and set maximum logsize to something you see appropriate | 15:47 |
DocScrutinizer05 | Skry: no use in suggesting good advice to renitent trolls | 15:48 |
kerio | nah, it's small enough to be reinstallable whenever i actually need to log stuff | 15:48 |
DocScrutinizer05 | hahaha | 15:48 |
DocScrutinizer05 | actually ver/log/syslog is small enough so I don't care | 15:49 |
*** ChanServ sets mode: +o merlin1991 | 15:53 | |
*** merlin1991 changes topic to "Maemo Community Seamless Software Update "CSSU" channel, http://wiki.maemo.org/Community_SSU | Known bugs: http://j.mp/communityssu-bugs | Channel logs: http://mg.pov.lt/maemo-ssu-irclog/ | Sources: http://gitorious.org/community-ssu/ | Latest version (testing): 21.2011.38-1Tmaemo5.1; (stable): 21.2011.38-1Smaemo4.1" | 15:54 | |
*** ChanServ sets mode: -o merlin1991 | 15:54 | |
DocScrutinizer05 | merlin1991: unfortunate you didn't allow prior testing of testing | 15:54 |
DocScrutinizer05 | merlin1991: I gather copying of repo wasn't that simple | 15:55 |
DocScrutinizer05 | btw why Tmaemo5.1 ? | 15:55 |
* DocScrutinizer05 checks changelog to learn about 5.0 | 15:56 | |
kerio | oh typos | 15:56 |
merlin1991 | when rearanging the busybox dependencies I managed to make a fail paste which fsckd 2 dependencies | 15:56 |
kerio | merlin1991: hmm, was the fix for the extra logging applied? | 15:56 |
DocScrutinizer05 | anyway changelog has no 5.1 | 15:56 |
kerio | http://wiki.maemo.org/Community_SSU/Changelog#T-maemo5.1 | 15:56 |
DocScrutinizer05 | ohh it has | 15:56 |
merlin1991 | scrolling helps ;) | 15:57 |
DocScrutinizer05 | indeed | 15:57 |
DocScrutinizer05 | ok, I'll immediately run a BM and then isntall update | 15:58 |
merlin1991 | also see: http://talk.maemo.org/showpost.php?p=1264791&postcount=793 | 15:59 |
*** luf has joined #maemo-ssu | 15:59 | |
merlin1991 | Pali: many thanks for the xterm patches, FINALLY I can set the text colour | 15:59 |
luf | merlin1991: curl also fixes the problem with some ciphers | 15:59 |
luf | There is also bug for it ... | 16:00 |
merlin1991 | luf: if you find the bugnumber fell free to add that to the changelog | 16:00 |
luf | So I can feel free to correct some typo also. Ok. | 16:01 |
merlin1991 | ofc, it's a wiki :) | 16:01 |
luf | https://bugs.maemo.org/show_bug.cgi?id=10233 | 16:01 |
povbot | Bug 10233: curl does not support certificates with a SHA-2 signature. | 16:01 |
luf | Are we able also to close the bug? | 16:01 |
*** X-Fade has quit IRC | 16:02 | |
*** X-Fade has joined #maemo-ssu | 16:02 | |
DocScrutinizer05 | merlin1991: http://talk.maemo.org/showpost.php?p=1264798&postcount=794 | 16:06 |
DocScrutinizer05 | merlin1991: please edit 793 to include more verbose explanation what exactly to do, in which sequence. You know about DAUs ;-) | 16:12 |
DocScrutinizer05 | sth like >>in HAM there will show up (among others): "CSSU-T 5 system update" and "opername-2.0". Either select "opername-2.0" *first* and install it, only *then* install CSSU-T. Or select "update all" option in HAM, which will automatically update opername-2.0 first...<< | 16:14 |
DocScrutinizer05 | (if that's^^^ correct. Otherwise adapt. Consider typeset bold) | 16:15 |
Pali | I think update all should work | 16:15 |
Pali | if you did not installed operator name widget, then everything is OK | 16:16 |
DocScrutinizer05 | I think so too, since opername probably is before T | 16:16 |
DocScrutinizer05 | alphabetically | 16:16 |
Pali | and if you installed it (e.g. via old cssu version) then you cannot update metapackage until you update operator name widget to new version (or you will uninstall it) | 16:16 |
Pali | and if HAM can solve dependency graph (like apt-get) then all is OK | 16:17 |
Pali | somebody wihtout cssu-devel/thumb should try what happends | 16:17 |
*** luf has left #maemo-ssu | 16:19 | |
kerio | oh so now operator-name-cbs-widget isn't actually in the metapackage anymore? | 16:21 |
kerio | neat | 16:21 |
Pali | kerio, is transitional package and can be removed | 16:22 |
kerio | i see | 16:22 |
* kerio currently waiting for fmg to give him the go | 16:23 | |
* kerio will do a BM backup in the meantime | 16:23 | |
kerio | *mean time | 16:23 |
*** andre__ has quit IRC | 16:23 | |
*** andre__ has joined #maemo-ssu | 16:24 | |
DocScrutinizer05 | dammit no, cbsms is after maemo-T | 16:24 |
DocScrutinizer05 | when I click "update all" | 16:24 |
DocScrutinizer05 | do I need to update cssu-enabler? I think I kicked it after installation | 16:27 |
DocScrutinizer05 | I don't plan to roll back cssu-t installation | 16:27 |
kerio | DocScrutinizer05: it used to be uninstallable, not sure if that's changed | 16:28 |
Pali | new enabler is needed for new HAM | 16:28 |
DocScrutinizer05 | ooh, ok | 16:29 |
Pali | now gpg keys and cssu repositories are properly installed into maemo system | 16:29 |
DocScrutinizer05 | so let's see how that works on *update* | 16:29 |
Pali | and new ham rejected on my n900 to install anything from cssu repositories without new enabler | 16:29 |
kerio | Pali: so the new cssu enabler enables the domain change in HAM? | 16:30 |
kerio | or whatevs | 16:30 |
Pali | yes | 16:30 |
Pali | cssu repositories have correct domain settings in ham | 16:30 |
kerio | oh, they're installed as "system repositories" now? neat | 16:31 |
Pali | as "trusted" | 16:31 |
kerio | the ones you can't modify, right? i see | 16:31 |
Pali | yes | 16:31 |
kerio | do i have to run community-ssu-enabler again then? | 16:31 |
Pali | packages from trusted repositories cannot be updated to version from non trusted (by ham) | 16:32 |
Pali | before -enabler script hacked ham to ignore it and then installed cssu | 16:32 |
Pali | now this hack was removed | 16:32 |
merlin1991 | kerio: nope just installing the new version is enough | 16:32 |
Pali | no, you do not need to run -enabler script again | 16:33 |
Pali | you only need to update enabler package, nothing more | 16:33 |
merlin1991 | btw DocScrutinizer05 we do have an uninstaller now :) | 16:33 |
DocScrutinizer05 | I know | 16:33 |
DocScrutinizer05 | [2012-09-11 15:27:38] <DocScrutinizer05> I don't plan to roll back cssu-t installation | 16:33 |
merlin1991 | :) | 16:34 |
DocScrutinizer05 | so will I have to (and be allowed to) uninstall enabler after update, again? | 16:34 |
kerio | Pali: hm, won't this cause duplications in the repo list? | 16:35 |
kerio | the .install adds a custom repo | 16:35 |
merlin1991 | have to no, allowed to no again | 16:36 |
DocScrutinizer05 | :-/ | 16:36 |
Pali | DocScrutinizer05, enabler can be uninstalled if you install cssu | 16:36 |
merlin1991 | *uninstall cssu* | 16:37 |
DocScrutinizer05 | I hope I'm at least allowed to nuke the icon off my desktop and menu | 16:37 |
Pali | kerio, no, duplicites are removed | 16:37 |
kerio | Pali: neat | 16:37 |
merlin1991 | DocScrutinizer05: there should be no icon whatsoever | 16:37 |
DocScrutinizer05 | mhm | 16:37 |
kerio | is the repo removed when community-ssu-enabler is removed? | 16:37 |
Pali | no | 16:37 |
Pali | wait | 16:37 |
DocScrutinizer05 | then I guess I have completely wrong idea about what's cssu-enabler | 16:37 |
Pali | it is removed when you uninstall package enabler | 16:38 |
merlin1991 | DocScrutinizer05: cssu enabler contains the system files for repo url and gpg keys + the script to bootstrap the initial cssu installation | 16:38 |
Pali | cssu-enabler package which install gpg keys + domains to system and contains script which install or uninstall cssu | 16:38 |
DocScrutinizer05 | and wtf is started via icon click on initial cssu installation then? | 16:38 |
merlin1991 | the script, but if your device is not fsckd you should have no icon | 16:39 |
DocScrutinizer05 | I'm not sure I get it | 16:39 |
Pali | it install CSSU metapackage | 16:39 |
Pali | after installing cssu metapackage postinst will remove icon | 16:40 |
Pali | also when enabler is updated it check if metapackage is installed and remove icon too | 16:40 |
DocScrutinizer05 | do you say the script started by icon is the enabler script but installing (aka updating) cssu-enabler will *not* create the icon to start it? | 16:41 |
Pali | it will unpack icon from enabler package | 16:41 |
DocScrutinizer05 | mhm | 16:41 |
Pali | but postinst will delete it | 16:41 |
DocScrutinizer05 | we'll see | 16:42 |
merlin1991 | DocScrutinizer05: why don't you try it out before you go on a rant about a possible problem? :D | 16:42 |
merlin1991 | also I did update here and no icon anywhere | 16:42 |
Pali | I tested it more times and it worked | 16:43 |
merlin1991 | DocScrutinizer05: the important part in the enabler postinst: http://paste.debian.net/189316/ | 16:44 |
DocScrutinizer05 | anyway, seems update went smoothly | 16:46 |
DocScrutinizer05 | and no icon | 16:46 |
DocScrutinizer05 | (at least not obvious so far) | 16:46 |
DocScrutinizer05 | merlin1991: "update all" first asked my about confirmation for installing cssu-t update | 16:48 |
DocScrutinizer05 | I concluded it would also try to *install* it first | 16:49 |
DocScrutinizer05 | and aborted to first install opername manually | 16:49 |
merlin1991 | tbh I have nfc how update all works, thus *should* | 16:49 |
merlin1991 | in theory if it does the dependency check with the other udpates in mind it should run | 16:49 |
DocScrutinizer05 | I had no mood to try what happens once I run into half-installed cssu-t that detects conflicts then. Since you also don't mention what will be the recoil and symptoms of such a conflict | 16:50 |
kerio | huh... what? | 16:51 |
DocScrutinizer05 | so we leave that for DAUs now | 16:51 |
kerio | why is 21.2011.38-1Tmaemo4-thumb16 a greater version than 21.2011.38-1Tmaemo5.1 | 16:51 |
kerio | ? | 16:51 |
kerio | :o | 16:51 |
merlin1991 | dafuq? | 16:51 |
jon_y | 16 > 5? | 16:51 |
jon_y | oh | 16:51 |
jon_y | t > T? | 16:52 |
kerio | merlin1991: that's what apt-cache policy tells me | 16:52 |
jon_y | ascii tables1 | 16:52 |
kerio | jon_y: try again | 16:52 |
freemangordon | :D | 16:52 |
freemangordon | could be because it is - not + | 16:52 |
jon_y | ah wait | 16:52 |
kerio | freemangordon: have you done magic to my apt | 16:52 |
jon_y | 4-thumb16 vs 5.1 | 16:53 |
freemangordon | kerio: tpt only? i was thinking it is cast on the whole device :S | 16:53 |
DocScrutinizer05 | straight silly compare was "-">"."? | 16:54 |
kerio | DocScrutinizer05: it's still 4 vs 5 | 16:54 |
kerio | before that | 16:54 |
DocScrutinizer05 | ooh yeah | 16:54 |
freemangordon | merlin1991: is -1Tmaemo5 correct? isn'n +1Tmaemo5 better? | 16:54 |
merlin1991 | hm dpkg --compare-versions 21.2011.38-1Tmaemo4-thumb16 lt 21.2011.38-1Tmaemo5.1 has an exit code of 1 | 16:55 |
merlin1991 | freemangordon: the -1 is part of the nokia version | 16:55 |
merlin1991 | we'd have to add the + before the T | 16:55 |
freemangordon | aah, yeah | 16:55 |
merlin1991 | though I need to check how that goes against the current version | 16:55 |
DocScrutinizer05 | honestly you want to update a thumb version to a plain T version? | 16:55 |
merlin1991 | btw adding the + doesn't either | 16:56 |
merlin1991 | fsckd version compare | 16:56 |
freemangordon | try without .1 | 16:56 |
kerio | DocScrutinizer05: there's no reason it shouldn't work :) | 16:56 |
DocScrutinizer05 | haha | 16:56 |
merlin1991 | freemangordon: still fsckd | 16:56 |
freemangordon | dafuq? | 16:56 |
DocScrutinizer05 | kerio: sure, the update will work flawlessly | 16:56 |
DocScrutinizer05 | just your system is fscked after that | 16:57 |
freemangordon | DocScrutinizer05: why? | 16:57 |
DocScrutinizer05 | hmm, indeed | 16:57 |
freemangordon | it will be slower and with less free RAM, but fsckd? | 16:57 |
DocScrutinizer05 | it's just de-thumbified, not fucked | 16:57 |
DocScrutinizer05 | until you get something into -thumb that's *needed* to make those parts work that aren't dethumbified | 16:58 |
DocScrutinizer05 | like enabling of thumb-patch in kernel during init | 16:58 |
kerio | nah, the workaround is enabled by default | 16:58 |
kerio | which is weird, but convenient | 16:59 |
DocScrutinizer05 | haha, maybe on your kernel | 16:59 |
kerio | it's Pali's kernel! | 16:59 |
DocScrutinizer05 | no, it's your kernel | 16:59 |
freemangordon | kerio: it's Titan's kernel :P | 16:59 |
kerio | k | 16:59 |
merlin1991 | anyway kerio just wait for freemangordon todo his update :) | 16:59 |
kerio | hm, i wonder if HAM will prompt the update | 16:59 |
freemangordon | merlin1991: we should fix that version bullshit | 17:00 |
DocScrutinizer05 | and no matter if system fscked or just de-thumbified, it's insane to roll back stuff to a basically lesser version | 17:00 |
freemangordon | maybe not now, but still | 17:00 |
freemangordon | DocScrutinizer05: :nod: | 17:01 |
DocScrutinizer05 | just to update to the new thumb version a few days later | 17:01 |
kerio | a few *days*? D: | 17:01 |
kerio | omg i gotta update | 17:01 |
freemangordon | kerio: what was that noise then? | 17:01 |
kerio | nah, just kidding | 17:01 |
freemangordon | yeah, verry funny | 17:02 |
freemangordon | very* | 17:02 |
DocScrutinizer05 | he's practicing trolling a lot lately | 17:02 |
kerio | HAM doesn't prompt the update either, anyway | 17:02 |
kerio | dpkg versions are silly | 17:02 |
freemangordon | DocScrutinizer05: and he's getting better | 17:02 |
DocScrutinizer05 | indeed | 17:02 |
merlin1991 | freemangordon: well personally I'd drop the nokia part in our version string, to simplify the whole thing, but I don't know if there aren't some mp-fremantle-* version comparisons in place somewhere | 17:02 |
DocScrutinizer05 | scary | 17:02 |
freemangordon | merlin1991: well, we'd better keep what is for -T and invent something new for -thumb | 17:03 |
DocScrutinizer05 | yoh | 17:03 |
merlin1991 | use +thumbX instead? | 17:04 |
freemangordon | instead what? | 17:04 |
DocScrutinizer05 | why? | 17:04 |
merlin1991 | instead of -thumbX | 17:04 |
kerio | if you do that, people won't be able to update | 17:04 |
kerio | because -thumbX will be greater than +thumbX | 17:04 |
merlin1991 | because the +whatever is compared differently | 17:04 |
DocScrutinizer05 | doesn't it work pretty nice right now, NOT "up"dating -thumb to lesser T version | 17:04 |
freemangordon | DocScrutinizer05: no, the ideas is new -T to be with higher version of latest -thumb | 17:05 |
DocScrutinizer05 | sorry? | 17:05 |
kerio | the weird thing is that it works fine with other packages | 17:05 |
freemangordon | in some critical fix needs to be distributed and I am unable to do it | 17:05 |
merlin1991 | freemangordon: to be more exact 5+whatever is bigger than 5 but never bigger than 6 for dpkg | 17:05 |
freemangordon | s/in/in case/ | 17:05 |
infobot | freemangordon meant: in case some critical fix needs to be distributed and I am unable to do it | 17:05 |
kerio | right now, apt is prompting me to update qt and obexd | 17:05 |
DocScrutinizer05 | and no matter if system fscked or just de-thumbified, it's insane to roll back stuff to a basically lesser version | 17:06 |
DocScrutinizer05 | (repost) | 17:06 |
freemangordon | DocScrutinizer05: the is why I issued a warning ont TMO "do not update to new -testing" :D | 17:06 |
merlin1991 | DocScrutinizer05: a -T update would never roll back to a lower version only to a version without thumb | 17:07 |
merlin1991 | DocScrutinizer05: that's why all the dependencies in our mp are >= | 17:07 |
DocScrutinizer05 | well, another way to handle that. Probably you want to replace cssu-t repo by cssu-thumb repo | 17:07 |
kerio | DocScrutinizer05: completely replace? ouch | 17:07 |
freemangordon | DocScrutinizer05: again, we invent that in case I got hit by a bus | 17:07 |
kerio | still doesn't solve the "critical bugfix" thing, tho | 17:08 |
DocScrutinizer05 | again | 17:08 |
DocScrutinizer05 | and no matter if system fscked or just de-thumbified, it's insane to roll back stuff to a basically lesser version | 17:08 |
freemangordon | whatever | 17:08 |
kerio | DocScrutinizer05: it's not a fucking lesser version if it doesn't have a critical bug | 17:08 |
DocScrutinizer05 | where "lesser" means "not thumbified" | 17:08 |
DocScrutinizer05 | kerio: eh?? | 17:08 |
merlin1991 | DocScrutinizer05: kerio talks about the case where there is an update for $package in -T which is arm but has some nasty bug fixed | 17:09 |
DocScrutinizer05 | so what? | 17:09 |
DocScrutinizer05 | I'm talking about the 'normal' case | 17:09 |
freemangordon | hmm, DocScrutinizer05 assumes -thumb an advantage? WTF, I FEEL constrained. wHAT DATE IS TODAY, i NEED THAT MARKED IN RED IN MY CALENDAR :p | 17:09 |
freemangordon | sorry for the caps | 17:10 |
kerio | freemangordon: you should've used +thumbY, but you haven't - how are you going to fix this? | 17:11 |
kerio | while still maintaining the ability to update from HAM for the non-techy users | 17:11 |
DocScrutinizer05 | I'm just waiting for the release that actually fucks up thumb systems, due to replacing something with a version that doesn't provide some mandatory thumb support or simply is so much bigger in size that it crashes in update on thumb systems | 17:11 |
kerio | DocScrutinizer05: wat? | 17:12 |
DocScrutinizer05 | ooh, btw, what will settings-aboutCSSU tell? XP | 17:13 |
kerio | flavour: testing, right now | 17:13 |
DocScrutinizer05 | ooh, for -thumb? *BAD* | 17:14 |
kerio | it is testing! | 17:14 |
DocScrutinizer05 | no it's not | 17:14 |
kerio | freemangordon: ooh, other packages have a +cssu or +0cssu i think, so the issue wouldn't come up | 17:14 |
kerio | it does come up for hildon-desktop, for instance, which has no +foo in the version | 17:16 |
kerio | 1:2.2.142-21.1-thumb1 is greater than 1:2.2.142-22 | 17:16 |
DocScrutinizer05 | I'm not going to support dumbass users by telling them "no matter what settings_About-CSSU shows! If you *ever* installed cssu-thumb, I can't help since that's not a proper cssu-t install, it has a lot of different binaries" | 17:16 |
kerio | freemangordon: this is going to be painful to fix, i'm afraid :c | 17:16 |
kerio | DocScrutinizer05: Version: 21.2011.38-1Tmaemo4-thumb16; Flavor: Testing | 17:17 |
* kerio files a bug for the wrong spelling of "Flavour" | 17:17 | |
DocScrutinizer05 | kerio: FUCK YOUr FLAVOUR! | 17:17 |
kerio | freemangordon: the supposed way to fix all these kinds of version mismatches is to increase the epoch | 17:18 |
DocScrutinizer05 | honestly... | 17:18 |
kerio | but you'd have to get merlin1991 to do that for -testing as well :s | 17:18 |
kerio | DocScrutinizer05: <3 | 17:19 |
merlin1991 | kerio: actually if the spelling in the widget is wrong it is a valid bug report | 17:19 |
kerio | DocScrutinizer05: see? :D | 17:19 |
DocScrutinizer05 | no, we won't mess with testing to accomodate -thumb requirements | 17:19 |
DocScrutinizer05 | kerio: stop trolling | 17:19 |
DocScrutinizer05 | you're extremely antagonizing | 17:20 |
merlin1991 | also kerio the spelling is right | 17:20 |
kerio | not in en_GB :) | 17:21 |
DocScrutinizer05 | I thought thumb repo got invented to decouple from trsting repo | 17:21 |
DocScrutinizer05 | kerio: stop trolling | 17:21 |
kerio | DocScrutinizer05: no, -thumb is supposed to match -testing | 17:21 |
DocScrutinizer05 | what a BS statement is that? | 17:21 |
kerio | it's supposed to be functionally equivalent to cssu-testing | 17:22 |
DocScrutinizer05 | so what? | 17:22 |
DocScrutinizer05 | ~dict decouple | 17:22 |
infobot | Dictionary 'decouple' (4): regard as unconnected; "you must dissociate these two events!"; "decouple our foreign policy from ideology" ;; reduce or eliminate the coupling of (one circuit or part to another) ;; eliminate airborne shock waves from (an explosive) ;; disconnect or separate; "uncouple the hounds" . | 17:22 |
kerio | they're very connected, though | 17:22 |
DocScrutinizer05 | BS | 17:22 |
DocScrutinizer05 | they are alternative | 17:23 |
DocScrutinizer05 | not complementary | 17:23 |
freemangordon | DocScrutinizer05: -thumb is supposed to follow -testing | 17:24 |
DocScrutinizer05 | you *can't* hope to pick _some_ essential bugfix updates from cssu-t repo while keeping other stuff on -thumb "level", whenever a new CSSU-T rolls out | 17:25 |
freemangordon | DocScrutinizer05: yes, you can | 17:25 |
DocScrutinizer05 | freemangordon: exactly, so why do you need cssu-t repo then? yeah the bus, BS | 17:25 |
*** arcean has joined #maemo-ssu | 17:25 | |
kerio | DocScrutinizer05: also, not everything is recompiled afaik | 17:25 |
freemangordon | DocScrutinizer05: because development happens in -T | 17:26 |
freemangordon | kerio: exactly | 17:26 |
DocScrutinizer05 | that's your probelm then, increase epoch of your -thumb stuff to avoid downgrading to cssu-t on new rev | 17:26 |
freemangordon | DocScrutinizer05: but that is intended behaviour | 17:27 |
freemangordon | it is not done by mistake | 17:27 |
kerio | freemangordon: otoh, the weird versioning thing is a mistake :c | 17:27 |
DocScrutinizer05 | shall I post it *again*?? | 17:27 |
freemangordon | DocScrutinizer05: no need | 17:27 |
freemangordon | kerio: yes, but we will fix that somehow | 17:28 |
DocScrutinizer05 | oh well, I simply don't care about that -thumb stuff, too weird | 17:28 |
freemangordon | DocScrutinizer05: but your reasoning is not necessarily valid for everyone | 17:28 |
DocScrutinizer05 | no concept how to maintain | 17:28 |
DocScrutinizer05 | so do whatever you like, on -thumb | 17:29 |
kerio | now look who's antagonizing | 17:29 |
freemangordon | and that is why newer -T is supposed to have bigger version when out. because I am not the best debian maintainer around and could easily screw something | 17:30 |
DocScrutinizer05 | just please save me from comments like "Then we need to increase epoch on all stuff in -testing" or similar nonsense | 17:30 |
kerio | DocScrutinizer05: no, i said that you'd have to do that to use the standard fix for this kind of thing | 17:31 |
freemangordon | when did I say that? | 17:31 |
DocScrutinizer05 | kerio did | 17:31 |
kerio | notice the... conditional tense? | 17:31 |
kerio | conditional form? | 17:31 |
freemangordon | but kerio can do nothing in -thumb, i miss your point | 17:31 |
DocScrutinizer05 | stfu kerio | 17:31 |
freemangordon | DocScrutinizer05: calm down, ok? | 17:32 |
DocScrutinizer05 | my point? we just updated testing and are discussing -thumb since 1h now | 17:32 |
DocScrutinizer05 | no feedback whatsoever about testing update form any of you | 17:32 |
freemangordon | DocScrutinizer05: I wouldn't update my primary device while at work | 17:33 |
kerio | i only have one n900 :c | 17:33 |
DocScrutinizer05 | so you got my point now? | 17:33 |
DocScrutinizer05 | if not I don't mind either | 17:33 |
DocScrutinizer05 | cya peeps | 17:33 |
kerio | DocScrutinizer05: oh i'm sorry, i didn't know me and freemangordon were the only ones in here who can test the new -T release | 17:35 |
DocScrutinizer05 | I'm temped to kick you for continued trolling despite warnings | 17:35 |
* Lava_Croft whistles a familiar tune | 17:36 | |
freemangordon | DocScrutinizer05: kicking everyone out of the channel wouldn't make CSSU better or more tested. Though it is your right being the chanop | 17:36 |
Lava_Croft | freemangordon recognized the tune! | 17:36 |
Lava_Croft | and fast too! | 17:36 |
DocScrutinizer05 | freemangordon: you try to join kerio's trollig efforts? That will still not cause me kicking *everybody* | 17:37 |
* freemangordon kicks himself on his way home, bbl | 17:37 | |
Lava_Croft | one day DocScrutinizer05 is all about how we should be friendly since we all like hacking our lovely n900 | 17:38 |
Lava_Croft | the next day he is all about kicking folks | 17:38 |
DocScrutinizer05 | I didn't know kerio was "folks" | 17:39 |
Lava_Croft | you already basically mentioned two | 17:39 |
DocScrutinizer05 | check your definition of "basically" | 17:39 |
Lava_Croft | but im smart enough to know that having any kind of discussion with you when you are in this mood is fruitless | 17:39 |
Lava_Croft | take a hike, get some fresh air | 17:39 |
Lava_Croft | it will do you good! | 17:39 |
Lava_Croft | (unless you live in some big city with shit air quality) | 17:40 |
DocScrutinizer05 | I tried that when kerio thought he needs to highlight me | 17:40 |
DocScrutinizer05 | with further trolling nonsense | 17:40 |
Lava_Croft | ignore hilights for a while | 17:40 |
Lava_Croft | and dont make em beep | 17:40 |
Lava_Croft | anyhow, you get the point | 17:41 |
*** M4rtinK has quit IRC | 17:41 | |
Lava_Croft | and im fairly certain kerio got it:) | 17:41 |
kerio | yeah, i'll just shut up when ops with PMS are around | 17:41 |
* Lava_Croft sighs | 17:41 | |
Lava_Croft | silly child | 17:41 |
Lava_Croft | I basically defended you for a little bit, but you waste no time in proving me wrong | 17:42 |
Lava_Croft | I should have known better | 17:42 |
*** toxaris has joined #maemo-ssu | 18:03 | |
*** BCMM has quit IRC | 18:11 | |
*** toxaris has quit IRC | 18:13 | |
*** toxaris has joined #maemo-ssu | 18:14 | |
Pali | please test patched ke-recv which adds support for configuring usbnetwork (by default all disabled) | 18:26 |
Pali | https://gitorious.org/~pali/community-ssu/pali-ke-recv | 18:26 |
*** arcean_ has joined #maemo-ssu | 18:30 | |
*** arcean has quit IRC | 18:39 | |
merlin1991 | Pali: you need one more thing taken care of in the script | 18:45 |
merlin1991 | dnsmasq shouldn't give a gateway ip if nat is disabled | 18:46 |
kerio | Pali: put it in -devel | 18:58 |
Pali | I do not want to start qemu | 18:58 |
Pali | scripts was not tested yet | 18:58 |
*** Woody14619 has quit IRC | 18:59 | |
*** Woody14619 has joined #maemo-ssu | 18:59 | |
*** Woody14619 has quit IRC | 18:59 | |
*** Woody14619 has joined #maemo-ssu | 18:59 | |
*** arcean_ is now known as arcean | 19:08 | |
DocScrutinizer05 | Pali: >>/usr/sbin/iptables -t nat -D POSTROUTING ! -o lo -j MASQUERADE<< ?? "!" really? not "1"? | 19:13 |
DocScrutinizer05 | Pali: -D, --delete chain rulenum | 19:13 |
DocScrutinizer05 | Pali: I'm no expert, but looks like a typo | 19:14 |
Pali | correct is also: -D chain rule-specification | 19:15 |
DocScrutinizer05 | ooh I see | 19:15 |
DocScrutinizer05 | should've finished manpage all way down first :-) | 19:15 |
*** NIN101 has joined #maemo-ssu | 19:15 | |
DocScrutinizer05 | [!] -o, --out-interface name | 19:16 |
Pali | maybe we can limit masquerading only for usb0 | 19:16 |
Pali | but we can use usbnetwork for 1) connect N900 to PC with NAT or 2) connect PC to N900 | 19:18 |
Pali | also more iptables rules for security should be added | 19:18 |
*** RST38h has quit IRC | 19:18 | |
Pali | but n900 has disabled iptables, so is without firewall by default... | 19:19 |
Pali | so additional rules depends on users... | 19:19 |
kerio | is there no kernel support or it's just the userland missing? | 19:19 |
Pali | kerio, missing iptables package (is only in extras) and I do not know if iptables is enabled in default kernel | 19:20 |
Pali | kernel-power has it included | 19:20 |
Pali | DocScrutinizer05, I pushed package iptables-persistent from ubuntu which run iptables-restore < /etc/iptables/<some_file> if exists | 19:21 |
Pali | so confguration is now possible with one config file | 19:21 |
Pali | (or two for ipv6) | 19:21 |
DocScrutinizer05 | no iptables in stock kernel afaik | 19:22 |
kerio | i think it's probably one of the things worth including if CSSU eventually decides to do a kernel update | 19:22 |
DocScrutinizer05 | sidenote: if [ $? != 0 ]; then ## better use [$? -ne 0] since it's basically numeric, but I see it's already like that in original script | 19:22 |
DocScrutinizer05 | another sidenote: if [ "$USBNETWORK_DNSMASQ" = "1" ]; then ### better use a function call, like >> if config-set USBNETWORK_DNSMASQ; then << | 19:25 |
DocScrutinizer05 | config-set(){ if [ "$1" = "1" -o "$1" = "true" -o "$1" = "yes" -o "$1" = "y"]; then return 0; fi; return 1} | 19:29 |
DocScrutinizer05 | sorry | 19:29 |
DocScrutinizer05 | s/$1/${$1}/ | 19:30 |
DocScrutinizer05 | s/$1/${$1}/g | 19:30 |
*** M4rtinK has joined #maemo-ssu | 19:44 | |
*** toxaris has quit IRC | 20:10 | |
*** Free-MG has joined #maemo-ssu | 20:13 | |
kerio | why does community-testing hold so many past versions of packages? | 20:19 |
kerio | same goes for the extras repos | 20:19 |
amiconn | Hmm, new cssu-t | 20:23 |
* amiconn wonders whether it has the fix for this %$&/§"% operator name widget | 20:24 | |
amiconn | Hmm, no offline mode + reboot? | 20:27 |
amiconn | Is this expected? | 20:28 |
kerio | amiconn: the new operator name widget worksforme | 20:34 |
kerio | and i had problems with the older version | 20:34 |
amiconn | The old version had a fairly big bug. Unfortunately I can't even verify right now - the bug only hit me when roaming | 20:36 |
kerio | yep, same bug i was affected by | 20:36 |
kerio | :) | 20:36 |
Pali | roaming is fixed | 20:36 |
kerio | it's kinda easy for me though, i have to go roaming when in 2g | 20:36 |
Pali | I fixed problem when operator name was empty and when not changed in roaming | 20:37 |
Pali | also added handler for dbus signal "operator changed" | 20:37 |
kerio | Pali: there's a weird log line that's not disabled, which makes me a sad panda :( | 20:39 |
kerio | Pali: "service provider name match" | 20:39 |
Pali | where is written? | 20:40 |
Pali | ok, I see | 20:41 |
Pali | kerio, fixed: https://gitorious.org/community-ssu/operator-name-cbs-widget/commit/ff96ce3653b9951158badc150cb4c62aec28b1fc | 20:44 |
kerio | =D | 20:44 |
DocScrutinizer05 | damn, sure we already fixed that, and kerio should know damn well | 21:26 |
kerio | i always said it wasn't fixed yet! | 21:26 |
DocScrutinizer05 | amiconn: yes, fixed opername applet in new CSSU, see changelog! | 21:27 |
DocScrutinizer05 | kerio: you've actually been asked to test the fix iirc | 21:28 |
DocScrutinizer05 | amiconn: reboot happened here, please see http://talk.maemo.org/showpost.php?p=1264791&postcount=793 | 21:29 |
kerio | DocScrutinizer05: i don't have a working toolchain, i'd need binaries to do that | 21:29 |
*** RST38h has joined #maemo-ssu | 21:31 | |
amiconn | Hmm, "About CSSU" still says Tmaemo4.1, even after manual reboot | 21:36 |
DocScrutinizer05 | amiconn: you're supposed to install opername applet first, then both cssu-t and cssu enabler | 21:37 |
DocScrutinizer05 | in HAM | 21:37 |
amiconn | ham offered both opername applet and cssu enabler, I chose "update all" as usual | 21:37 |
merlin1991 | cssu-t is enough it will pull in the enabler update aswell | 21:37 |
merlin1991 | amiconn: ham is stupid, do hte opername applet first :D | 21:38 |
DocScrutinizer05 | amiconn: thanks for confirming it doesn't work | 21:38 |
DocScrutinizer05 | merlin1991: please again edit #793 | 21:38 |
amiconn | Now there are "no updates available" | 21:39 |
DocScrutinizer05 | since everybody would just pick "update all" | 21:39 |
amiconn | Why are dependencies on maemo f**ed up like that? | 21:39 |
DocScrutinizer05 | Nokia invented mega-MP | 21:40 |
gregoa | the dependencies are fine, it seems that HAM is silly | 21:40 |
kerio | that's an understatement :) | 21:40 |
DocScrutinizer05 | no, actually dependencies are lazy/sloppy | 21:40 |
merlin1991 | DocScrutinizer05: edited | 21:40 |
merlin1991 | amiconn: because ham uses apt worker and udate all is just a nice way to say do those all in single steps | 21:41 |
DocScrutinizer05 | Nokia decided to just go "screwit, it's all in MP anyway, so why get *real* dependencies right?" | 21:41 |
merlin1991 | which will fail if the cssu update goes first | 21:41 |
gregoa | well, "aptitude dist-upgrade" did what I expected it to do (i.e. remove the opername package all by itself. in one run.) | 21:41 |
DocScrutinizer05 | OUCH | 21:41 |
* amiconn does not understand why Nokia tried to reinvent the wheel (but made it worse) instead of just relying on apt and one of the numerous existing front-ends (maybe with a custom touch ui) | 21:44 | |
kerio | because fukken Elop | 21:44 |
kerio | or something like that | 21:44 |
amiconn | That was way before Elop | 21:44 |
kerio | i know :) | 21:44 |
kerio | proper dependencies are actually kinda hard to maintain | 21:44 |
kerio | and i think they wanted to control the updates released to the users | 21:45 |
kerio | nokia phones are upgraded in bulk | 21:45 |
gregoa | apt worker updating packages one by one is a different question than dependencies | 21:45 |
*** luf has joined #maemo-ssu | 21:46 | |
wmarone | numerous existing front-ends? | 21:47 |
kerio | wmarone: synaptic | 21:47 |
DocScrutinizer05 | btw prepare for shitstorm, as most users simply will click "update all2 nevertheless, when they get notification [!] | 21:47 |
wmarone | kerio: OH ok, I thought you meant GUI | 21:47 |
wmarone | as in the entire UI, not package manager | 21:48 |
kerio | DocScrutinizer05: well, it upgrades fine if you do "Update all" twice, doesn't it? | 21:48 |
DocScrutinizer05 | nfc | 21:48 |
amiconn | kerio: Nope, at least not here | 21:48 |
kerio | :( | 21:48 |
kerio | meh, at least it's testing and not stable | 21:48 |
DocScrutinizer05 | meh, at least meh | 21:49 |
* DocScrutinizer05 prepares for nightshift | 21:49 | |
RST38h | Nightshift where | 21:49 |
RST38h | Here? | 21:50 |
DocScrutinizer05 | sure | 21:51 |
DocScrutinizer05 | and on friggin tmo | 21:52 |
amiconn | Why is this ham thing so darn slow? | 21:52 |
kerio | amiconn: because fukken Elop, or something like that :P | 21:52 |
DocScrutinizer05 | because it's building dependencies | 21:52 |
DocScrutinizer05 | something other fap apps don't | 21:52 |
amiconn | apt-get dist-upgrade would be faster, wouldn't it? | 21:52 |
kerio | amiconn: eeh, never properly worked for me | 21:53 |
kerio | (on maemo) | 21:53 |
RST38h | You are still trolling tmo? | 21:53 |
RST38h | Nooooo | 21:53 |
DocScrutinizer05 | the question is: why does it still fsck up when there's a conflict that can get solved by simply installing in right sequence? | 21:53 |
amiconn | Not necessarily on maemo, but on debian with similarly specced hardware | 21:53 |
kerio | amiconn: it's not a matter of hardware :) | 21:53 |
kerio | and yeah, apt-get is much faster | 21:53 |
amiconn | Seems it's finally upgrading cssu properly | 21:54 |
DocScrutinizer05 | amiconn: apt-get dist-upgrade is *DEPRECATED* | 21:54 |
DocScrutinizer05 | known to fuck up system | 21:54 |
amiconn | [20:53:39] <amiconn> Not necessarily on maemo, but on debian with similarly specced hardware | 21:54 |
kerio | actually i think it's kinda deprecated on debian too | 21:54 |
gregoa | no | 21:55 |
kerio | you're supposed to use aptitude for big updates that apt-get can't get through | 21:55 |
amiconn | Nope | 21:55 |
kerio | (or apt-get install manually) | 21:55 |
gregoa | debian changed recommendations from apt-get to aptitude and back to apt-get, because at different points in time one or the other was slightly more intellignet | 21:55 |
amiconn | For Debian 5.0 -> 6.0 they *recommend* using apt-get instead of aptitude | 21:55 |
*** wmarone_ has joined #maemo-ssu | 21:56 | |
DocScrutinizer05 | on maemo all intelligence is doomed to fail | 21:56 |
DocScrutinizer05 | thanks Nokia and their inventions | 21:56 |
kerio | because of the bigass metapackage? | 21:56 |
amiconn | Not entirely automatic dist-upgrade though, but several steps. First a plain 'upgrade', then install current kernel+udev, reboot, and only then a dist-upgrade | 21:56 |
merlin1991 | hm I did dist-upgrade from 5 to 6 | 21:57 |
merlin1991 | worked fine | 21:57 |
*** NIN101 has quit IRC | 21:57 | |
*** gri has quit IRC | 21:57 | |
*** wmarone has quit IRC | 21:57 | |
*** infobot has quit IRC | 21:57 | |
*** Sicelo has quit IRC | 21:57 | |
merlin1991 | DocScrutinizer05: what was the pkg name of hte fsckd ovi sms msg shit? | 21:57 |
DocScrutinizer05 | fine! did you also evaluate it? | 21:57 |
* amiconn usually follows their upgrade instructions | 21:57 | |
DocScrutinizer05 | OOOONOES infobot ;-( | 21:57 |
Pali | fedora will have offline updates :D | 21:58 |
Pali | thanks lennart! | 21:58 |
*** NIN101 has joined #maemo-ssu | 21:58 | |
*** gri has joined #maemo-ssu | 21:58 | |
*** Sicelo has joined #maemo-ssu | 21:58 | |
* amiconn wonders why ham temporarily forgot all its catalogs | 21:59 | |
amiconn | Restarting ham fixed that | 21:59 |
NIN101 | yay. | 21:59 |
* kerio looks at amiconn | 21:59 | |
Pali | where is infobot? | 21:59 |
kerio | amiconn: is this really the first time you used HAM? | 21:59 |
amiconn | nope | 21:59 |
merlin1991 | Pali: I tried you u-boot the other day and whilst it worked fine to boot the nemo image on my sd card, the attached kernel didn't boot maemo anymore | 22:00 |
merlin1991 | afaik it should have the p1.3 kernel attached? | 22:00 |
Pali | u-boot-flasher has attached pr1.3 kernel | 22:00 |
amiconn | Ah, and the usual bt-disabled-after-cssu-update :\ | 22:00 |
* DocScrutinizer05 dances a jig for his new magic revive powers | 22:00 | |
merlin1991 | Pali: any idea why it would refuse to boot maemo then? | 22:00 |
Pali | no idea, I did not have any problems... | 22:01 |
Pali | merlin1991, try to use framebuffer kernel | 22:01 |
*** infobot has joined #maemo-ssu | 22:01 | |
*** ChanServ sets mode: +v infobot | 22:01 | |
DocScrutinizer05 | infobot: wb | 22:01 |
infobot | thx | 22:01 |
merlin1991 | Pali: does the pkg with the various kernels contain that? | 22:01 |
Pali | I think yes | 22:02 |
Pali | it is only tar.gz | 22:02 |
merlin1991 | anything I can look for especially? | 22:02 |
Pali | try boot framebuffer kernel and see where it failing... | 22:03 |
merlin1991 | okay | 22:05 |
* DocScrutinizer05 blames modules | 22:05 | |
merlin1991 | first I need to find my sd card :D so I know which device I tried it on :D | 22:05 |
merlin1991 | DocScrutinizer05: it is a stock kernel on a device with only stock kernel modules | 22:05 |
DocScrutinizer05 | hmm | 22:05 |
DocScrutinizer05 | strange then | 22:05 |
merlin1991 | now where was that u-boot thread again D: | 22:06 |
DocScrutinizer05 | merlin1991: we *might* have postponed cssu-t update by 24h or more, to allow all users to update opername first | 22:12 |
DocScrutinizer05 | now I think we should have a story at hand about how to recover | 22:12 |
merlin1991 | Pali: it fails to mount the root fs cannot open "ubi0:rootfs erorr -19 | 22:13 |
Pali | I need more messages | 22:13 |
merlin1991 | how can I keep it from shutting down so fast after the kernel panic? | 22:13 |
DocScrutinizer05 | >>ovi sms msg shit?<< notmynokia | 22:13 |
DocScrutinizer05 | cherry | 22:14 |
Pali | merlin1991, enable R&D mode + disable watchdogs | 22:14 |
merlin1991 | okay | 22:14 |
Pali | merlin1991, I think somebody already wrote problem that rootfs cannot be mounted | 22:14 |
DocScrutinizer05 | ~jrtools | 22:14 |
infobot | somebody said jrtools was http://wiki.maemo.org/User:Joerg_rw/tools | 22:14 |
DocScrutinizer05 | ^f cherry | 22:15 |
merlin1991 | Pali: what is the disable watchdog argument for flasher? | 22:15 |
Pali | $ flasher-3.5 --enable-rd-mode --set-rd-flags=no-omap-wd | 22:15 |
DocScrutinizer05 | flasher --set-rd-flags | 22:16 |
DocScrutinizer05 | will show a list | 22:16 |
Pali | merlin1991, can you upload your parsed /proc/atags (booted without u-boot) | 22:16 |
DocScrutinizer05 | no-omap-wd no-ext-wd no-lifeguard-reset | 22:16 |
merlin1991 | I'll get them together with the framebuffer msgs | 22:17 |
DocScrutinizer05 | omap-wd is hw-wd on omap, ext-wd is on GAIA, lifeguard is basically dsme, aiui | 22:17 |
Pali | yes, correct | 22:18 |
merlin1991 | is it a comma seperated list? | 22:18 |
Pali | yws | 22:18 |
DocScrutinizer05 | or maybe not even GAIA but bb5? | 22:18 |
DocScrutinizer05 | yep, afaik | 22:18 |
Pali | DocScrutinizer05, in kernel are only: CONFIG_OMAP_WATCHDOG=m CONFIG_TWL4030_WATCHDOG=m | 22:19 |
DocScrutinizer05 | at least I recently noticed the bb5 can reset omap and of course omap can reset bb5 | 22:19 |
DocScrutinizer05 | I guess you can't disable bb5 wd | 22:20 |
DocScrutinizer05 | just shut down cmt completely | 22:20 |
DocScrutinizer05 | or maybe it's not even used in fremantle | 22:21 |
merlin1991 | Pali: I guess I could test on my other devices aswell to see if it is a hw specific thing | 22:21 |
Pali | merlin1991, ok | 22:21 |
Pali | merlin1991, also provide atags, see thread: http://talk.maemo.org/showthread.php?t=84560 | 22:21 |
DocScrutinizer05 | merlin1991: amiconn: what's the proper recovery path from updating cssu-t5 first conflicts? | 22:23 |
amiconn | It seems that one just needs to update again. This didn't work for me first time because of ham hiccup | 22:24 |
merlin1991 | DocScrutinizer05: install oper name, do update | 22:25 |
Pali | and then remove operator name | 22:25 |
DocScrutinizer05 | [2012-09-11 20:39:02] <amiconn> Now there are "no updates available" | 22:25 |
Pali | wait | 22:25 |
Pali | HAM will show update after 2-3 minutes after clicking on button | 22:25 |
DocScrutinizer05 | [2012-09-11 20:48:27] <kerio> DocScrutinizer05: well, it upgrades fine if you do "Update all" twice, doesn't it? [2012-09-11 20:48:38] <amiconn> kerio: Nope, at least not here | 22:26 |
amiconn | [20:59:10] -*- amiconn wonders why ham temporarily forgot all its catalogs | 22:26 |
amiconn | [20:59:22] <amiconn> Restarting ham fixed that | 22:26 |
merlin1991 | amiconn: when apt-worker is still busy when you hit the catalouges button then you get an emtpy list after a timeout | 22:27 |
Pali | problem is that after each operation ham will parse all package lists and this take a lot of time | 22:27 |
DocScrutinizer05 | ok, so the recommendation is: quit ham, start ham, do update opername, quit ham, start ham, do cssu-t update | 22:27 |
Pali | second problem is with apt-worker | 22:27 |
merlin1991 | just run xterm in the background with top, when you see that apt-worker isn't hogging the cpu anymore you should be able to get the list | 22:27 |
Pali | if apt worked crash or fail, HAM stop working... | 22:27 |
merlin1991 | fsckd battery for the rtc | 22:28 |
Pali | if apt-worker freeze then HAM not working after starting again too | 22:28 |
DocScrutinizer05 | ok, so the recommendation is: quit ham, wait 5 min, start ham, do update opername, quit ham, wait 5 min, start ham, do cssu-t update | 22:28 |
Pali | it is possible to charge rtc battery? | 22:28 |
merlin1991 | I really need a set of those batteries, I'm getting tired of setting the time over and over again | 22:28 |
merlin1991 | DocScrutinizer05: the wait 5 min is after start ham | 22:28 |
Pali | merlin1991, use operator time | 22:28 |
merlin1991 | Pali: operator does not provide time :/ | 22:29 |
merlin1991 | and ntp needs at least some kind of proper date in order to work | 22:29 |
Pali | use gps for time :D | 22:29 |
DocScrutinizer05 | haha | 22:29 |
merlin1991 | i have nothing installed that does that | 22:29 |
DocScrutinizer05 | a-gps needs time to work | 22:29 |
merlin1991 | though it would be the next best option | 22:29 |
Pali | DocScrutinizer05, how agps working? | 22:30 |
DocScrutinizer05 | +-10min maybe | 22:30 |
Pali | it is only SW solution? | 22:30 |
kerio | merlin1991: find a way to install ntpdate | 22:30 |
Pali | or needs also some HW? | 22:30 |
DocScrutinizer05 | basically yes | 22:30 |
merlin1991 | Pali: do you need just a cat /proc/atags or do I need to run your funky binary? | 22:30 |
merlin1991 | kerio: like I said, ntp needs at least some sort of right date | 22:30 |
Pali | merlin1991, /proc/atags is binary | 22:30 |
kerio | that's why i said ntpdate | 22:30 |
Pali | so if you can send me unmodified binary, it is ok | 22:30 |
DocScrutinizer05 | merlin1991: ntp has an option to override the "less than a hour" restriction | 22:31 |
DocScrutinizer05 | and also ntpd now is the official replacement for ntpdate | 22:31 |
Pali | $ cat file | xxd | 22:32 |
Pali | $ uuencode -m file file | 22:32 |
merlin1991 | DocScrutinizer05: I'm using rdate (which talks ntp) but when your own date is wayyyy too wrong then it fails when calculating the timing difference to the ntp server | 22:32 |
DocScrutinizer05 | xxd: no binary | 22:32 |
Pali | or something else... | 22:32 |
kerio | DocScrutinizer05: i thought that the option that made ntpd work like ntpdate wasn't in maemo ntpd | 22:32 |
merlin1991 | Pali: I still have the atags binary around | 22:33 |
Pali | DocScrutinizer05, xxd is used by vim | 22:33 |
Pali | :%!xxd | 22:33 |
Pali | convert current file to binary | 22:33 |
Pali | :%!xxd -r | 22:33 |
Pali | convert current hexdump back | 22:33 |
Pali | merlin1991, or check if atags are still same as your reported: http://talk.maemo.org/showpost.php?p=1214052&postcount=14 | 22:35 |
Pali | and which is problematic | 22:36 |
Pali | merlin1991, can you try to boot rescue os via u-boot and copy atags from rescue os? | 22:36 |
merlin1991 | Pali: atags are still the same, they only differ in BOOTREASON (pwr key vs sw reset) | 22:37 |
Pali | ok | 22:37 |
Pali | I will write bootmenu item for rescue os | 22:37 |
Pali | ~rescueos | 22:37 |
infobot | [rescue-os] http://206.253.166.96/N900/rescueOS/ | 22:37 |
merlin1991 | my errors are here: http://paste.debian.net/189406/ | 22:38 |
DocScrutinizer05 | inde3ed friggin opennntpd doesn't know -g | 22:38 |
DocScrutinizer05 | I'm pretty sure I got ntpdate on T900, wonder where from | 22:39 |
Pali | merlin1991, here is rescue os item file for u-boot: http://talk.maemo.org/showthread.php?p=1264989#post1264989 | 22:42 |
Pali | rescue os not using ubifs rootfs, so it should boot | 22:43 |
merlin1991 | Pali: where should I drop the rescuefs files? | 22:43 |
Pali | internal or external mmc | 22:43 |
Pali | mydocs is ok | 22:43 |
merlin1991 | the other n900 booted fine with u-boot | 22:44 |
Pali | just update file names, filesystem and mmc partittion in config file | 22:44 |
Pali | which HW rev is OK and which incorrect? | 22:44 |
DocScrutinizer05 | dman, tmo users are not thanks-happy on [Announce] CSSU Testing thread | 22:46 |
merlin1991 | 2101 is ok | 22:46 |
DocScrutinizer05 | ooooh, hv-rev incompatibilities? :-) | 22:46 |
DocScrutinizer05 | prolly due to different nand types on different hw-revs | 22:47 |
merlin1991 | arf now my n900 is silly | 22:49 |
merlin1991 | 2104 fails | 22:50 |
kerio | merlin1991: oh noes D: | 22:50 |
* kerio has a 2104 | 22:50 | |
merlin1991 | hm I also have a revision 2204 device here, I'll try how that one does | 22:51 |
merlin1991 | ewww I'll have to reflash this one beforehand | 22:53 |
Pali | seems that your error is same as somebody reported | 22:58 |
*** Free-MG has quit IRC | 23:00 | |
kerio | Pali: regarding different hw revs? or regarding 2104? | 23:02 |
Pali | somebody reported same problem as merlin | 23:02 |
merlin1991 | Pali: where on Mydocs do I have to drop the kernel image, inside bootmenu.img.d? | 23:05 |
Pali | merlin1991, drop it where you want | 23:07 |
Pali | just adjust path in config file | 23:07 |
Pali | in config file must be full path without leading '/' | 23:07 |
merlin1991 | err the question is does it assume the bootmenu.img.d path by default or not? | 23:07 |
Pali | no | 23:08 |
Pali | do not use this bootmenu.img.d | 23:08 |
merlin1991 | what is it? :D | 23:08 |
Pali | this is for autogenerated files | 23:08 |
merlin1991 | and why does it contain the stock and the fb kernel, and how the fuck do their item files work then? | 23:08 |
Pali | and I forgot that you need to generate uimage and uinitrd from rescue os files | 23:09 |
merlin1991 | ... how? | 23:09 |
Pali | u-boot can load only signed images with its types | 23:09 |
Pali | merlin1991, or better, drop these files in /boot/ | 23:09 |
*** andre__ has quit IRC | 23:09 | |
Pali | and remove FSTYPE and MMC drive lines | 23:09 |
Pali | u-boot-update-bootmenu will then sign it | 23:10 |
merlin1991 | okay | 23:10 |
Pali | and generate files to bootmenu.img.d | 23:10 |
Pali | post on TMO was edited | 23:11 |
Pali | u-boot binary which is flashed to nand contains u-boot binary and attached pr1.3 kernel | 23:12 |
Pali | so nolo load both to RAM and start first image (u-boot). u-boot then find kernel and can boot it | 23:13 |
DocScrutinizer05 | no real news | 23:13 |
Pali | then uboot load script file bootmenu.cmd from MyDocs and execute it (if exists) | 23:13 |
Pali | this script is generated by u-boot-update-bootmenu and contains additional entries for other kernels/OSs | 23:14 |
Pali | u-boot-update-bootmenu will parse config files and if ITEM_DEVICE is not specified it copy file from /boot to MyDocs/bootmenu.img.d and sign it | 23:15 |
merlin1991 | hm how do I get the atags out of there? | 23:15 |
Pali | in rescue os you can enable usb networking | 23:15 |
Pali | and login via telnet | 23:16 |
Pali | also you can mount MyDocs | 23:16 |
Pali | cat /proc/atags > ... mydocs... | 23:16 |
merlin1991 | no I can't same error as when I boot pr kernel from u-boot | 23:16 |
Pali | and via usb mass storage you can copy file out | 23:16 |
DocScrutinizer05 | dafaq, /me needs some breakfast | 23:17 |
DocScrutinizer05 | *before* nightshift | 23:17 |
Pali | merlin1991, is rescue os working from u-boot? | 23:17 |
merlin1991 | yes it booted | 23:17 |
merlin1991 | hm I have usb networking up, but no ssh server? | 23:18 |
Pali | telnet | 23:18 |
kerio | merlin1991: why are you doing this in rescue-os? | 23:19 |
*** toxaris has joined #maemo-ssu | 23:19 | |
Pali | because if atags are fsck then only rescue os can show it | 23:19 |
kerio | i see | 23:19 |
Pali | because in atags are stored mtd layout (for maemo) | 23:19 |
Pali | but upstream kernel has hardcoded mtd layout | 23:19 |
Pali | and does not depends on atags | 23:19 |
Pali | ok, now I'm going offline | 23:20 |
Pali | bye | 23:20 |
kerio | where are the atags written? | 23:20 |
kerio | aww :( | 23:20 |
Pali | /proc/atags | 23:20 |
merlin1991 | hm I can telent in but nothing happens when I type on the telnet console | 23:20 |
kerio | no, i mean | 23:20 |
kerio | where does the kernel get those? | 23:20 |
Pali | from RAM | 23:20 |
kerio | i see | 23:20 |
Pali | bootloader will store atags to RAM to well known address | 23:21 |
*** Pali has quit IRC | 23:21 | |
DocScrutinizer05 | http://talk.maemo.org/showthread.php?p=1265003#post1265003 | 23:22 |
DocScrutinizer05 | damn, how can pali vanish the very 30s you need him? | 23:23 |
kerio | DocScrutinizer05: haha | 23:23 |
kerio | also no, it's a bb-p problem | 23:23 |
DocScrutinizer05 | orly? | 23:23 |
DocScrutinizer05 | cooooool | 23:23 |
DocScrutinizer05 | isn't bb-p the latest hottest shit around? | 23:24 |
kerio | yep | 23:24 |
kerio | the only issue is updating bb-p itself | 23:24 |
kerio | not sure why it's related to cssu, that one | 23:24 |
DocScrutinizer05 | toldya don't mess with system shell | 23:24 |
kerio | unless that dude also upgraded busybox-power at teh same time | 23:25 |
kerio | DocScrutinizer05: if i told you that with a proper replacement for the busybox package there would be no such problem, would you tell me to stop trolling? | 23:25 |
kerio | right now the packaging for bb-p is a kludge | 23:25 |
DocScrutinizer05 | I'd tell you "read chanlog, it's exactly what I suggested for a clean solution of this mess" | 23:26 |
DocScrutinizer05 | ooh, actually not | 23:26 |
DocScrutinizer05 | since the "proper replacement" actually is an independent installation of bb-p on /opt | 23:27 |
luf | kerio: stop trolling :D :D | 23:27 |
kerio | i meant a conflicts+provides for "busybox", but yeah | 23:27 |
DocScrutinizer05 | the problem now is that bb-p actually replaces bb | 23:28 |
DocScrutinizer05 | without any sane rationale why we can't do better | 23:28 |
kerio | the problem is that it doesn't actually replace it, but it uses preinst/postinst scripts to kludge itself in | 23:29 |
DocScrutinizer05 | so cssu-t bb update obviously is prone to run into trouble | 23:29 |
DocScrutinizer05 | :nod: | 23:29 |
kerio | and it also does a weird thing with the symlinks to the busybox functions | 23:29 |
kerio | bb-p provides stuff like bzip2 | 23:29 |
kerio | but it doesn't actually make the link if you have it installed already | 23:30 |
kerio | i have no fscking idea of what happens if you install bzip2 *after* busybox-power | 23:30 |
merlin1991 | kerio: nothing special, the bzip2 binary replaces the bb-power symlink | 23:31 |
merlin1991 | DocScrutinizer05: we had a discussion with the bb-power maintainer about this some time ago | 23:31 |
merlin1991 | users of bb-power have to reinstall bb-power in order to get it back | 23:31 |
merlin1991 | but there is no way we can go around this due to the hacky postinst replacing of busybox bb-power does | 23:31 |
kerio | merlin1991: actually, a dpkg diversion with an exception for the busybox-power package itself would work much better | 23:32 |
DocScrutinizer05 | yep, but we got new problems | 23:32 |
kerio | it wouldn't help for the extra symlinks though | 23:32 |
DocScrutinizer05 | http://talk.maemo.org/showthread.php?p=1265003#post1265003 | 23:32 |
*** toxaris has quit IRC | 23:33 | |
DocScrutinizer05 | merlin1991: we honestly should go for some 5 betatesters, from a non-public repo, before rolling out new cssu releases | 23:40 |
DocScrutinizer05 | I wouldn't know which potential betatester would've noticed the bb-p problem, but generally it's for sure worth the 30min additional effort on your side | 23:41 |
kerio | DocScrutinizer05: honestly i still can't figure out why he had that problem | 23:42 |
DocScrutinizer05 | and we could've simply rolled out opername today and postponed cssu-t for 48h | 23:42 |
DocScrutinizer05 | kerio: I guess new cssu.enabler's uninstall function tried to detect (and actually detected) modified files (here: for bb-p) and tried to do some diff to allow rollback/unnstall | 23:43 |
amiconn | ...which would not have helped when post-poning updates due to limited internet access | 23:43 |
DocScrutinizer05 | amiconn: sure | 23:44 |
merlin1991 | DocScrutinizer05: your guess is bs | 23:44 |
DocScrutinizer05 | ok, granted | 23:44 |
DocScrutinizer05 | been wild handwaving | 23:44 |
kerio | DocScrutinizer05: it could've been the ke-recv update, actually | 23:44 |
* amiconn usually never updates when travelling, with two main reasons | 23:44 | |
merlin1991 | the facts are: we updated the pkg busybox which contains /bin/busybox, so dpkg removed /bin/busybox (which at this time actually is the bb-power binary) and replaced it with our slightly patched binary | 23:45 |
amiconn | (1) No PC available, should a recovery become available. (2) Volume cap on mobile data | 23:45 |
merlin1991 | thus all the symlinks which point to "applets" inside bb-power don't work anymore because there is a reduced busybox now in place | 23:45 |
amiconn | s/available\./necessary./ | 23:45 |
amiconn | Hmm, no infobot... | 23:46 |
kerio | merlin1991: not pager or diff though | 23:46 |
DocScrutinizer05 | damn, diff symlink should be standard bb | 23:46 |
kerio | indeed | 23:46 |
kerio | DocScrutinizer05: was that guy even using HAM? | 23:46 |
merlin1991 | standard bb does not have diff | 23:46 |
DocScrutinizer05 | ~pet amiconn | 23:46 |
* infobot reaches over and pets amiconn on the head, saying "There, there! It'll all be ok!" | 23:46 | |
kerio | ...wat | 23:47 |
kerio | merlin1991: wat | 23:47 |
kerio | how the fuck can it be a UNIX system without diff | 23:47 |
amiconn | eh? infobot being lazy then? | 23:47 |
merlin1991 | kerio: in case you do have "diff" avaiable you should check where it actually comes from | 23:47 |
DocScrutinizer05 | amiconn: \ | 23:47 |
merlin1991 | kerio: we're missing patch aswell | 23:47 |
kerio | Package diff is a virtual package provided by: busybox | 23:47 |
DocScrutinizer05 | amiconn: infobot doesn't do regex | 23:48 |
kerio | i don't even | 23:48 |
amiconn | Hmm, but there used to be a bot doing that... | 23:48 |
merlin1991 | kerio: so what? if you have extras enabled you'll always get that, even if you still have stock bb installed | 23:48 |
kerio | ...i don't even | 23:48 |
kerio | is there no diff in the sdk tools repo? :c | 23:48 |
merlin1991 | hm funky busybox does have diff in the provides list | 23:49 |
kerio | meh, nvm, it would uninstall busybox | 23:49 |
DocScrutinizer05 | amiconn: and even if it did, you "matched for . while you wrote , | 23:49 |
kerio | anyway yeah, diff comes from bb-p | 23:49 |
kerio | so... i dunno lol | 23:49 |
merlin1991 | I guess we should fix that, since busybox does not provide diff at all | 23:49 |
* merlin1991 just checked on a device with stock pr 1.3 | 23:49 | |
DocScrutinizer05 | so, I dunno who the fuck is using diff then | 23:49 |
kerio | DocScrutinizer05: apt | 23:49 |
DocScrutinizer05 | so why didn't we see similar crap since PR1.0? | 23:51 |
kerio | DocScrutinizer05: that dude was using apt | 23:52 |
DocScrutinizer05 | either diff been in bb-basic since all times (or provided as separate binary by Nokia-MP), or it never got used by apt before | 23:52 |
*** NIN101 has quit IRC | 23:52 | |
merlin1991 | DocScrutinizer05: never used | 23:52 |
kerio | it doesn't *use* it, unless you ask for it when there's a conflicting configuration file | 23:52 |
merlin1991 | it is not part of stock busybox | 23:52 |
DocScrutinizer05 | weird shit | 23:53 |
kerio | DocScrutinizer05: ok, so it appears that it's not a busybox-power problem | 23:53 |
gregoa | since when and why would apt use diff? | 23:53 |
kerio | gregoa: it doesn't use diff internally | 23:53 |
DocScrutinizer05 | indeed | 23:53 |
gregoa | ah, conffile handling. well that's dpkg and not apt :) | 23:53 |
kerio | "this configuration file was changed, and there's an update, y to install the new one, n to keep the old, d to show a diff" | 23:53 |
DocScrutinizer05 | I just can't figure nobody used that option when he ran into conflicts so far | 23:54 |
merlin1991 | hm neither apt nor dpkg depend on diff | 23:54 |
merlin1991 | (according to the dpkg database) | 23:54 |
gregoa | or they have diffutils-gnu installed (and $PATH adjusted) | 23:54 |
DocScrutinizer05 | failwale then | 23:54 |
kerio | DocScrutinizer05: maybe the latest ke-recv update is the first that had a common cause of conflicts | 23:54 |
DocScrutinizer05 | open a ticket | 23:55 |
kerio | namely, /etc/default/mmc-mount if you had kernel-power installed | 23:55 |
DocScrutinizer05 | kerio: nah! | 23:55 |
DocScrutinizer05 | there are dozen of users messing with their scripts and shit | 23:55 |
DocScrutinizer05 | .bashrc | 23:55 |
DocScrutinizer05 | .profile | 23:55 |
DocScrutinizer05 | whatnot else | 23:56 |
kerio | those are *never* | 23:56 |
kerio | er | 23:56 |
kerio | those are *never* modified by packages | 23:56 |
kerio | nothing ships (or should ship, at least) files or configuration files inside /home/user/ | 23:56 |
DocScrutinizer05 | ooh? not even when updating bash2->bash3->bash4 ? | 23:56 |
gregoa | in debian /usr/bin/diff is in diffutils which is an essential package, so no need for an explicit dependency. - no idea about dpkg + diff in maemo | 23:56 |
kerio | DocScrutinizer05: bash3 has a "bash-setup" script that does that if you want to | 23:57 |
DocScrutinizer05 | rrright | 23:57 |
*** lizardo has quit IRC | 23:57 | |
kerio | anyway, .bashrc, .profile, .bash_profile are not tracked by dpkg in any way | 23:57 |
DocScrutinizer05 | open a ticket | 23:57 |
kerio | should the next busybox update by cssu include diff? | 23:58 |
DocScrutinizer05 | if only to close as won't-fix | 23:58 |
DocScrutinizer05 | we can discuss this on the ticket | 23:58 |
DocScrutinizer05 | obviously it's a sleeping bug | 23:59 |
DocScrutinizer05 | since beginning of times | 23:59 |
DocScrutinizer05 | either dpkg is fsckd, or dependencies and bb-b are fsckd | 23:59 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!