*** M4rtinK has quit IRC | 00:06 | |
*** BCMM has quit IRC | 00:06 | |
*** M4rtinK has joined #maemo-ssu | 00:12 | |
*** M4rtinK has quit IRC | 00:28 | |
*** dhbiker has quit IRC | 01:06 | |
*** nox- has joined #maemo-ssu | 01:07 | |
*** M4rtinK has joined #maemo-ssu | 01:12 | |
*** arcean has quit IRC | 01:16 | |
*** FIQ has joined #maemo-ssu | 01:26 | |
FIQ | ok so I have a strange problem with CSSU installation, HAM first load something, then seems to do nothing. Then after a while, it simply said "Process failed" (might not be exact wording due to localized msg) and there's no icon | 01:30 |
---|---|---|
FIQ | I had a very similar issue last time I installed cssu (pretty much first thing I did that time, so no devel unless backupmenu required it, which I've forgot), but then there was an icon | 01:31 |
DocScrutinizer05 | tadzik: I think you're asking for mime-types | 01:33 |
FIQ | While CSSU seemed to install fine then, any install afterwards of anything reissued the "Please accept these TOU" of cssu-install | 01:33 |
DocScrutinizer05 | tadzik: http://wiki.maemo.org/Mime_Types#Using_hildon-mime-open | 01:33 |
FIQ | so basically both of my cssu installations has failed strangely, and both that time and now, it's pretty much clean rootfs | 01:34 |
DocScrutinizer05 | FIQ: that for sure needs investigation. Please wait til the knowledgeable people respond | 01:37 |
DocScrutinizer05 | particularly merlin1991 (and mohammadAG) might want to have a look. | 01:38 |
FIQ | I have used apt to install a few packages from extras+extras-devel (the following: bash xchat vgb gpspmgui coreutils-gnu findutils-gnu mlocate unzip) and some from HAM (backupmenu, openssh server+client) and I've changed default to /bin/bash. That's basically everything from factory state to now (might have forgot something, but nothing major) | 01:38 |
FIQ | s/default/default shell/ | 01:39 |
infobot | FIQ meant: I have used apt to install a few packages from extras+extras-devel (the following: bash xchat vgb gpspmgui coreutils-gnu findutils-gnu mlocate unzip) and some from HAM (backupmenu, openssh server+client) and I've changed default shell to /bin/bash. That's... | 01:39 |
DocScrutinizer05 | oh lol, changed default shell | 01:39 |
merlin1991 | FIQ: the cssu enabler mechanism runs a script that was never tested under bash | 01:39 |
FIQ | ohhh | 01:39 |
FIQ | so making /bin/sh default should work? | 01:40 |
FIQ | or has things broken already? | 01:40 |
merlin1991 | yeah, though I have nfc in what state you are now thanks to the broken install | 01:40 |
FIQ | hm | 01:40 |
FIQ | merlin1991: that can explain *this* fail | 01:40 |
FIQ | but not the previous one, which gave strange problems too | 01:41 |
FIQ | and then everything I did was installing backupmenu and making a backup | 01:41 |
FIQ | (from factory state, that is) | 01:41 |
DocScrutinizer05 | well, it seems nobody can reproduce this, so far | 01:42 |
FIQ | ok | 01:42 |
FIQ | I'll try to switch shell and try again | 01:42 |
DocScrutinizer05 | if you actually can reproduce failure to install CSSU on a factory image ülus BM, we got a new problem | 01:42 |
merlin1991 | FIQ: did you get as far as runing the enabler script from the app menu? | 01:42 |
FIQ | merlin1991: no | 01:42 |
FIQ | i got to step 5 on wiki | 01:43 |
FIQ | i.e. adding repo (I think), wait to HAM to do something | 01:43 |
merlin1991 | FIQ: ok time to get some info | 01:43 |
* DocScrutinizer05 indly wonders if CSSU repo is broken now too | 01:43 | |
merlin1991 | please pastebin /etc/apt/sources.list.d/hildon-application-manager.list | 01:44 |
merlin1991 | also the output of dpkg -l community-ssu-enabler | 01:44 |
FIQ | not found | 01:45 |
merlin1991 | the package or the file? | 01:45 |
DocScrutinizer05 | or pastebin? | 01:46 |
FIQ | http://pastebin.com/6caU7zZx is the sources list | 01:46 |
FIQ | merlin1991: the package | 01:46 |
DocScrutinizer05 | community-ssu-enabler not installed | 01:46 |
merlin1991 | FIQ: ok then you basically only added the repo but did nothing else | 01:46 |
DocScrutinizer05 | for some reason HAM blew chunks on installing enabler | 01:47 |
FIQ | DocScrutinizer05: no, No packages found matching community-ssu-enabler. | 01:47 |
merlin1991 | well change your shell back to /bin/sh and try again | 01:47 |
FIQ | merlin1991: ok | 01:47 |
merlin1991 | so far nothing happened to you system cssu wise :) | 01:47 |
FIQ | ok :å | 01:47 |
FIQ | s/å/p/ | 01:48 |
infobot | FIQ meant: ok :p | 01:48 |
DocScrutinizer05 | >>HAM first load something, then seems to do nothing. Then after a while, it simply said "Process failed" (might not be exact wording due to localized msg) and there's no icon<< conclusive | 01:49 |
DocScrutinizer05 | might even be connection to repo | 01:49 |
DocScrutinizer05 | HAM log... | 01:50 |
DocScrutinizer05 | or "out of memory"? | 01:50 |
DocScrutinizer05 | sth... | 01:50 |
FIQ | ehm | 01:50 |
FIQ | i'm very skeptical to OOM | 01:51 |
FIQ | connection could be the issue | 01:51 |
FIQ | also how are you supposed to know when HAM is done? | 01:51 |
DocScrutinizer05 | there are weird checks for available free storage in original HAM | 01:51 |
FIQ | it just sits there, but wiki says that's supposed to happen | 01:51 |
DocScrutinizer05 | yeah, that might actually be a bit unfortunate with HAM | 01:52 |
DocScrutinizer05 | basically as soon as you can do a new operation in HAM it's supposed to be finished so much that you can close it | 01:53 |
merlin1991 | FIQ: yeah you can run top in a terminal | 01:53 |
DocScrutinizer05 | afaik | 01:53 |
merlin1991 | as long as there is an apt-worker process eating your cpu ham is working :) | 01:53 |
DocScrutinizer05 | apt-worker keeps working even when you close HAM | 01:53 |
DocScrutinizer05 | so no problem there | 01:54 |
FIQ | good to know | 01:54 |
DocScrutinizer05 | (unless our new brilliant HAM kills apt-worker) | 01:54 |
FIQ | also what happens if you attempt to start the process again? | 01:55 |
FIQ | when it's started, that is | 01:55 |
DocScrutinizer05 | since it just installs cssu-enabler, there shouldn't be much harm | 01:56 |
FIQ | because IIRC I did that when I tried to install cssu the first time because it stalled | 01:56 |
DocScrutinizer05 | usual disclaimer: aiui | 01:56 |
FIQ | as I didn't realize it was normal behaviour for HAM to do apparently nothing but in fact do something | 01:57 |
DocScrutinizer05 | merlin1991: could we add a notifier to cssu-enabler post-install? | 01:58 |
FIQ | how long should the process take? | 01:59 |
DocScrutinizer05 | HAM? | 02:00 |
FIQ | yes | 02:00 |
DocScrutinizer05 | depends on how many repos you got enabled, I guess | 02:00 |
FIQ | woo | 02:00 |
DocScrutinizer05 | with extras-devel it usually takes 7 min here | 02:00 |
FIQ | it asked if I wanted to install enabler | 02:00 |
FIQ | it took precisely 7min :p | 02:01 |
FIQ | seemed to work | 02:01 |
DocScrutinizer05 | lesson learnt: don't make bash your default shell | 02:01 |
FIQ | I don't like busybox though | 02:02 |
* DocScrutinizer05 neither | 02:02 | |
DocScrutinizer05 | ~jrtools | 02:02 |
infobot | jrtools is, like, http://wiki.maemo.org/User:Joerg_rw/tools | 02:02 |
DocScrutinizer05 | ok, I have to start it by typing `bash` as first command to xterm | 02:03 |
FIQ | maybe I should try to edit .shrc or whatever the equavilence to .bashrc is, and check for PS1 in order to spawn bash on top of it | 02:03 |
FIQ | it's hackish, but i'm lazy | 02:03 |
DocScrutinizer05 | ~jrtools | 02:03 |
infobot | well, jrtools is http://wiki.maemo.org/User:Joerg_rw/tools | 02:03 |
DocScrutinizer05 | messybox is sourcing bashrc | 02:03 |
FIQ | so messybox thinks it's bash? | 02:04 |
DocScrutinizer05 | you *might* check if you got an interactive shell or not. But probably that doesn't pan out either, in the end | 02:04 |
FIQ | fair enough, i'll type bash manually | 02:04 |
DocScrutinizer05 | some retard maintained thought messybox == bash, yes | 02:04 |
DocScrutinizer05 | FIQ: check my nifty PATH joggling | 02:05 |
FIQ | i'd rather not touch busybox, but i guess i'm forced as maemo depend on it | 02:05 |
FIQ | hmm | 02:05 |
*** dhbiker has joined #maemo-ssu | 02:06 | |
FIQ | you mean placing /usr/bin/gnu first? | 02:06 |
DocScrutinizer05 | juggling even | 02:06 |
DocScrutinizer05 | yep | 02:06 |
FIQ | I already do that :P | 02:06 |
FIQ | to not have to bother with busybox binary | 02:06 |
DocScrutinizer05 | hmm, actually my recent files look different | 02:07 |
FIQ | export PATH="/usr/bin/gnu:"$PATH":$HOME/bin is my path atm | 02:08 |
FIQ | ("bin" actually contains shellscripts but whatever) | 02:09 |
DocScrutinizer05 | I once got a nifty sed cmdline, but I seem too ill and dizzy to locate it | 02:14 |
*** kolp has quit IRC | 02:14 | |
DocScrutinizer05 | sorry | 02:14 |
FIQ | also | 02:15 |
FIQ | why does HAM (not only for cssu, but that's one of the things) show "(null) Install anyway?" when installing? | 02:15 |
DocScrutinizer05 | dafaq ~user/.bashrc vanished | 02:15 |
FIQ | evil maemo | 02:16 |
DocScrutinizer05 | or I'm starting to lose my mind | 02:16 |
MrPingu | Hmm, what I don't like or liked about the bash thingy; as soon you ran sudo gainroot from bash, it returns you to busybox | 02:16 |
FIQ | MrPingu: I just made sudo usable as on linux desktop | 02:16 |
FIQ | by making an own file in sudoers.d | 02:16 |
DocScrutinizer05 | well, that's 'normal' since you open a new default shell | 02:16 |
FIQ | "Upgrade for Maemo" guess I'll be afk shortly | 02:17 |
MrPingu | Yes, I know but I don't like it | 02:17 |
MrPingu | I always close xterm by issuing exit | 02:17 |
DocScrutinizer05 | if you don't like what the command you invoked is doing, find a better command | 02:18 |
FIQ | ................. | 02:18 |
FIQ | That was stupid | 02:18 |
FIQ | HAM closed everything, *then* proceed to tell me that there wasn't enough battery | 02:18 |
MrPingu | It would mean in the worst case I would have to issue it 4 times :P | 02:19 |
MrPingu | DocScrutinizer05: You got a point there! | 02:19 |
DocScrutinizer05 | MrPingu: btw bash for root seems semi-safe | 02:19 |
DocScrutinizer05 | while bash for user definitely never been | 02:19 |
FIQ | is there a reason that you shouldn't use "proper sudo"? | 02:20 |
DocScrutinizer05 | paradox but true | 02:20 |
DocScrutinizer05 | nah, sudo is used all the time, but sudo isn't `root` aka sudo gainroot | 02:20 |
FIQ | i.e. enable yourself to do "sudo (whatever)" instead of logging in as root and do "(whatever)" | 02:20 |
DocScrutinizer05 | if you log in via `root` you get root's environment ($HOME etc) | 02:21 |
FIQ | well.. yeah | 02:21 |
MrPingu | semi-safe still don't like that :P For now I stick to busybox, till I hate it so much and can't bear it anymore | 02:21 |
DocScrutinizer05 | if you log in via `sudo gainroot` you're still in user's environment | 02:22 |
FIQ | I use sudo as on a desktop, basically | 02:22 |
MrPingu | DocScrutinizer05: Thanks for the passwd for root, love it :) | 02:22 |
FIQ | i.e. no "root" or "sudo gainroot" | 02:22 |
DocScrutinizer05 | if you do sudo $somecmd, then you're quite obviously in user's environment still | 02:22 |
FIQ | yes I know | 02:23 |
MrPingu | But sudser fvcks it up | 02:23 |
DocScrutinizer05 | sudser is brainfsck | 02:23 |
FIQ | MrPingu: I did it manually | 02:23 |
MrPingu | FIQ: would love to hear how, but I really need to get some sleep now | 02:23 |
MrPingu | Good night! | 02:24 |
DocScrutinizer05 | MrPingu: I thought a lot how to make root password more safe. First step: make HAM ask for root password, so nobody can install crap to nuke passwords | 02:24 |
FIQ | made a file in sudoers.d, typed "user ALL = (ALL) ALL" and ran update-sudoers | 02:25 |
DocScrutinizer05 | I discussed it with kerio a whole night long | 02:25 |
FIQ | /etc/sudoers.d, that is | 02:25 |
*** MrPingu has quit IRC | 02:25 | |
DocScrutinizer05 | FIQ: definitely a poor idea | 02:26 |
DocScrutinizer05 | in my book | 02:26 |
FIQ | hmmmm | 02:26 |
FIQ | how so? | 02:26 |
DocScrutinizer05 | it defeats the whole idea of root permissions | 02:26 |
FIQ | DocScrutinizer05: maemo is a single user environment | 02:27 |
DocScrutinizer05 | you as well could set UID for user to 0 | 02:27 |
DocScrutinizer05 | FIQ: definitely not | 02:27 |
*** FireFly has quit IRC | 02:27 | |
DocScrutinizer05 | there's users "root", "user", "apt", "ntpd", "dahellknows".... | 02:28 |
FIQ | fair enough, but I don't see the problem | 02:28 |
DocScrutinizer05 | not exactly a singleuser system | 02:28 |
FIQ | sudo gainroot gives you root access in user environment | 02:29 |
DocScrutinizer05 | so what? | 02:29 |
DocScrutinizer05 | sudo gainroot asks for root password on my machine | 02:29 |
FIQ | the above gives you the same, just with letting you to not have to do sudo gainroot before | 02:29 |
FIQ | sudo gainroot also does it w/o password by default, while the above *do* ask the password, for user | 02:30 |
DocScrutinizer05 | FIQ: chmod -R a+rwx / | 02:30 |
FIQ | why would you do that? | 02:31 |
DocScrutinizer05 | FIQ: I'm not going to debate with you how to fuck up a fubar concept even more | 02:31 |
*** FireFly has joined #maemo-ssu | 02:31 | |
DocScrutinizer05 | FIQ: same question to your "user ALL = (ALL) ALL" | 02:32 |
DocScrutinizer05 | why would you do that? | 02:32 |
DocScrutinizer05 | since chmod -R a+rwx / work much better, by not even requiring a sudo in front of any braindamaged command user may want to nuke system with | 02:33 |
FIQ | DocScrutinizer05: it also gives all other users root | 02:33 |
DocScrutinizer05 | either use permissions, the put it to purpose properly, or not use them at all, no? | 02:34 |
DocScrutinizer05 | <FIQ> DocScrutinizer05: maemo is a single user environment | 02:35 |
DocScrutinizer05 | ((<FIQ> DocScrutinizer05: it also gives all other users root)) so what? | 02:36 |
DocScrutinizer05 | either you bother about security, or not | 02:36 |
DocScrutinizer05 | the most dangerous processes run under user anyway | 02:37 |
DocScrutinizer05 | multiuser setup and user separation is for security purposes. Not much sense in neutering it for user | 02:38 |
DocScrutinizer05 | even less sense in improving that neutering | 02:38 |
DocScrutinizer05 | Nokia fsckd it up by running stuff like HAM under user and not daring to ask for a password when it invokes apt-worker under root | 02:39 |
FIQ | DocScrutinizer05: I'm "user", and last time I checked, I'm the owner of the n900. sudo gainroot gives user full root (without even a password by default), my proposal does it too. The difference is that in one way needs me to first type a command to allow root commands, then a second do what I want, then exit when done with a third command while my line lets you do it with one command | 02:39 |
FIQ | Also, my thing ask for a password. sudo gainroot doesn't | 02:40 |
DocScrutinizer05 | meh | 02:40 |
FIQ | Do you disallow root for yourself on your desktop machines too? | 02:41 |
DocScrutinizer05 | who the fuck disallows anything? | 02:41 |
DocScrutinizer05 | sorry, this exhausts me too much. should stay in bed curing my flu | 02:42 |
FIQ | I don't see the problem here, but I might miss something here, I don't deny it | 02:43 |
DocScrutinizer05 | ~jrtools | 02:43 |
infobot | somebody said jrtools was http://wiki.maemo.org/User:Joerg_rw/tools | 02:43 |
FIQ | (Also chmod 777 is a bad idea for other reasons too, as it screws things up in other ways, but that's irrelevant) | 02:44 |
DocScrutinizer05 | FIQ: I missed sth: your setup asks for password - user password though, which is kinda silly ubuntu bullshit, but kinda works for maemo since you can't change your own user password on maemo | 02:47 |
FIQ | [01:41:30] <FIQ> Do you disallow root for yourself on your desktop machines too? -- i.e. you don't give your normal user account full sudo as you think it's bad security principe | 02:47 |
FIQ | oh you thought i had nopasswd? | 02:47 |
DocScrutinizer05 | yes, my users never have root access by using their *own* password | 02:48 |
FIQ | hmm | 02:48 |
DocScrutinizer05 | root access should need root password, not user's password | 02:48 |
FIQ | let me think now | 02:48 |
FIQ | changing password can be done on the user's account | 02:48 |
DocScrutinizer05 | n8 again | 02:48 |
FIQ | (normally) | 02:49 |
FIQ | so basically asking for user password is like asking for no password at all | 02:49 |
DocScrutinizer05 | not as user, under maemo (thanks busybox passwd) | 02:49 |
FIQ | yeah, i said "(normally)" | 02:49 |
DocScrutinizer05 | only root can change user's password on maemo | 02:49 |
FIQ | I think I get your point now | 02:49 |
FIQ | though in practice maemo is a joke to get root on due to HAM | 02:52 |
FIQ | but still | 02:52 |
DocScrutinizer05 | see my comment several lines above | 02:52 |
FIQ | yeah I saw it | 02:52 |
DocScrutinizer05 | actually my comments (plural) | 02:52 |
FIQ | no need to break it more | 02:53 |
DocScrutinizer05 | [2012-11-16 01:24:46] <DocScrutinizer05> MrPingu: I thought a lot how to make root password more safe. First step: make HAM ask for root password, so nobody can install crap to nuke passwords | 02:53 |
FIQ | yeah | 02:53 |
FIQ | saw the line | 02:53 |
FIQ | well you've modified HAM in cssu, making that shouldn't be too hard :p | 02:54 |
DocScrutinizer05 | if we had a better sudo executable (more recent) we could solve this pretty nicely | 02:54 |
FIQ | also what about the autologin? | 02:55 |
DocScrutinizer05 | but yeah, we might fix HAM as well | 02:55 |
FIQ | (on boot, that is) | 02:55 |
DocScrutinizer05 | another PITA | 02:55 |
FIQ | it's just "user" but still | 02:55 |
DocScrutinizer05 | it should attach PAM to device lockcode and PIN query | 02:55 |
DocScrutinizer05 | actually lockcode should be a PAM module | 02:56 |
FIQ | PIN is easy to dodge if you just want to access the system | 02:56 |
DocScrutinizer05 | and H-D startup get done by sth like XDM | 02:57 |
FIQ | lockcode or asking for user's password, it is | 02:57 |
FIQ | My previous device asked for lockcode on startup :p | 02:58 |
DocScrutinizer05 | honestly they ought have implemented lockcode query into NOLO | 02:59 |
DocScrutinizer05 | as well as charging ;-P | 03:00 |
FIQ | shouldn't it be possible to query on boot? | 03:00 |
FIQ | i mean, there's bootmenu, shouldn't be too hard to make a "bootloader" that asks for password before proceeding | 03:01 |
DocScrutinizer05 | nah, it will still aloow resueOS and flashing | 03:01 |
DocScrutinizer05 | both done by NOLO | 03:01 |
FIQ | well flashing nukes system | 03:01 |
FIQ | like reinstalling the OS on a desktop :p | 03:02 |
DocScrutinizer05 | yep, but on desktop you got device password on BIOS level | 03:02 |
DocScrutinizer05 | if you want | 03:02 |
DocScrutinizer05 | rendering device basically useless for thieves | 03:02 |
DocScrutinizer05 | no such thing on N900 | 03:03 |
DocScrutinizer05 | thanks NOLO (our 'BIOS') | 03:03 |
FIQ | True | 03:03 |
DocScrutinizer05 | o/ | 03:04 |
FIQ | well as long as the OS isn't open doors, I'm happy, if people really want to get my data, go on and steal my HDD, I'm not *that* paranoid :3 | 03:04 |
FIQ | bye | 03:04 |
*** Jade has quit IRC | 03:06 | |
FIQ | (also isn't uboot lowlevel enough?) | 03:06 |
*** Jade has joined #maemo-ssu | 03:09 | |
*** Jade has joined #maemo-ssu | 03:09 | |
*** Jade has quit IRC | 03:23 | |
*** Jade has joined #maemo-ssu | 03:26 | |
*** Jade has joined #maemo-ssu | 03:26 | |
DocScrutinizer05 | uboot comes after NOLO, flashing / rescueOS booting is done *in* NOLO | 03:35 |
*** M4rtinK has quit IRC | 03:36 | |
DocScrutinizer05 | actually even NOLO can get coldflashed by a new one that wouldn't ask for password | 03:36 |
DocScrutinizer05 | since ROMBOOT allows coldflashing | 03:36 |
FIQ | there we go, CSSU installed | 03:37 |
DocScrutinizer05 | but afaik NOLO resp the XLoader header at least needs a signature to get executed by ROMBOOT, so nobody could tamper XLoader | 03:37 |
FIQ | HAM took literally 15min to show update... | 03:38 |
DocScrutinizer05 | hmm, | 03:38 |
DocScrutinizer05 | terrible | 03:38 |
FIQ | once I booted from new battery | 03:38 |
FIQ | the update itself took 8min | 03:38 |
DocScrutinizer05 | but don't worry, updates are less painful | 03:38 |
FIQ | so HAM took less time to do what I actually wanted than it took to show me it, even though it showed me previously | 03:39 |
FIQ | yaay! | 03:39 |
DocScrutinizer05 | once the automatic check for updates shows "maemo 5 CSSU update available" it doesn't take forever to show that update in HAM. Then you fire "update that" and forget | 03:39 |
FIQ | i "disabled" the check | 03:40 |
FIQ | so I went manually opening HAM and press update button | 03:40 |
FIQ | that task took 15min | 03:40 |
DocScrutinizer05 | yeah, that's when it downloads *all* repo 'indexes' and builds databese with dependencies | 03:41 |
FIQ | hmmm | 03:41 |
FIQ | can I uninstall these stuff taking up place on my appmenu w/o problems? | 03:42 |
DocScrutinizer05 | on my 300MHz 150MB RAM laptop the same procedure in Yast took a day, literally | 03:42 |
FIQ | amazon, ap news, documents to do, facebook, foreca, that is | 03:42 |
DocScrutinizer05 | "these stuff"? | 03:42 |
FIQ | sounds.. "fun" | 03:43 |
FIQ | i mentioned which i meant :p | 03:43 |
DocScrutinizer05 | yeah sure, apt-get purge | 03:43 |
FIQ | wonder what they're called.. hm | 03:43 |
DocScrutinizer05 | apt-cache search | 03:43 |
FIQ | yeah | 03:44 |
FIQ | also your HAM doesn't kill apt-worker on exit | 03:44 |
DocScrutinizer05 | first remove the friggin desktop widget though, in desktop edit mode | 03:44 |
FIQ | i don't use the desktop | 03:45 |
FIQ | removed everything etc | 03:45 |
FIQ | blank bg | 03:45 |
FIQ | :p | 03:45 |
DocScrutinizer05 | then it's safe to apt-get purge that crap | 03:45 |
DocScrutinizer05 | (though foreca isn't that bad, I actually use it) | 03:46 |
FIQ | argh didn't found the name for documents to go stuff | 03:48 |
DocScrutinizer05 | hmm, sorry can't look it up here anymore ;-) | 03:49 |
DocScrutinizer05 | or maybe I can? | 03:49 |
FIQ | found it | 03:50 |
FIQ | "dtg-installer" | 03:50 |
DocScrutinizer05 | dtg-installer | 03:50 |
DocScrutinizer05 | :-D | 03:50 |
FIQ | what about the er | 03:51 |
FIQ | "Handbook" and the "how to start" entries? :3 | 03:52 |
FIQ | or that mahjong game for that matter | 03:52 |
*** nox- has quit IRC | 04:15 | |
*** Jade has quit IRC | 04:39 | |
*** dafox has quit IRC | 04:40 | |
*** Jade has joined #maemo-ssu | 04:42 | |
*** luf has joined #maemo-ssu | 04:45 | |
*** luf has quit IRC | 05:19 | |
*** amiconn has quit IRC | 05:28 | |
*** amiconn_ has joined #maemo-ssu | 05:28 | |
*** amiconn_ is now known as amiconn | 05:28 | |
*** luf has joined #maemo-ssu | 05:51 | |
*** DocScrutinizer05 has quit IRC | 06:03 | |
*** DocScrutinizer05 has joined #maemo-ssu | 06:03 | |
*** luf has quit IRC | 06:31 | |
*** Jade has quit IRC | 08:00 | |
*** Jade has joined #maemo-ssu | 08:03 | |
*** Jade has joined #maemo-ssu | 08:03 | |
kerio | DocScrutinizer05: CSSU replacement for foreca-widget! | 08:10 |
*** chainsawbike has quit IRC | 08:27 | |
*** chainsawbike has joined #maemo-ssu | 08:28 | |
tadzik | DocScrutinizer05: (mime types), no, I mean the code that's responsible for reacting to longpress, drawing the context menu, making it disappear etc | 09:07 |
kerio | gtk, maybe? | 09:16 |
*** X-Fade has quit IRC | 09:28 | |
*** X-Fade has joined #maemo-ssu | 09:29 | |
*** FIQ has quit IRC | 09:44 | |
tadzik | no, it's somewhere above Gtk, I think | 10:03 |
tadzik | What I want to do is to make the context menu not disappear when you stop pressing the screen | 10:03 |
tadzik | so you can actually see what popped out | 10:03 |
tadzik | in Gtk the context menu does not disappear when you cease pressing right mouse button, so I guess that's have to be some hildon shamanism | 10:04 |
*** BCMM has joined #maemo-ssu | 10:49 | |
*** luf has joined #maemo-ssu | 11:04 | |
*** dhbiker has quit IRC | 11:25 | |
*** kolp has joined #maemo-ssu | 12:05 | |
*** dhbiker has joined #maemo-ssu | 12:21 | |
DocScrutinizer05 | tadzik: for some reason I sometimes achieve to not close context menu when lifting stylus. I guess it's rather a bug that hits when touchpoint coords move minimally on lifting stylus, rather than an option designed in on purpose | 12:33 |
vi__ | It happens when you lift off the very edge of the menu perimeter. | 12:37 |
vi__ | I believe the relevent code is in hildon-home. | 12:37 |
tadzik | DocScrutinizer05: indeed, it is possible | 12:37 |
tadzik | then that's a bug worth fixing, imho | 12:38 |
tadzik | (and will probably turn out to be a PITA) | 12:42 |
DocScrutinizer05 | nah, the fix should be 'easy': just don't close the menu when pen-up event happens outside of menu area, no matter if coords moved away from pen-down coords or not | 13:01 |
DocScrutinizer05 | actually it seems to me this is indeed a function that has to be coded in on purpose. Probably just needs commenting out | 13:02 |
DocScrutinizer05 | if you wanna do a real fix, then check for distance between pen-down and pen-up event and close menu only when distance is >threshold | 13:04 |
DocScrutinizer05 | maybe, just *maybe*, there's already such check for distance, and somebody simply set a fsckng low threshold | 13:05 |
tadzik | possibly | 13:05 |
kerio | the context menu does indeed disappear if you right-click-and-hold and then release the right mouse button | 13:05 |
tadzik | I'll have to locate the code first; will look more after $dayjob | 13:05 |
kerio | iirc | 13:06 |
tadzik | not unles you move it over some menu item | 13:07 |
tadzik | if you click, hold, drag away it stays | 13:07 |
*** Pali has joined #maemo-ssu | 13:14 | |
*** arcean has joined #maemo-ssu | 13:21 | |
vi__ | kerio: That implies it is a bug that is a result of the emulated push+hold right click. | 13:31 |
*** kolp_ has joined #maemo-ssu | 13:39 | |
*** kolp has quit IRC | 13:42 | |
*** NIN101 has joined #maemo-ssu | 14:11 | |
DocScrutinizer05 | tadzik: ((if you click, hold, drag away it stays)) unlike on touchscreen where it vanishes if you dare to drag | 14:32 |
DocScrutinizer05 | tadzik: which is a bit strange | 14:32 |
tadzik | DocScrutinizer05: yeah. I was talking about desktop gtk | 14:33 |
tadzik | in hildon that's strange and annoying imho | 14:33 |
tadzik | but for example Opera got it right on maemo | 14:33 |
DocScrutinizer05 | (the absolute contrary behavior between mouse RMB and stylus longtouch) | 14:33 |
DocScrutinizer05 | also I find anchor of context menu got chosen rather unfortunate (menu appears to right of touchpoint, so you basically *always* cover it with stylus/finger) | 14:35 |
tadzik | true | 14:35 |
tadzik | room for improvement! \o/ | 14:36 |
DocScrutinizer05 | what a fool chosen those parameters? | 14:36 |
DocScrutinizer05 | obviously somebody who tested on PC screen in emulator | 14:36 |
DocScrutinizer05 | with mouse | 14:36 |
tadzik | entire maemo feels this way for me | 14:36 |
tadzik | well, almost | 14:37 |
tadzik | it's not quite utilizing the potential in a touchscreen interface | 14:37 |
kerio | nah | 14:37 |
tadzik | like the status area. Clicking small stuff with mouse is fine. Clicking small stuff with big finger is tricky | 14:37 |
tadzik | call me blasphemer, but I think android does it right :) | 14:38 |
tadzik | this particular thing | 14:38 |
tadzik | although n900 does not have much space above screen, so it may not be such a good idea | 14:38 |
vi__ | DocScrutinizer05: obviously someone who was left handed... | 14:39 |
DocScrutinizer05 | I'm using fingernail. I actually love the smaller touchareas in fremantle, compared to finger-only GUI design of e.g. harmattan | 14:39 |
vi__ | A WITCH! | 14:39 |
DocScrutinizer05 | vi__: yep, or lefthanded | 14:40 |
DocScrutinizer05 | but then that guy ought have *forced* a "lefthand/righthand" user setting option into fremantle screen settings | 14:41 |
DocScrutinizer05 | after all at least that guy *should* be aware of the problems of the other part of mankind that doesn't use same hand as the rest | 14:42 |
kerio | DocScrutinizer05: he did that on purpose | 14:42 |
DocScrutinizer05 | revenge | 14:42 |
kerio | to inflict vengeance to every right-handed person on the planet! | 14:42 |
kerio | s/on the planet/that happened to be using a n900, so very few people/ | 14:42 |
infobot | kerio meant: to inflict vengeance to every right-handed person that happened to be using a n900, so very few people! | 14:42 |
DocScrutinizer05 | actually this is a pretty good filed and opportunity for a mad useful improvement in CSSU | 14:43 |
DocScrutinizer05 | field* | 14:43 |
DocScrutinizer05 | arcean: ^^^ | 14:44 |
DocScrutinizer05 | arcean: last ~150 lines | 14:44 |
DocScrutinizer05 | arcean: in short: implement (or check if already there) a threshold for drag after stylus-down, to make context menu close. Do not close when threshold distance isn't exceeded. and create a config option to change anchor point of context menu to arbitrary of 4 corners | 14:46 |
*** Timo has joined #maemo-ssu | 15:16 | |
*** Timo has quit IRC | 16:06 | |
*** M4rtinK has joined #maemo-ssu | 16:07 | |
*** BCMM has quit IRC | 16:08 | |
*** M4rtinK has quit IRC | 16:10 | |
*** luf has quit IRC | 16:13 | |
*** M4rtinK has joined #maemo-ssu | 16:17 | |
*** M4rtinK has quit IRC | 16:36 | |
*** arcean has quit IRC | 16:41 | |
*** M4rtinK has joined #maemo-ssu | 16:44 | |
*** M4rtinK has quit IRC | 16:49 | |
*** Timo has joined #maemo-ssu | 17:58 | |
*** arcean has joined #maemo-ssu | 18:01 | |
*** toxaris has joined #maemo-ssu | 18:17 | |
*** toxaris has quit IRC | 18:47 | |
*** Timo has quit IRC | 19:04 | |
*** Linger has joined #maemo-ssu | 19:17 | |
Linger | Good morning | 19:17 |
Linger | I'm a fellow with some fairly easy technical questions. | 19:19 |
Linger | Questions regarding IRC. | 19:19 |
wmarone | ask, don't wait | 19:20 |
Linger | Ok, thank you. I'm an Aussie. I have not been on IRC for 10 years. Where would I go do talk to Aussies? | 19:20 |
DocScrutinizer05 | not on freenode. channels on freenode are technical topics, not geographical | 19:21 |
Linger | I'm 26. Used to me on IRC a lot when I was a teen. But I lost interest for a while. | 19:21 |
wmarone | well that's not a technical question | 19:21 |
Linger | Fair point | 19:22 |
DocScrutinizer05 | we got some aussies in every maemo related channel | 19:22 |
Linger | Well I would be happy to ask a technical question, as I have an InfoSec background | 19:22 |
*** M4rtinK has joined #maemo-ssu | 19:23 | |
kerio | Linger: did you just stumble upon #maemo-ssu? | 19:23 |
Linger | Truth is, it's Friday night here in Aus. I've had a few beers. I'm looking for some interesting discussion before I crash. | 19:23 |
Linger | Yes I did | 19:23 |
kerio | this is a channel dedicated to a community project for a niche OS for niche smartphones | 19:24 |
Linger | However I'd be more than happy to leave you gents alone. I just don't know where the Aussies go on IRC anymore. I've been out of it for a long time. | 19:25 |
Linger | I apologise. | 19:25 |
Linger | I have no intention to take the discussion off topic | 19:26 |
kerio | perhaps you might want to connect to a less-specific irc network? | 19:26 |
Linger | I'm open to suggestions | 19:26 |
Linger | Like I said, I have no idea where to go. | 19:27 |
kerio | \_o_/ | 19:27 |
kerio | i have no idea, sorry | 19:27 |
Linger | I'm a somewhat drunk InfoSec fag. Aussie. | 19:27 |
Linger | No problem mate. | 19:28 |
*** _rd has joined #maemo-ssu | 19:28 | |
Linger | Would it be a problem to entertain my somewhat drunken self on this channel for a while? | 19:28 |
Linger | Given that there does not seem to be much else going on. | 19:29 |
kerio | are you kidding me | 19:30 |
Linger | I can ensure my questions are technology based. | 19:30 |
kerio | questions about what | 19:30 |
kerio | there's 53 people here | 19:31 |
kerio | there's probably about 20 times in #linux or whatever | 19:31 |
Linger | And none of them seem to be active. | 19:31 |
DocScrutinizer05 | so you're interested in monologues? | 19:31 |
Linger | I've worked as an InfoSec analyst for a while. | 19:32 |
DocScrutinizer05 | tease us | 19:32 |
Linger | Surely theres some middle ground | 19:32 |
Linger | Monologues to me relates to the English literature definition. | 19:33 |
kerio | DocScrutinizer05: he's annoying and off-topic - but then again, there's not really anything else going on, here | 19:34 |
DocScrutinizer05 | I'm not yet ready to kick him | 19:34 |
Linger | Fair call. | 19:34 |
DocScrutinizer05 | off topic isn't a problem as long as nobody complains | 19:35 |
Linger | Is InfoSec off topic? | 19:35 |
DocScrutinizer05 | chances for somebody complaining increase the more the talk is OT though | 19:35 |
DocScrutinizer05 | not per se | 19:35 |
kerio | wtf is infosec anyway? | 19:35 |
Linger | Information Security | 19:36 |
Linger | I did InfoSec for a very large company, that you would all know. | 19:37 |
Linger | You cunts are interested in tech issues. I imagine you may have questions | 19:37 |
kerio | DocScrutinizer05: we could make him audit maemo :D | 19:38 |
Linger | This is on topic no? | 19:38 |
*** _rd has quit IRC | 19:38 | |
DocScrutinizer05 | 1. warning: speech | 19:38 |
*** ChanServ sets mode: +o kerio | 19:40 | |
DocScrutinizer05 | just in case | 19:40 |
Linger | Ask a cunt who was considered one of the three people within the organisation who was considered knowledgeable enough to make judgements on InfoSec issues. | 19:41 |
Linger | ask anything | 19:41 |
Linger | A global organisation. 70,000+ users | 19:42 |
Linger | Is this not intersting? | 19:43 |
Linger | No? I do remember how IRC works. You can reply to me if you like. | 19:48 |
Linger | I have a good job. I have made the most of my IT experience. | 19:50 |
Linger | So what? | 19:50 |
Linger | Kick/Ban me. | 19:50 |
DocScrutinizer05 | how to get the root certificate of any security infra (here: the one Nokia implemented with aegis) | 19:52 |
DocScrutinizer05 | (you know, for the simple questions we got google and wikipedia nowadays) | 19:52 |
* Lava_Croft strokes his beard | 19:53 | |
*** _rd has joined #maemo-ssu | 19:55 | |
Linger | Ok. So I'm a guy who works on security on Pc's. I recently got a new phone. A Samsung S3 $g. | 19:56 |
Linger | Please tell me what issues I way have with this phone. | 19:57 |
Linger | 4g | 19:57 |
Linger | rather | 19:57 |
Linger | No? | 20:00 |
Linger | It seems as though IRC is as I remember it. | 20:01 |
Linger | So how long are you neck beards? | 20:03 |
Linger | IRC is where you feel like you have some power | 20:04 |
kerio | DocScrutinizer05: hold on, why am i opped? | 20:05 |
DocScrutinizer05 | just in case, since I'm busy | 20:05 |
*** _rd has quit IRC | 20:05 | |
Linger | Sd. | 20:08 |
Linger | Sad | 20:08 |
Linger | And then what? | 20:09 |
Linger | Cook some bacon and eggs | 20:11 |
kerio | Linger: you're being annoying | 20:11 |
kerio | stop or leave | 20:11 |
*** Linger has quit IRC | 20:13 | |
kerio | hm | 20:13 |
*** kerio sets mode: -o kerio | 20:14 | |
*** _rd has joined #maemo-ssu | 20:14 | |
*** Pali has quit IRC | 20:15 | |
*** _rd has quit IRC | 20:16 | |
*** NIN102 has joined #maemo-ssu | 20:35 | |
*** NIN101 has quit IRC | 20:37 | |
*** _xnt14 has quit IRC | 20:50 | |
*** _xnt14 has joined #maemo-ssu | 20:59 | |
freemangordon | merlin1991: ping | 21:16 |
*** dhbiker has quit IRC | 21:17 | |
merlin1991 | freemangordon: pong | 21:33 |
*** arcean has quit IRC | 21:33 | |
*** Woody14619b has quit IRC | 21:40 | |
*** Woody14619 has joined #maemo-ssu | 21:41 | |
*** arcean has joined #maemo-ssu | 21:59 | |
freemangordon | merlin1991: http://talk.maemo.org/showpost.php?p=1295049&postcount=59 | 22:13 |
freemangordon | tklock leaks like mad :( | 22:14 |
freemangordon | I need an advise and/or help what to do | 22:14 |
merlin1991 | let's start easy for me, what is tklock? | 22:15 |
freemangordon | swipe-to-unlock thingie | 22:15 |
arcean | oh nooo | 22:15 |
freemangordon | arcean: yes :( | 22:15 |
freemangordon | try it | 22:15 |
arcean | looks like everything with 'systemui' in a name leaks | 22:16 |
freemangordon | yes | 22:16 |
merlin1991 | freemangordon: does it need anything special on device to start leaking like mad? | 22:16 |
freemangordon | nothing afaik | 22:16 |
freemangordon | well, didn't tested on stock, only on -thumb | 22:16 |
freemangordon | but that should be irrelevant | 22:16 |
merlin1991 | anything special to test that? | 22:16 |
freemangordon | nothing | 22:16 |
* merlin1991 does not even know how to get per process ram on linux :$ | 22:17 | |
freemangordon | check /proc/`pidof systemui`/smaps | 22:17 |
freemangordon | [heap] section | 22:17 |
freemangordon | it leaks about 16k on unlock | 22:17 |
freemangordon | now, the real problem is that there is no x86 binary | 22:18 |
freemangordon | and I am just not good enough with GTK to rewrite it | 22:18 |
merlin1991 | is the smaps output a huge thing? | 22:19 |
merlin1991 | for me it goes over the xterm backsroll buffer | 22:19 |
freemangordon | cat /proc/`pidof systemui`/smaps | more | 22:19 |
freemangordon | it is the 3rd thing | 22:19 |
freemangordon | BTW that leak is without any sms-es, mails, etc | 22:20 |
freemangordon | didn't check what happens if there is some icon on unlock screen | 22:20 |
merlin1991 | hm it happily stays @ 1296kb here after an unlock (stock device) | 22:22 |
*** dhbiker has joined #maemo-ssu | 22:23 | |
merlin1991 | leaks only from time to time, not always | 22:24 |
freemangordon | yeah, in chunks of 48k | 22:24 |
merlin1991 | I managed todo some 3 unlocks without an increase in heap size | 22:24 |
freemangordon | or something | 22:24 |
freemangordon | yep 48k/3 unlocks | 22:24 |
kerio | dafuq, osso-systemui-tklock can be uninstalled | 22:24 |
merlin1991 | /proc/something should be live right? | 22:25 |
merlin1991 | kerio: obviously | 22:25 |
kerio | will shit break if i do? | 22:25 |
merlin1991 | it was only hold by the old @nokia metapackge | 22:25 |
merlin1991 | kerio: nfc | 22:25 |
kerio | ok, i have to make a backup anyway | 22:25 |
freemangordon | well, it is usefull for those of us who don't have cases | 22:25 |
merlin1991 | freemangordon: yep it definitely leaks when you use the slider | 22:26 |
freemangordon | merlin1991: yeah, I know :( | 22:26 |
merlin1991 | arcean: what else with systemui did leak? | 22:27 |
freemangordon | btw the first time you use it (after reboot) heap usage jumps from 700k to 1400k | 22:27 |
merlin1991 | freemangordon: that's probably a gtkpixbuf for the backgroundimage | 22:27 |
freemangordon | so, not only it leaks, it seems to keep the background in memory | 22:27 |
kerio | merlin1991: 700kb for a 800x480 bitmap? | 22:27 |
arcean | merlin1991: osso-systemui-alarm :P | 22:27 |
freemangordon | merlin1991: exactly my thoughts | 22:27 |
freemangordon | arcean: well, that should be ok now, check the latest commits | 22:28 |
merlin1991 | kerio: do the maths, 3 byte each pixel | 22:28 |
freemangordon | and that's pretty instane | 22:28 |
kerio | hm | 22:28 |
kerio | that's about 1mb | 22:28 |
freemangordon | yeah | 22:28 |
merlin1991 | yeah | 22:28 |
kerio | neat! | 22:29 |
freemangordon | of memory which gets lost | 22:29 |
freemangordon | effectively | 22:29 |
Woody14619 | Actually... IT shouldn't, as it reloads it on re-lock anyway... | 22:29 |
freemangordon | most probably it does not reload it | 22:29 |
Woody14619 | I say, having written a script that updates the screen, so I know it in fact does. | 22:29 |
merlin1991 | freemangordon: or it reuses the same pixbuffer and loads a new image into it ;) | 22:30 |
freemangordon | well, no idea what it does, but it is a piece of crap | 22:30 |
freemangordon | merlin1991: could be | 22:30 |
Woody14619 | merlin1991: That may be to.. It's a fixed-size picture. | 22:30 |
Woody14619 | I wonder which is worse? Having the pages allocated, or freeing an allocing each time, causing fragmentation? | 22:31 |
merlin1991 | well I see a point in keeping the pixbuf around, but the leak is definitely a nogo | 22:31 |
freemangordon | Woody14619: if you have pages allocated and DON'T LEAK makes sense | 22:32 |
kerio | just keep dat shit on the stack | 22:32 |
* Woody14619 nods | 22:32 | |
kerio | or, better, uninstall it | 22:32 |
freemangordon | though if I am to write such a think, i'll free the pixbuf when not needed | 22:33 |
freemangordon | *thing | 22:33 |
merlin1991 | freemangordon: well in this case the probably had a minimal delay till displayed requirement | 22:33 |
merlin1991 | and loading the image into memory produces "some" delay, especially when the phone does something else at the time the user hits the button | 22:34 |
freemangordon | merlin1991: agree, makes sense | 22:34 |
freemangordon | though by the time you need it, there is a very high probability it was already swapped ;) | 22:35 |
merlin1991 | also "grep -A 9 heap /proc/`pgrep systemui`/smaps" is way more readable than the huge output to more :D | 22:35 |
freemangordon | :D | 22:35 |
freemangordon | now, we have a couple of options here: | 22:36 |
merlin1991 | theoretically swapping back in has better performance than reading from the the disc ;), plus you still get read of the alloc | 22:36 |
freemangordon | merlin1991: in case swap and fs does not clash :P | 22:37 |
freemangordon | the options: | 22:37 |
freemangordon | 1. RE | 22:37 |
freemangordon | that will be the harest way as there is no x86 binary | 22:37 |
freemangordon | 2. rewrite | 22:37 |
merlin1991 | hm I'd try to snoop the interfaces and rewrite | 22:37 |
kerio | hahaha | 22:37 |
kerio | i uninstalled osso-systemui-tklock | 22:37 |
freemangordon | merlin1991: we have the interfaces | 22:38 |
merlin1991 | but the sms stuff probably is odd todo | 22:38 |
merlin1991 | also for the inlays such as missed calls, msgs and whatnot? | 22:38 |
DocScrutinizer05 | haha, now I see why I get uptimes of weeks while others suffer from swap fragmentation and low memory after 3..5 days - I never use that tklock screen | 22:38 |
merlin1991 | DocScrutinizer05: first time we have a valid reason (even in your book) to replace a closed component :D | 22:38 |
kerio | i can't lock my screen | 22:39 |
DocScrutinizer05 | yes, definitely | 22:39 |
kerio | this was fun :D | 22:39 |
kerio | should i try reinstalling it, or should i just restore the backup? | 22:39 |
merlin1991 | kerio: what happens if you hit the powerbutton and press the lock screen button? | 22:39 |
freemangordon | well, according to IDA tklock uses stuff like sqlite3_exec :) | 22:39 |
merlin1991 | uh | 22:39 |
kerio | merlin1991: the powerkey menu disappears and nothing happens | 22:39 |
merlin1991 | kerio: well reinstall is probably enough | 22:40 |
DocScrutinizer05 | dafaq wut? sqlite3? in tklock? | 22:40 |
freemangordon | yeah | 22:40 |
DocScrutinizer05 | are they mad now? | 22:40 |
merlin1991 | probably for the sms db | 22:40 |
merlin1991 | just my wild guess | 22:40 |
freemangordon | gouys, just do apt-cache show osso-systemui-tklock on your devices ;) | 22:40 |
freemangordon | merlin1991: yeah, most probably | 22:41 |
freemangordon | *guys | 22:41 |
freemangordon | look at the maintainer fueld :P | 22:41 |
freemangordon | *field | 22:41 |
freemangordon | as kerio pointed, there is ext- part | 22:42 |
merlin1991 | kerio: the pkg has no install scripts and just provides a lib so installing should be all you need | 22:42 |
freemangordon | (besides the maintainer is she) | 22:42 |
DocScrutinizer05 | freemangordon: hm? | 22:42 |
DocScrutinizer05 | yeah, ext- part | 22:42 |
freemangordon | seems like it is not written by nokia, but some third party | 22:42 |
merlin1991 | freemangordon: should I contact her? | 22:42 |
DocScrutinizer05 | sure | 22:42 |
freemangordon | merlin1991: yeah, lets try | 22:42 |
DocScrutinizer05 | many things are outsourced | 22:43 |
merlin1991 | freemangordon: I guess the best attempt would be to ask for some information in order to write a replacement | 22:43 |
DocScrutinizer05 | that's one of main reasons not to open source stuff | 22:43 |
freemangordon | merlin1991: hmm, what info do you need? | 22:43 |
DocScrutinizer05 | Nokia doesn't hold (C) | 22:43 |
merlin1991 | freemangordon: well I'm most probably not going to rewrite it ;) | 22:44 |
freemangordon | toldya, we have almost everything we need to write systemui plugins | 22:44 |
freemangordon | if not everything | 22:44 |
merlin1991 | hm why contact then? | 22:44 |
freemangordon | ask for the source code :P | 22:44 |
merlin1991 | we can't possibly get a sourcedump unless she can get her and nokias legal department to relicense the fun | 22:44 |
freemangordon | forget that | 22:45 |
freemangordon | no way to happen | 22:45 |
merlin1991 | and I'm not going to push some under the hand source dump to the cssu repo | 22:45 |
freemangordon | merlin1991: ok, i'll push it :P | 22:45 |
freemangordon | just give it to me :D | 22:46 |
merlin1991 | btw freemangordon where are the docs on what we know about systemui plugins? | 22:46 |
freemangordon | merlin1991: no docs afaik | 22:46 |
freemangordon | well, there is some systemui-dev package | 22:46 |
merlin1991 | that's kinda something I miss about all the re work around maemo, nobody ever documents his findings for others | 22:46 |
freemangordon | with only docs in it, didn't look what exactly is written there | 22:47 |
DocScrutinizer05 | merlin1991: +1 | 22:47 |
kerio | dafuq, what's that new osso-systemui-alarm? | 22:47 |
kerio | freemangordon: ! | 22:47 |
kerio | freemangordon: !!! | 22:47 |
kerio | freemangordon: ^_^ | 22:47 |
arcean | kerio: fixed memory leaks :P I hope ;] | 22:48 |
*** M4rtinK has quit IRC | 22:48 | |
freemangordon | kerio: https://gitorious.org/community-ssu/osso-systemui-alarm/commit/373fbcaaafdedee51744a1b696814fa6555a7670 | 22:48 |
kerio | freemangordon: yay | 22:49 |
freemangordon | merlin1991: yes, you are right re documentation | 22:49 |
freemangordon | on the other hand it is still WIP. alarmui that is | 22:49 |
freemangordon | merlin1991: i can do tklock replacement skeleton, but can't do the GTK part | 22:52 |
*** M4rtinK has joined #maemo-ssu | 22:52 | |
freemangordon | on the other hand... can we use Qt? | 22:52 |
kerio | don't do the gtk part | 22:52 |
merlin1991 | freemangordon: nah | 22:52 |
freemangordon | reason? | 22:52 |
merlin1991 | afaik gtk is in memory all the time | 22:53 |
freemangordon | so? only needed Qt pages will be in memory | 22:53 |
DocScrutinizer05 | don't use qt for system stuff | 22:53 |
freemangordon | ok, ok | 22:54 |
DocScrutinizer05 | my 2 cents | 22:54 |
freemangordon | I know you are right, was just hoping... :D | 22:54 |
merlin1991 | yeah qt is nice to write :) | 22:54 |
merlin1991 | looks a billion times better when you look at the source compared to gtk for me | 22:54 |
freemangordon | though... shortcutd anyone? | 22:54 |
merlin1991 | hm? | 22:55 |
* DocScrutinizer05 waves and hopes for a break | 22:55 | |
freemangordon | it is written in Qt | 22:55 |
merlin1991 | well mags lock screen overlay is written in qt aswell | 22:55 |
freemangordon | arcean | 22:55 |
freemangordon | ? | 22:55 |
DocScrutinizer05 | those are no system replacement stuff though | 22:55 |
freemangordon | wanna do it? | 22:56 |
freemangordon | DocScrutinizer05: :nod: | 22:56 |
DocScrutinizer05 | aiui on my stock n900 there's no qt installed at all | 22:56 |
merlin1991 | DocScrutinizer05: yep | 22:56 |
freemangordon | DocScrutinizer05: no way | 22:56 |
merlin1991 | but cssu always ships qt | 22:56 |
freemangordon | afaik | 22:56 |
DocScrutinizer05 | o/ | 22:57 |
DocScrutinizer05 | --->dinner | 22:57 |
freemangordon | too bad nicolai is no longer interested much in CSSU | 22:57 |
merlin1991 | ah wait pr1.3 depends on qt | 22:57 |
freemangordon | MohammadAG ??? | 22:57 |
freemangordon | MohammadAG: ping | 22:57 |
merlin1991 | but I do remember a time where I had to install qt on maemo in order to run some qt things | 22:58 |
merlin1991 | must have been 1.1 | 22:58 |
DocScrutinizer05 | well, since then they fixed their dependencies ;-P | 22:58 |
freemangordon | yeah, but >=PR12 ships Qt with it | 22:58 |
DocScrutinizer05 | publish a call for GTK-devels | 22:59 |
freemangordon | merlin1991: yeah ^^^ | 23:00 |
DocScrutinizer05 | there have to be 1 or 2 left over | 23:00 |
arcean | DocScrutinizer05: re context menu: http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtkmenu.c#line3175 | 23:04 |
freemangordon | arcean: i owe you a beer. Or whatever you like to drink | 23:04 |
arcean | it's already implemented, but treshold value is low | 23:04 |
freemangordon | this is the next thing in maemo after the alarmui (stock) I hate :) | 23:05 |
arcean | freemangordon: uhm why :D ? | 23:05 |
freemangordon | context menu | 23:05 |
freemangordon | the way it appears | 23:05 |
arcean | yup, it's broken by design (tm) :) | 23:06 |
DocScrutinizer05 | arcean: g_object_get (gtk_widget_get_settings (widget), 3167 "gtk-dnd-drag-threshold", &threshold, 3168 NULL); <--what does that mean? | 23:06 |
arcean | developer can change the treshold value in the application's code | 23:07 |
*** freemangordon_ has joined #maemo-ssu | 23:08 | |
DocScrutinizer05 | err what? it's for fetching a hardcoded constant? | 23:08 |
DocScrutinizer05 | not a gconf key maybe? | 23:09 |
freemangordon_ | The fuck, power outage. What's going on with that recently :-( | 23:09 |
arcean | DocScrutinizer05: no, it's read/write property | 23:09 |
freemangordon_ | Object property | 23:09 |
DocScrutinizer05 | sorry, nfc. gibberish to me | 23:09 |
DocScrutinizer05 | anyway this *should* be a gconf key | 23:10 |
*** freemangordon has quit IRC | 23:10 | |
DocScrutinizer05 | maybe not at this location, maybe somewhere in init | 23:11 |
DocScrutinizer05 | http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtkmenu.c#line3140 | 23:11 |
arcean | so the developer can overwrite our default value, it makes sense :) | 23:11 |
freemangordon_ | arcean: so the default is bad? | 23:12 |
arcean | if you move slightly your finger, the context menu will disappear or even won't show | 23:13 |
freemangordon_ | Yeah, we all know that :-D | 23:14 |
arcean | maybe it's ok with stylus but for sure not with finger | 23:14 |
*** sixwheeledbeast has joined #maemo-ssu | 23:14 | |
freemangordon_ | Hmm, still no power, wtf has happened? | 23:15 |
arcean | heh, the default value is 8 pixels | 23:17 |
DocScrutinizer05 | arcean: the question is: can we fix it? preferably in a way so threshold is a gconf key (or otherwise user configurable)? | 23:17 |
*** freemangordon_ has quit IRC | 23:19 | |
DocScrutinizer05 | arcean: aiui replacing #3166 by a call to read a gconf key should already work | 23:19 |
DocScrutinizer05 | s/work/suffice/ | 23:20 |
infobot | DocScrutinizer05 meant: arcean: aiui replacing #3166 by a call to read a gconf key should already suffice | 23:20 |
arcean | I'll check | 23:21 |
DocScrutinizer05 | hmm, for anchor I found http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtkmenu.c#line2995 so far. will be fun to find where coords to show that popup menu are set | 23:23 |
*** freemangordon has joined #maemo-ssu | 23:25 | |
DocScrutinizer05 | hmmm http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtkmenu.c#line2914 | 23:26 |
*** kolp_ is now known as kolp | 23:26 | |
DocScrutinizer05 | or rather here: http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtkmenu.c#line2947 ? | 23:27 |
*** don_falcone has joined #maemo-ssu | 23:28 | |
DocScrutinizer05 | http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtkmenu.c#line2923 | 23:28 |
*** FredrIQ|n900 has joined #maemo-ssu | 23:28 | |
DocScrutinizer05 | aiui that's for showing menu either above or below touchpoint? | 23:29 |
arcean | the gconf-library depends on gtk, so it's a bit problematic | 23:30 |
DocScrutinizer05 | yeah, when I click in lower right corner of screen, then popup menu will show to the upper left of touchpoint. So it's already all there, just conditions for picking one or the other are crap | 23:32 |
freemangordon | arcean: afaik those properties have default values | 23:32 |
arcean | freemangordon: yes, and it's 8pix by default | 23:33 |
freemangordon | WTF? | 23:33 |
freemangordon | well, lets change that to 128 or something. Problem solved | 23:33 |
arcean | http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtksettings.c#line331 | 23:33 |
DocScrutinizer05 | arcean: why is gconf-lib depends gtk a problem? | 23:34 |
DocScrutinizer05 | isn't that gtk we're looking at ? | 23:34 |
*** FredrIQ|n900 is now known as FIQ|n900 | 23:35 | |
DocScrutinizer05 | arcean: I just can't believe all that stuff in line 331 and before is just a nice way to write a comment. There has to be some function and meaning behind all this | 23:36 |
DocScrutinizer05 | I.E. a way to change the value from default set there | 23:36 |
arcean | DocScrutinizer05: we can change the default value to the another, hardcoded one | 23:37 |
DocScrutinizer05 | like hildon-desktop --gtk-dnd-drag-threshold=65 | 23:37 |
DocScrutinizer05 | nobody would create sucha structure with comment and all, just to read "8" at another point in code | 23:38 |
DocScrutinizer05 | g_param_spec_int sounds pretty much like "specify parameter" to me | 23:39 |
DocScrutinizer05 | parameter in turn makes my think about commandline parameters or configfile parameters | 23:39 |
DocScrutinizer05 | and G_MAXINT is probably the upper limit allowable for that parameter. Who's going to check that, when, and where? | 23:41 |
DocScrutinizer05 | http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtksettings.c#line44 | 23:43 |
DocScrutinizer05 | particularly GTK_SETTINGS_SOURCE_RC_FILE | 23:44 |
*** nox- has joined #maemo-ssu | 23:44 | |
DocScrutinizer05 | I'd suspect sth like this whole crap looking for /etc/gtk/filemanager.rc and in there for a line gtk-dnd-drag-threshold=120 | 23:45 |
* DocScrutinizer05 maybe should ask rasterman, he should know a bit about gtk afaik | 23:47 | |
freemangordon | DocScrutinizer05: not that I am much in GTK, but it most probably looks in the current theme | 23:48 |
DocScrutinizer05 | or just scrutinize that code a tad longer, and find out where else GTK_SETTINGS_SOURCE_RC_FILE is used | 23:48 |
DocScrutinizer05 | theme, YEAH | 23:48 |
DocScrutinizer05 | sounds about right | 23:48 |
freemangordon | "/usr/share/themes/alpha/..." or whatever the path was | 23:49 |
DocScrutinizer05 | freemangordon: look at http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtksettings.c#line44 please. What's the ideas that gives you? | 23:50 |
DocScrutinizer05 | http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtksettings.c#line194 hmm | 23:52 |
freemangordon | I am too tired right now to grok that | 23:52 |
arcean | hmm, gtk-dnd-drag-threshold is defined in my gtkrc file | 23:52 |
freemangordon | well, default should come from the theme | 23:52 |
freemangordon | arcean: and what is the value? | 23:53 |
DocScrutinizer05 | http://gitorious.org/community-ssu/gtk/blobs/master/gtk/gtksettings.c#line533 | 23:53 |
arcean | freemangordon: 30pix | 23:53 |
arcean | default theme | 23:53 |
freemangordon | does not seem used | 23:54 |
freemangordon | so it could be in a rogn section or something | 23:54 |
freemangordon | *wrong | 23:54 |
DocScrutinizer05 | arcean: where's that file? (path) | 23:54 |
arcean | DocScrutinizer05: /usr/share/themes/default/gtk-2.0 | 23:55 |
DocScrutinizer05 | thnaks | 23:55 |
freemangordon | arcean: no matter what, we should change the default | 23:55 |
freemangordon | anyway, I need some sleep | 23:58 |
freemangordon | night | 23:58 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!