| Venemo_N950 | also happy you didn't see it yesterday :P yesterday's version was garbage | 00:00 |
|---|---|---|
| blueslee_ | i was happy to use the old one, it was the only usable one with colours and it is simple, exact what i needed | 00:00 |
| Venemo_N950 | :) | 00:01 |
| Venemo_N950 | I tried to keep it simple | 00:01 |
| djszapi | I saw it yesterday, and I took screenshots bruhaha :D | 00:01 |
| Venemo_N950 | djszapi, :) | 00:01 |
| *** blueslee_ has quit IRC | 00:02 | |
| *** blueslee has quit IRC | 00:03 | |
| *** blueslee has joined #harmattan | 00:03 | |
| blueslee | Venemo_N950: where ate the bugs? | 00:04 |
| Venemo_N950 | blueslee, it doesn't handle connection failures well + it doesn't always handle disconnections well | 00:05 |
| Venemo_N950 | and changing the color doesn't work for djszapi | 00:05 |
| Venemo_N950 | and I can get the manage servers page into a buggered state | 00:06 |
| djszapi | + long lines are not visible. | 00:06 |
| blueslee | is he a chameleon or why he wants to change his colour?;-) | 00:06 |
| djszapi | + timestamping is buggy | 00:06 |
| Venemo_N950 | they are now, aren't they, djszapi? | 00:06 |
| *** stlpaul has joined #harmattan | 00:07 | |
| Venemo_N950 | djszapi, timestamping only takes effect for new messages (after you switch the setting) | 00:07 |
| Venemo_N950 | blueslee :P | 00:07 |
| djszapi | Venemo_N950: which is not a good imo | 00:07 |
| blueslee | okay, i will give you my feedback | 00:07 |
| djszapi | + settings page is a bit strange, to say the least. | 00:08 |
| blueslee | the issue with the long lines is also present in other apps | 00:08 |
| Venemo_N950 | djszapi, perhaps not, but it would take effort to remove all timestamps from the html. | 00:08 |
| djszapi | should be fairly simple, right ? | 00:08 |
| Venemo_N950 | qwertzuiopasdfghjklyxcvbnm | 00:08 |
| *** Free-MG has quit IRC | 00:08 | |
| Venemo_N950 | qwertzuiopasdfghjklyxcvbnmqwertzuiopasdfghjklyxcvbnmqwertzuiopasdfghjklyxcvbnm | 00:09 |
| Venemo_N950 | it wraps for me | 00:09 |
| djszapi | Venemo_N950: you can practice it in a dummy channel :D | 00:09 |
| Venemo_N950 | djszapi, I usually use #mychannel, but it seems to work now. | 00:09 |
| jonni | blueslee: yes | 00:10 |
| djszapi | Venemo_N950: joining a channel should be in joining order imo, not in alphabetical order | 00:11 |
| djszapi | otherwise it is pretty hard to get my favourite channel in the beginning... | 00:12 |
| blueslee | Venemo_N950, i even get a notification on the lock screen | 00:12 |
| Venemo_N950 | djszapi, yeah, I plan to make them rearrangeable | 00:12 |
| Venemo_N950 | djszapi, there are also two more things I need to refactor in the code :) | 00:12 |
| djszapi | not rearrangable what I meant | 00:13 |
| *** Swerige has quit IRC | 00:13 | |
| djszapi | the same ordering as in irssi | 00:13 |
| Venemo_N950 | plus, I'm surprised that for half a year, nobody noticed the limitations of the autocompletion implementation. :) | 00:13 |
| blueslee | jonni, i want to give you some feedback concerning bug 123 | 00:13 |
| djszapi | new channels and queries go to the end. | 00:13 |
| _MeeGoBot_ | Bug https://harmattan-bugs.nokia.com/show_bug.cgi?id=123 maj, Highest, ---, ext-risto.lahti, REOP, accounts passwords are not remembered accross reboots | 00:13 |
| Venemo_N950 | djszapi, not even you | 00:14 |
| djszapi | "not even you" ? :) | 00:14 |
| tehdely | i think today is the day | 00:14 |
| tehdely | today is the day that i will flash my N9 and void its warranty | 00:14 |
| jonni | blueslee: ah yes that one, if you visited rescue kernel while beeing on PR1.2 (even with -b flag), then on next reboot, you are actually on open mode, untill complete reflash. That would explain why your accouts data dont work since most likely data was encrypted with closed mode keys, vs after reboot you are actually using open mode keys. | 00:15 |
| tehdely | that's precisely what i'm about to do! | 00:17 |
| tehdely | btw is there a 1.2 rescue kernel yet | 00:17 |
| Venemo_N950 | djszapiN9, yes, you have very keen eyes for bugs, thus I expected you to find this one sooner or later :P but you haven't | 00:17 |
| tehdely | i only need to boot into one for long enough for it to make the phone realize i've messed with it | 00:17 |
| djszapi | Venemo_N950: I am pretty sure I found :P | 00:17 |
| jonni | tehdely: you can use pr1.1 or whatever rescue kernel which is on fix malf page | 00:17 |
| djszapi | you probably lost it among the lines ;) | 00:17 |
| Venemo_N950 | djszapi, no, you have not mentioned it to me yet | 00:18 |
| djszapi | :D :D | 00:18 |
| tehdely | jonni: i assume you're refering to this page: http://wiki.meego.com/N950/Fixing_MALF_state | 00:18 |
| jonni | yep | 00:18 |
| *** niqt has quit IRC | 00:19 | |
| tehdely | itsnotabigtruck had some specific directions as far as the ordering | 00:19 |
| tehdely | i want to do a clean flash of 1.2 and then immediately boot that rescue kernel | 00:19 |
| tehdely | can i do this: flasher -f -F main.bin -F emmc.bin --flash-only=mmc | 00:19 |
| tehdely | immediately followed by this: $ sudo flasher --load -k vmlinuz-2.6.32.20112201-11.2-adaptation-n950-bootloader -n initrd.img-rescue-2.6.32.20112201-11.2-n950 --boot | 00:20 |
| *** rzr is now known as rZr | 00:21 | |
| tehdely | this should be fun :> | 00:21 |
| tehdely | found the logs... he suggested a reboot in between | 00:21 |
| tehdely | http://mg.pov.lt/harmattan-irclog/%23harmattan.2012-03-01.log.html | 00:21 |
| tehdely | booya | 00:21 |
| blueslee | jonni, i was on pr 1.1 with notifications that pr 1.2 is available, flashed the rescue kernel for a final backup and after that i updated to pr 1.2 and the issue arised .. flashing again the rescue image was my attempt to repair it | 00:22 |
| jonni | tehdely: if you reboot in between it generates aegis storages with closed mode keys, if you really want fully clean open mode you should not reboot in between, but instead run some other command in between so -l -b would work | 00:23 |
| Venemo_N950 | so anyway, good evening guys! | 00:24 |
| tehdely | i want fully clean, because i want saved accounts to wrok | 00:24 |
| *** Venemo_N950 has quit IRC | 00:24 | |
| tehdely | in what state is the phone after it's been flashed? | 00:24 |
| tehdely | i am led to believe it requires a reboot before more can be done to it | 00:24 |
| tehdely | or could you just flash it 5x in a row and it wouldn't really care | 00:24 |
| jonni | blueslee: so you did somekind backup/restore? I've never done any restores since that usually causes problems | 00:24 |
| jonni | tehdely: I've had a script somewhere which allowed it to -l -b without reboot... but dont remember where my script is | 00:25 |
| tehdely | what was the complication that you had to resolve | 00:25 |
| tehdely | as far as being able to do one after the other | 00:25 |
| jonni | lets see if I can find my script, it was some flasher command, then sleep 1.5 and then running the flasher -l -b.... | 00:26 |
| ieatlint | to anyone else who may be wondering... despite the qtmobility docs now saying that harmattan is a completely supported platform for connectivity/nfc, it still indeed does not support tag type specific access | 00:27 |
| ieatlint | here's to hoping there's a pr1.3, and that it does solve this | 00:27 |
| djszapi | 1.3 is a bugfix release. | 00:28 |
| ieatlint | i can have my hopes | 00:29 |
| tehdely | some more Qs | 00:29 |
| tehdely | i did an OTA update to 1.2 when I got my phone; it was nearly the first thing i did | 00:29 |
| tehdely | (it still had issues) | 00:29 |
| djszapi | ieatlint: just if you still wear the pants. | 00:29 |
| tehdely | i'm assuming that all that was going on behind the scenes was an apt-get dist-upgrade | 00:29 |
| tehdely | now, user-installed packages are also installed using dpkg and their manifests stored in the same database, right? | 00:30 |
| ieatlint | djszapi, actually most of my hopes seem to involve not wearing pants | 00:30 |
| tehdely | so how can you flash a newer version (assuming you _don't_ erase the EMMC) without wiping out software you've installed | 00:30 |
| tehdely | or is there some sort of overlay where "system" and "user | 00:30 |
| tehdely | " packages are actually stored on different bits of flash.. | 00:30 |
| tehdely | that would be very Symbianesque if so | 00:30 |
| *** dfaure has quit IRC | 00:31 | |
| *** ajalkane has quit IRC | 00:32 | |
| *** machine1 has quit IRC | 00:32 | |
| blueslee | jonni, i hoped to do so but after flashing the rescue images with pr 1.2 the well known problem with the warnings and reboot cycles arised | 00:33 |
| *** piggz_ has quit IRC | 00:33 | |
| blueslee | i had no chance to restore backups | 00:33 |
| jonni | tehdely: if I remember right it was something like flasher -i; flasher -R;sleep 1.5;flasher -l -b... after the flash command. | 00:33 |
| *** piggz_ has joined #harmattan | 00:33 | |
| Corsac | btw, I sent my N9 to warranty before pr1.2 was out, when I get the device back should I upgrade before or after restoring backups? | 00:33 |
| jonni | Corsac: I would just setup accounts by hands and not using any restore functions, but thats just me. | 00:34 |
| *** machine1 has joined #harmattan | 00:35 | |
| djszapi | +1 here. | 00:35 |
| jonni | blueslee: what do you mean "flashing rescue images" as I dont know what images you mean on that sentanse? | 00:35 |
| Corsac | jonni: I don't really care about accounts, but I do care about sms, contacts and calendar | 00:36 |
| Corsac | it should be possible to push contacts and calendar using syncml, but it's not possible for smses | 00:36 |
| jonni | Corsac: well you should restore on the same pr that you generated the backup with. | 00:36 |
| Corsac | ok | 00:36 |
| blueslee | jonni, rescue kernel for n950 | 00:37 |
| tehdely | once i've booted that rescue kernel btw, should i just reboot again | 00:38 |
| jonni | blueslee: flashed or load booted? | 00:38 |
| admiral0 | itsnotabigtruck: it works! | 00:38 |
| tehdely | that should be enough to flip the bit that sends the phone into Open Mode right | 00:38 |
| jonni | blueslee: if you flashed it, then you are never supposed to get into an UI, as rescure kernels only job is to enable usb device without ui. | 00:39 |
| djszapi | ieatlint: okay, it seems I am also beyond my first qt5 patch accepted, yay \o/ | 00:39 |
| tehdely | will there be a qt5 dance? | 00:39 |
| tehdely | i remember the qt4 dance fondly... | 00:39 |
| ieatlint | congrats! :) | 00:39 |
| *** piggz_ has quit IRC | 00:40 | |
| djszapi | tehdely: I can dance crazily without Qt, too :) | 00:40 |
| blueslee | jonni, right, i made dd copies of the three partitions with that method | 00:40 |
| jonni | blueslee: (all the instructions say that you are not supposed to flash the rescue kernel, only load boot). | 00:40 |
| *** araujo has quit IRC | 00:41 | |
| jonni | blueslee: and if you did dd copies of them in pr1.1 and did dd restore to pr1.2 you will get malf everytime, since aegis checksums will fail. | 00:41 |
| admiral0 | itsnotabigtruck: ping | 00:42 |
| blueslee | jonni, add i said restoring will not work, i did a --load ... --boot with the flasher | 00:45 |
| jonni | blueslee: (as if you dd restore pr1.1 rootfs to pr1.2 /etc/aegis_vdata will have missmatch to running kernel, you cannot dd copy/restore pr1.1 rootfs to pr1.2) | 00:45 |
| tehdely | damn it i should have downloaded the rofs earlier | 00:47 |
| tehdely | slow Nokia Care Suite server is slow | 00:47 |
| djszapi | yes | 00:48 |
| blueslee | jonni, now i also know that its not possible | 00:48 |
| tehdely | is it ok i flash an EMMC for a 16GB product code to my 64GB phone | 00:49 |
| tehdely | i want the 059J1Z5 (rest of the world) one | 00:49 |
| tehdely | i don't really need scandianavian maps :> | 00:49 |
| tehdely | oh wait, there is a 64GB one | 00:50 |
| blueslee | jonni, what do you think about reopening bug 123? | 00:51 |
| _MeeGoBot_ | Bug https://harmattan-bugs.nokia.com/show_bug.cgi?id=123 maj, Highest, ---, ext-risto.lahti, REOP, accounts passwords are not remembered accross reboots | 00:51 |
| blueslee | jonni, i think the problem is related to a missing internet connection, that's my opinion | 00:53 |
| itsnotabigtruck | admiral0: back online | 00:54 |
| admiral0 | ok | 00:54 |
| tehdely | itsnotabigtruck: hey-yo | 00:54 |
| tehdely | i'm going to try the clean flash + boot into rescue kernel today | 00:54 |
| tehdely | we were discussing whether it woudl be possible to execute the load command after the flash command without rebooting the phone in between | 00:55 |
| tehdely | i'm going to assume that "no" | 00:55 |
| itsnotabigtruck | tehdely: hmm | 00:55 |
| *** Natunen has quit IRC | 00:55 | |
| blueslee | good night guys, i am tired | 00:55 |
| itsnotabigtruck | i don't think anyone's actually tested it, so i have no idea if it's required | 00:55 |
| tehdely | night | 00:55 |
| *** blueslee has quit IRC | 00:56 | |
| tehdely | because if i reboot into pr1.2 without having booted into rescue kernel, won't it go ahead and generate the keys for account storage/etc. | 00:56 |
| itsnotabigtruck | after flashing the image, you're still booted into the "ape-algo" (device flasher) | 00:56 |
| itsnotabigtruck | but if you boot from that, i'm pretty sure it sends you back to the bootloader to boot | 00:56 |
| itsnotabigtruck | the question is, does it use the previous bootloader or the new bootloader | 00:56 |
| itsnotabigtruck | probably the latter, which means that you don't need the reboot in between | 00:56 |
| tehdely | good Q | 00:56 |
| tehdely | so flash, then load | 00:57 |
| tehdely | i'm going to try that and see what happens | 00:57 |
| itsnotabigtruck | yeah, and if the disclaimer screen doesn't pop up, then run flasher -i again and unplug, then immediately replug the usb | 00:57 |
| itsnotabigtruck | (so that you catch it on the reboot)( | 00:57 |
| pa | has anybody here ever used syncevolution on N800? | 00:58 |
| pa | (or also N9/950) | 00:58 |
| tehdely | i will be very curious, btw, if i can get gogoc ported to the phone | 01:00 |
| tehdely | and have it run every time it associates with a WAP | 01:00 |
| tehdely | evne better, have it work over 3G | 01:00 |
| tehdely | so i can have IPv6 everywhere :> | 01:00 |
| tehdely | it reuqires a tun/tap device | 01:01 |
| tehdely | i wonder what's the difference between the ROW and SWAP firmwares, btw | 01:01 |
| tehdely | probably nothing | 01:01 |
| *** djszapi has left #harmattan | 01:02 | |
| *** decoy has quit IRC | 01:06 | |
| *** decoy has joined #harmattan | 01:06 | |
| itsnotabigtruck | tehdely: nothing | 01:06 |
| itsnotabigtruck | if the name's the same, it's the same firmware | 01:06 |
| itsnotabigtruck | same as that | 01:06 |
| *** marsje has quit IRC | 01:06 | |
| itsnotabigtruck | (well, if it's coming from nokia, since nokia uses a systematic naming scheme) | 01:07 |
| itsnotabigtruck | all 001 images for pr1.2 are the same | 01:07 |
| itsnotabigtruck | the other files aren't necessarily the same | 01:07 |
| itsnotabigtruck | but the other files aren't used by flasher, they're mostly for label printing and calibration | 01:08 |
| *** marsje has joined #harmattan | 01:08 | |
| tehdely | ok, you can not boot a kernel after the flashing mode, you do have to reboot | 01:29 |
| *** admiral0 has quit IRC | 01:29 | |
| *** admiral0 has joined #harmattan | 01:29 | |
| tehdely | man, second time this has happened | 01:32 |
| tehdely | i can't seem to get the phone not to boot | 01:32 |
| tehdely | it's going into setup again | 01:32 |
| DocScrutinizer | could someone check if mtd_debug is available on N9? | 01:33 |
| DocScrutinizer | I got it on N950 at least | 01:34 |
| tehdely | itsnotabigtruck: flasher -R reboots; it doens't power off | 01:35 |
| tehdely | is there any way i can power off the phone after flash | 01:35 |
| tehdely | i can't seem to 'catch' it early enough in the boot cycle on the next boot | 01:35 |
| DocScrutinizer | wtf is >>ConF^B^@^@^@cert-sw-enos^@^@<< and >>Mrmxxx_enos_0.2011.12-5_005.bin^@<< ? | 01:36 |
| DocScrutinizer | (/dev/mtd1 if you wonder) | 01:37 |
| *** djszapiN9 has quit IRC | 01:38 | |
| pa | how does it work to request an N950? | 01:39 |
| *** djszapi has joined #harmattan | 01:39 | |
| DocScrutinizer | there are no more left over | 01:43 |
| pa | ah i see | 01:45 |
| pa | but is N9 still produced, or that went out as well? | 01:45 |
| djszapi | still produced. | 01:46 |
| pa | here it seems it's on sale everywhere, and no one restocks | 01:46 |
| ieatlint | the n9 targets very specific audiences very well... but it's difficult for sales people to sell, when it's a dead end device that lacks the apps that android/iphone have | 01:47 |
| tehdely | itsnotabigtruck: so i put it in open mode after it had already booted into PR1.2 once | 01:47 |
| tehdely | it went as far as asking me for my SIM PIN before i powered off | 01:47 |
| ieatlint | my guess is no one is restocking it because they saturated the geek market | 01:47 |
| tehdely | but i just added a facebook account and rebooted, so let's see if it prompts for the password again | 01:48 |
| tehdely | if it doesn't, i should be in the clear... | 01:48 |
| tehdely | nope, didn't remember the password :< | 01:48 |
| djszapi | though the pants business is good, too | 01:49 |
| ieatlint | pants are overrated | 01:50 |
| DocScrutinizer | eh? | 01:50 |
| DocScrutinizer | pants? | 01:51 |
| ieatlint | yes, pantaloons | 01:51 |
| djszapi | xD | 01:51 |
| tehdely | ok i'm reflashing into pr1.2 again | 01:51 |
| * DocScrutinizer blames his ignorelist and forgets about it | 01:51 | |
| pa | go for kilt | 01:51 |
| djszapi | flashing 1.3 is funnier | 01:52 |
| tehdely | i'll pass on that | 01:52 |
| tehdely | i just wanted to do the magic two-step "clean flash, followed immediately by boot into rescue kernel" | 01:52 |
| tehdely | but i can't seem to interrupt the first boot into pr1.2 | 01:52 |
| tehdely | i want to flash, reboot, immediately load kernel... but no can do | 01:52 |
| DocScrutinizer | if itsnotabigtruck was correct in his description how that secure-bit works, I don't see how doing this 'magic 2step' can help to accomplish anything | 01:54 |
| tehdely | the magic 2step is designed to flip that bit | 01:54 |
| tehdely | but avoid the phone having ever booted into a non-"open" mode | 01:54 |
| DocScrutinizer | why would that be needed? | 01:54 |
| tehdely | so that it hasn't generated the aegis security keys | 01:55 |
| tehdely | they'll be incompatible with the ones in open mode | 01:55 |
| DocScrutinizer | errr | 01:55 |
| tehdely | and i wno't be able to save any passwords | 01:55 |
| tehdely | i just tested that that is in fact the case | 01:55 |
| tehdely | one boot into PR1.2 is one too many | 01:55 |
| DocScrutinizer | I'd guess zero boots into secure mode are one too many | 01:55 |
| DocScrutinizer | but maybe that may actually work | 01:55 |
| tehdely | it's a race | 01:56 |
| DocScrutinizer | not really | 01:56 |
| tehdely | i wish there were a flasher command to unequivocally say "POWER THE FUCK OFF" | 01:56 |
| tehdely | i want to power it off, unplug, run flasher -i again, replug, and be back into flashing mode | 01:56 |
| tehdely | but it stubbornly insists on booting | 01:56 |
| tehdely | i'm moving hella fast, too | 01:56 |
| DocScrutinizer | simple: switch off device after flashing, then prepare flasher with load-and-execute-kernel and fire, then plug in the powered-down device to USB | 01:57 |
| tehdely | how do i switch it off | 01:57 |
| tehdely | it's plugged in right now | 01:57 |
| tehdely | it's just been flashed | 01:57 |
| tehdely | is it safe to unplug | 01:57 |
| DocScrutinizer | you unplug it and press hold power button for >20s | 01:57 |
| tehdely | and i don't need to run 'flasher -i'? | 01:58 |
| tehdely | the whole load-and-execute shebang will also wait for the phone to be plugged in? | 01:58 |
| DocScrutinizer | sure, why not? | 01:58 |
| tehdely | sorry, i've only owned the device for a few days | 01:58 |
| tehdely | still picking this up :) | 01:58 |
| tehdely | alright, that worked. i was able to power it off | 01:59 |
| tehdely | now let's see if i can get a saved password in open mode | 01:59 |
| tehdely | that is a very pretty warning! | 02:00 |
| DocScrutinizer | this aegis stuff still spoils your day even while on openmode :-S | 02:00 |
| tehdely | i believe you | 02:00 |
| tehdely | it's alright | 02:01 |
| tehdely | the phone is so damn pretty i | 02:01 |
| tehdely | 'm willing to overlook a few inconveniences ;) | 02:01 |
| DocScrutinizer | after loading kernel to ram, you might have to power up the device again. do NOT unplug as it will shutdown and kernel in ram will evaporate | 02:03 |
| tehdely | oh i unmounted, unplugged, and booted into pr1.2 | 02:03 |
| tehdely | i only needed to load the kernel in RAM for long enough to flip that bit | 02:03 |
| DocScrutinizer | aaah I see | 02:03 |
| *** longthen has quit IRC | 02:03 | |
| tehdely | it was nice seeing it mount though. it actually mounted with a pretty icon on OS X | 02:04 |
| tehdely | i should've taken a screenshot | 02:04 |
| tehdely | it looked like an N9 ;) | 02:04 |
| djszapi | tehdely: oh, you also a Mac user ? | 02:04 |
| tehdely | yep | 02:04 |
| DocScrutinizer | yep what? | 02:04 |
| djszapi | nifty :) | 02:04 |
| tehdely | DocScrutinizer: "yep" to djszapi | 02:05 |
| tehdely | you got him on ignore or something? :> | 02:05 |
| DocScrutinizer | yup | 02:05 |
| *** longthen has joined #harmattan | 02:05 | |
| tehdely | ha | 02:05 |
| tehdely | man i need some pho | 02:08 |
| djszapi | tehdely: do you mind giving some help with qt5 patch testing on future on Mac ? | 02:08 |
| ieatlint | mmm, pho | 02:08 |
| tehdely | depends what kind of help | 02:09 |
| itsnotabigtruck | tehdely: did you manage to do it? | 02:09 |
| djszapi | tehdely: as said, testing. | 02:09 |
| tehdely | itsnotabigtruck: yes, now the "can we save a password across reboots" test | 02:09 |
| itsnotabigtruck | btw flasher -R says it's supposed to reboot but IMX it always shuts off | 02:09 |
| djszapi | build + unit test run. | 02:09 |
| tehdely | djszapi: perhaps. mike@teambarry.org <- me | 02:09 |
| tehdely | i run lion | 02:09 |
| tehdely | if you need someone building on lion, i'm your man | 02:09 |
| djszapi | :) | 02:09 |
| itsnotabigtruck | tehdely: you just put your unadulterated email address on the net, so you'll probably be getting moar spams now | 02:09 |
| tehdely | oh man | 02:09 |
| tehdely | that email address is already ruined | 02:09 |
| itsnotabigtruck | lol | 02:09 |
| DocScrutinizer | itsnotabigtruck: nope, it just reboots to act-dead mode | 02:09 |
| tehdely | i had an account on the MtGox bitcoin exchange | 02:09 |
| *** heymaste_ has quit IRC | 02:09 | |
| tehdely | and their whole accounts db was stolen | 02:10 |
| itsnotabigtruck | DocScrutinizer: hrm, ok | 02:10 |
| djszapi | -R does reboot. | 02:10 |
| tehdely | i get 419s like you wouldn't believe | 02:10 |
| tehdely | I guess someone figured Bitcoin users for a bunch of gullible pawns who would easily fall prey to a get-rich-quick scheme why I never... | 02:10 |
| djszapi | tehdely: feel free to any working one to me in private. | 02:10 |
| djszapi | send* | 02:10 |
| *** araujo has joined #harmattan | 02:10 | |
| tehdely | that's a perfectly good email, use that one | 02:11 |
| tehdely | ok let's reboot and see if the phone can log into facebook | 02:11 |
| djszapi | tehdely: you have just said "it is ruined" | 02:12 |
| djszapi | is it for something else then ? | 02:12 |
| tehdely | no i just mean spammers have it | 02:12 |
| tehdely | but zimbra does a good job of filtering it out | 02:12 |
| tehdely | i use that email plenty | 02:12 |
| tehdely | i appear to be logged into facebook! yay! | 02:14 |
| tehdely | itsnotabigtruck: ça marche! | 02:14 |
| DocScrutinizer | \o/ | 02:14 |
| DocScrutinizer | so I already start to wonder if it needs to be any kernel that you load to RAM and execute. Or if NOLO will flip that bit on trying to load some crap from /dev/urandom already | 02:16 |
| tehdely | what's the default pass for 'user' on newly-installed sshd | 02:16 |
| tehdely | or do i have to set one using passwd first | 02:16 |
| DocScrutinizer | rootme | 02:17 |
| djszapi | nothing | 02:17 |
| djszapi | just use developer | 02:17 |
| DocScrutinizer | ooh user. none set | 02:17 |
| tehdely | what is the use of the developer vs. user account | 02:17 |
| djszapi | after enabling the developer mode. | 02:17 |
| tehdely | which one should i be sshing into | 02:17 |
| djszapi | developer | 02:17 |
| DocScrutinizer | so yeah, you have to set one using passwd user in rootsh | 02:17 |
| tehdely | i'd rather set no pass and just use keys | 02:17 |
| tehdely | i wonder if the passwd binary on here supports the '-u' flag for unlock | 02:18 |
| tehdely | we'll find out! | 02:18 |
| *** risca has quit IRC | 02:18 | |
| DocScrutinizer | for some strange reason sshd refuses to allow login with keys on an account that has no password | 02:18 |
| tehdely | even after running 'passwd -u'? | 02:18 |
| tehdely | that's usually all that needs to be done | 02:18 |
| tehdely | so let's see if that worked | 02:18 |
| tehdely | hmm, looks like i need to create developer's home dir | 02:19 |
| tehdely | what is the use of the developer account vs user account | 02:19 |
| tehdely | why would i want to use one vs. the other | 02:19 |
| tehdely | 'user' is the one that my personal stuff runs as, yes? | 02:19 |
| DocScrutinizer | developer account is for downlink from SDK on PC afaik | 02:20 |
| djszapi | you do not need to create that | 02:20 |
| djszapi | just enable the developer mode. | 02:20 |
| DocScrutinizer | yes | 02:20 |
| djszapi | yes | 02:20 |
| *** benares_98 has joined #harmattan | 02:21 | |
| djszapi | but as a rule of thumb, I would suggest using the developer account over ssh. | 02:21 |
| tehdely | and what is the advantage thereof | 02:21 |
| djszapi | plus creds | 02:22 |
| djszapi | less trouble. | 02:22 |
| djszapi | working "by default" without assign password and so forth. | 02:22 |
| djszapi | assigning* | 02:22 |
| tehdely | ok confirm i can get in without setting a pass | 02:22 |
| tehdely | just have to 'passwd -u' | 02:23 |
| djszapi | never tried | 02:23 |
| tehdely | THIS IS UNIX I KNOW THIS | 02:23 |
| tehdely | ;) | 02:23 |
| djszapi | I normally set up my preferred password. | 02:23 |
| djszapi | for developer, that is. | 02:23 |
| tehdely | developer doesn't seem to have a home dir | 02:23 |
| djszapi | it does have. | 02:23 |
| tehdely | nope, no /home/developer on here | 02:24 |
| djszapi | sounds like a screwed up system | 02:24 |
| djszapi | if the developer mode is on, it should be there. | 02:24 |
| tehdely | let's install some more of the devmode packages | 02:24 |
| tehdely | maybe one of htem creates it | 02:25 |
| DocScrutinizer | it's just for SDK to log in | 02:25 |
| djszapi | so what is in the /home folder then ? | 02:25 |
| tehdely | just /home/user | 02:25 |
| djszapi | user and lostnfound only ? | 02:25 |
| tehdely | yep | 02:25 |
| tehdely | may also ahve to do with the emmc i used to flash | 02:25 |
| tehdely | duno | 02:25 |
| djszapi | and you can log in as developer ? | 02:25 |
| tehdely | no i can't | 02:25 |
| tehdely | and i can't create /home/developer and chown it developer either | 02:25 |
| djszapi | you have the developer mode enabled ? | 02:25 |
| tehdely | yes, i'm sshed in righ tnow | 02:25 |
| djszapi | and it installed the stuff fine, right ? | 02:25 |
| tehdely | yep | 02:25 |
| DocScrutinizer | forget about developer account | 02:25 |
| djszapi | disable developer mode, reboot, enable again | 02:26 |
| tehdely | also, how do i get opensh installed | 02:26 |
| tehdely | so far i don't seem to be benefiting from open mode at all.. what has this actually enabled me to do | 02:26 |
| djszapi | http://maemo.cloud-7.de/HARM/N9/openmode_kernel_PR1.1/opensh/ | 02:26 |
| tehdely | where do i get the kernel in question | 02:28 |
| tehdely | oh, one dir up | 02:28 |
| tehdely | is it ok to boot with a kernel intended for pr1.1 | 02:28 |
| djszapi | well, that should be nice to get updated | 02:29 |
| djszapi | but there was no proper effort yet apparently. | 02:30 |
| djszapi | at least it is not shared then in there. | 02:30 |
| itsnotabigtruck | so tehdely: any issues? | 02:30 |
| * itsnotabigtruck reads sb | 02:30 | |
| tehdely | i have no /home/developer | 02:30 |
| tehdely | that seems to be it so far | 02:30 |
| tehdely | i'mma try to get opensh on here | 02:30 |
| tehdely | this should be fun | 02:30 |
| itsnotabigtruck | tehdely: don't boot with a pr1.1 kernel | 02:31 |
| itsnotabigtruck | unless you downgrade the kernel modules too | 02:31 |
| tehdely | oh, don't? | 02:31 |
| itsnotabigtruck | but warning: that will screw up the camera | 02:31 |
| itsnotabigtruck | don't do it | 02:31 |
| tehdely | dang | 02:31 |
| tehdely | even just to get into sdk mode? | 02:31 |
| itsnotabigtruck | don't bother changing the kernel | 02:31 |
| itsnotabigtruck | a) turn on developer mode | 02:31 |
| itsnotabigtruck | b) download doc's opensh | 02:31 |
| itsnotabigtruck | c) AEGIS_FIXED_ORIGIN=com.nokia.maemo dpkg -i opensh.deb | 02:31 |
| itsnotabigtruck | as root | 02:31 |
| tehdely | that is all | 02:31 |
| tehdely | i don't need to be booted into the hacked kernel in order for that to work? | 02:32 |
| itsnotabigtruck | then opensh -c 'accli -I' and be in awe of all the caps | 02:32 |
| itsnotabigtruck | nope, you don't | 02:32 |
| djszapi | C) should not work at all | 02:32 |
| djszapi | if that works, that is just a bug | 02:32 |
| itsnotabigtruck | last time i checked, dpkg knows whether you're in open mode | 02:32 |
| tehdely | seems to have worked | 02:32 |
| djszapi | right, time to report it... | 02:32 |
| tehdely | oh my | 02:32 |
| itsnotabigtruck | tehdely: well, if it doesn't work it gets ignored | 02:32 |
| tehdely | i sure do seem to have a lot of privileges | 02:32 |
| itsnotabigtruck | djszapi: somehow i doubt that was a bug given that logic for it was WRITTEN RIGHT IN | 02:33 |
| tehdely | ahhh, this is MUCH BETTER | 02:33 |
| tehdely | mwahahahaha | 02:33 |
| djszapi | itsnotabigtruck: I am quite amazed how you know better how we designed aegis things. | 02:33 |
| djszapi | this is not the first try, and I am getting tired of you, quite frankly. | 02:33 |
| djszapi | time* | 02:33 |
| *** DocScrutinizer has quit IRC | 02:34 | |
| itsnotabigtruck | djszapi: i'm quite amazed at how often you contradict how it actually works, having worked on it... | 02:34 |
| tehdely | cool i can now ssh in key-based as develoepr or user | 02:35 |
| tehdely | yay, i own my phone | 02:35 |
| tehdely | i know you guys are busy bickering but i want to thank you both | 02:35 |
| tehdely | :) | 02:35 |
| itsnotabigtruck | heh | 02:35 |
| tehdely | and DocScrutinizer51 | 02:35 |
| djszapi | tehdely: this method hopefully is not going to work in the future though. | 02:35 |
| tehdely | well we'll just have to resort to a new one when that happens | 02:35 |
| djszapi | since it is a bug, it can be done at the momment. | 02:35 |
| itsnotabigtruck | :D | 02:35 |
| tehdely | there is still the problem, btw, that no appropriate libc6-dbg exists in the repos | 02:36 |
| tehdely | at least not one that matches the running kernel version | 02:36 |
| tehdely | so i can't install bunch of the devtools pkgs | 02:36 |
| tehdely | libc6-dbg: Depends: libc6 (= 2.10-0maemo18.1+0m6) but 2.10-0maemo20+0m7 is to be installed | 02:37 |
| djszapi | jonni ping | 02:37 |
| djszapi | tehdely: you are right, I should not bicker, just ignore.. | 02:40 |
| djszapi | if someone wanna know the things better than the authors, I cannot do too much. :/ | 02:40 |
| tehdely | i wonder if Track & Protect works in Open Mode | 02:42 |
| itsnotabigtruck | djszapi: take one look at the source of dpkg and you'll see what i mean | 02:42 |
| tehdely | i was thinking of signing up | 02:42 |
| djszapi | jonni please let met know, if you are back. | 02:42 |
| djszapi | when* | 02:42 |
| itsnotabigtruck | i'm not claiming i know more about aegis than you, i am claiming that maybe you need to double check on this one | 02:42 |
| djszapi | itsnotabigtruck: why do you hijack ? | 02:42 |
| djszapi | of course the dpkg source code is first of all not available anymore | 02:42 |
| djszapi | so this is the entry point where you are already wrong | 02:43 |
| djszapi | secondly, I mentioned the damn design | 02:43 |
| tehdely | that would be a GPL violation | 02:43 |
| djszapi | that me and my colleague decided together... | 02:43 |
| djszapi | but surely, you know better than we, of course /sighs/. | 02:43 |
| djszapi | tehdely: why ? | 02:43 |
| itsnotabigtruck | then i guess someone else changed the design | 02:44 |
| djszapi | tehdely: the dpkg real code is still gpl | 02:44 |
| tehdely | oh, this is just your wrapper | 02:44 |
| djszapi | tehdely: the wrapper itself is not | 02:44 |
| tehdely | gotcha | 02:44 |
| djszapi | itsnotabigtruck: yes, sure ... | 02:44 |
| djszapi | be smart man, continue :) | 02:44 |
| admiral0 | dpkg wrapper? | 02:44 |
| admiral0 | dafuq? | 02:44 |
| djszapi | anyway, I will fix this hole tomorrow | 02:45 |
| djszapi | and send the patch for review. | 02:45 |
| tehdely | interesting | 02:46 |
| tehdely | aegis rejects packages installed from opensh | 02:46 |
| admiral0 | djszapi: you know it's gonna make users angrier | 02:46 |
| djszapi | admiral0: opposite | 02:46 |
| djszapi | it is gonna make users happy | 02:46 |
| admiral0 | eventually i'm gonna ask my friends in bin-reverse to take a look at aegis | 02:46 |
| admiral0 | our security team at university is pretty good :) | 02:47 |
| djszapi | admiral0: they are not better than itsnotabigtruck :) | 02:47 |
| admiral0 | we didn't even start looking | 02:47 |
| admiral0 | we will | 02:48 |
| djszapi | he wanna know everything better, no matter if even better than the authors, or official sites :) | 02:48 |
| admiral0 | :D | 02:48 |
| djszapi | admiral0: which university is this ? | 02:48 |
| djszapi | if I may ask. | 02:48 |
| admiral0 | why are you interested? | 02:49 |
| admiral0 | legal things | 02:49 |
| admiral0 | ? | 02:49 |
| djszapi | because I know a good university which made quite many analyzes for google. | 02:49 |
| admiral0 | ah | 02:49 |
| *** stlpaul has quit IRC | 02:49 | |
| djszapi | about security, and they won a lot of money out of it. | 02:49 |
| admiral0 | we just do stuff for fun | 02:49 |
| admiral0 | ictf, defcon, ructfe | 02:50 |
| admiral0 | nothing to do with google | 02:50 |
| pa | sorry to bug you guys, again, no one ever used syncevolution? i managed to configure it, with two peers (two gmail_contacts account). i set the first as refresh from server, the second as one way from server | 02:51 |
| pa | the problem is that when i sync the second, only the contacts with ID> maxid from the first account are actually added (i think) | 02:51 |
| pa | now actually it didnt sync at all in the second peer | 02:53 |
| *** adlan has quit IRC | 02:57 | |
| *** DocScrutinizer has joined #harmattan | 03:06 | |
| *** M4rtinK has quit IRC | 03:12 | |
| tehdely | does anyone know the meaning of this: Mar 4 17:13:11 (2012) cellular: csd[655]: com.nokia.csd.simatk: simatk_report_event_notification:460 Building and sending ATK_REPORT_EVENT_NTF -message | 03:13 |
| tehdely | it shows up in syslog quite a bit | 03:13 |
| djszapi | probably no harm. | 03:14 |
| tehdely | on a lark, i'm gonna try installing ad-hac via the AppsForMeego client | 03:14 |
| tehdely | will be neat if it works | 03:14 |
| tehdely | worked! | 03:15 |
| tehdely | i gotta say | 03:16 |
| tehdely | i know some things are frustrating with this phone | 03:16 |
| tehdely | but as a long-time Nokia fanboy | 03:17 |
| tehdely | this is such a breath of fresh air | 03:17 |
| tehdely | that nothing can make me mad :> | 03:17 |
| djszapi | yay | 03:17 |
| djszapi | glad to hear that :) | 03:18 |
| tehdely | i wish some of the stuff that only made it into the Symbian maps suite were available but otherwise it's all good | 03:18 |
| tehdely | there's no Check In, separate Transport app, or ability to calculate time | 03:19 |
| tehdely | ~*~ alas ~*~ | 03:19 |
| *** deimos_ has quit IRC | 03:25 | |
| *** adlan has joined #harmattan | 03:27 | |
| *** arcean has quit IRC | 03:34 | |
| admiral0 | i assigned a pass to user | 03:39 |
| admiral0 | fuu | 03:39 |
| beford | reboot and tell us what happens | 03:43 |
| beford | xD | 03:43 |
| admiral0 | all ok | 03:46 |
| *** faenil has quit IRC | 03:50 | |
| *** admiral0 has quit IRC | 03:57 | |
| *** longthen has quit IRC | 04:00 | |
| *** jluisn has joined #harmattan | 04:02 | |
| *** jluisn has quit IRC | 04:12 | |
| *** risca has joined #harmattan | 04:12 | |
| Richlv | iptables v1.4.8: can't initialize iptables table `filter': Permission denied (you must be root) | 04:28 |
| Richlv | i am root :/ | 04:28 |
| virtuald | AEGIS_FIXED_ORIGIN override only allowed in OPEN MODE device | 04:47 |
| virtuald | hm | 04:47 |
| virtuald | richlv: develsh -c 'iptables -nL' | 04:48 |
| Richlv | virtuald, oh, interesting - thanks | 04:52 |
| Richlv | thought devel-su would be enough :) | 04:52 |
| *** nid0 has quit IRC | 05:09 | |
| *** ieatlint has quit IRC | 05:20 | |
| *** ieatlint has joined #harmattan | 05:20 | |
| *** oberling_ has joined #harmattan | 05:31 | |
| *** oberling has quit IRC | 05:33 | |
| virtuald | richlv: no the device is all locked down, it's an abomination to free software | 05:34 |
| itsnotabigtruck | virtuald: accli -I | 05:42 |
| itsnotabigtruck | are you in open mode? | 05:42 |
| itsnotabigtruck | if not, that might be why :p | 05:42 |
| itsnotabigtruck | and Richlv: get used to using 'devel-su -s /bin/develsh' | 05:42 |
| virtuald | =P yeah i read the backlog after i tried that | 05:44 |
| virtuald | what stops working in open mode? sorry for asking all the faq's | 05:45 |
| itsnotabigtruck | virtuald: very little but you have to completely wipe the phone while you go into open mode, or lots of things will stop working | 05:46 |
| Richlv | itsnotabigtruck, heh, changing shell was on my mind already, thanks :) | 05:48 |
| virtuald | ok | 05:48 |
| slingr | do i need to backup before upgrading to PR1.2? | 06:27 |
| tehdely | opensh, you complete me | 06:40 |
| *** Necrosporus has quit IRC | 06:42 | |
| *** Necrosporus has joined #harmattan | 06:44 | |
| djszapi | slingr: it does not hurt to make a backup. | 06:58 |
| djszapi | virtuald: not just open mode...about the AEGIS_FIXED_ORIGIN post | 07:00 |
| djszapi | it also works with R&D certificate. | 07:00 |
| djszapi | virtuald: and if you have an N9 with warranty, you theoritically lost the warranty with open mode (just a friendly reminder). | 07:00 |
| *** crevetor has quit IRC | 07:08 | |
| virtuald | haven't.put it in open mode yet and i may try to add an antenna to the fm transmitter amyway | 07:15 |
| tehdely | virtuald: i would _love_ to know if you pull that off | 07:23 |
| *** jaywink has joined #harmattan | 07:23 | |
| virtuald | tehdely: i don't even known if it's possible yet | 07:25 |
| jonni | tehdely: you should catch the right timing (most likely) with: sudo flasher -R;sleep 1.5;sudo flasher -l -b -k | 07:26 |
| tehdely | jonni: i pulled it off earlier | 07:26 |
| jonni | tehdely: roger | 07:27 |
| tehdely | yanked the usb cable and held the power button till it powered off | 07:27 |
| tehdely | and then had the flasher ready w/ kernel for when i plugged it abck in | 07:27 |
| virtuald | =) | 07:30 |
| *** heymaster has joined #harmattan | 07:34 | |
| *** mece has joined #harmattan | 07:56 | |
| *** beford has quit IRC | 07:59 | |
| mece | ZogG_laptop, *tadaa* | 08:00 |
| *** Stecchino_ has quit IRC | 08:00 | |
| *** Stecchino has joined #harmattan | 08:00 | |
| *** dm8tbr has quit IRC | 08:08 | |
| djszapi | Is it possible to reside a Column in a page right to the top of ToolBar with anchors ? I cannot seem to be able to manage that. | 08:08 |
| *** benares_98 has quit IRC | 08:09 | |
| *** Venemo_N950 has joined #harmattan | 08:10 | |
| djszapi | virtuald: I have not said you put it into open mode ... not sure about "fm transmitter anyway" post. | 08:13 |
| Venemo_N950 | fm transmitter? | 08:14 |
| Venemo_N950 | oh, good morning djszapi :) | 08:15 |
| djszapi | Venemo_N950: morn | 08:30 |
| *** xarcass has joined #harmattan | 08:31 | |
| gabriel9 | morning | 08:49 |
| gabriel9 | did anyone read this: http://www.omgubuntu.co.uk/2012/02/demand-for-kde-tablet-phenomenal/ | 08:49 |
| *** risca has quit IRC | 08:49 | |
| gabriel9 | i did not belive it will be this good | 08:49 |
| ZogG_laptop | mece: sup | 08:54 |
| mece | I'm on irc :) | 08:55 |
| ZogG_laptop | yeah i see it | 08:55 |
| ZogG_laptop | i'll query you | 08:55 |
| *** rigo has joined #harmattan | 09:17 | |
| *** heymaste_ has joined #harmattan | 09:18 | |
| *** gabriel9 has quit IRC | 09:19 | |
| *** heymaster has quit IRC | 09:20 | |
| *** longthen has joined #harmattan | 09:40 | |
| *** spenap has joined #harmattan | 09:47 | |
| *** Venemo has joined #harmattan | 09:51 | |
| *** Venemo has joined #harmattan | 09:51 | |
| *** gabriel9 has joined #harmattan | 09:51 | |
| *** gabriel9 is now known as gabriel9|work | 09:53 | |
| *** psycho_oreos has joined #harmattan | 09:57 | |
| *** djszapiN9 has joined #harmattan | 09:58 | |
| djszapiN9 | Venemo_N950, ping | 09:59 |
| djszapiN9 | why is it immediately red if i query someone ? | 09:59 |
| gabriel9|work | what is red? | 09:59 |
| djszapiN9 | gabriel9|work, my pants :) | 10:00 |
| gabriel9|work | dude, why red pants :D | 10:00 |
| djszapiN9 | the title of the query dialog (aka the other buddy) | 10:00 |
| *** leinir has quit IRC | 10:00 | |
| *** Saviq_ has joined #harmattan | 10:04 | |
| *** Saviq_ has quit IRC | 10:04 | |
| djszapiN9 | right, so aegis fixed origin only works in open mode, and the rescue hackery put the device into that | 10:09 |
| djszapiN9 | so it seems to work as i expected, luckily then. The context was vague to me. | 10:10 |
| djszapiN9 | tehdely, ^ | 10:10 |
| *** Arkenoi has joined #harmattan | 10:12 | |
| *** Natunen has joined #harmattan | 10:12 | |
| *** ZogG_laptop has quit IRC | 10:14 | |
| Venemo | djszapiN9, is it? | 10:15 |
| *** tbf has joined #harmattan | 10:15 | |
| djszapiN9 | Venemo, indeed | 10:15 |
| djszapiN9 | Venemo, plus it would be nice if i can turn the notification off. | 10:16 |
| *** Venemo_N950 has quit IRC | 10:16 | |
| djszapiN9 | plus the notifications could contain more information | 10:16 |
| Venemo | you can turn notifications off in settings | 10:17 |
| Venemo | what more info would you like? | 10:17 |
| Venemo | please type your suggestions into #irc-chatter, I must leave now but I'll read them later | 10:17 |
| djszapiN9 | which channel/query obviously | 10:17 |
| *** Natunen has quit IRC | 10:17 | |
| *** leinir has joined #harmattan | 10:24 | |
| *** leinir has joined #harmattan | 10:24 | |
| rigo | can anybody tell me how to uninstall the facebook thing on the N9? The phone doesn't let me, which is weird | 10:25 |
| djszapiN9 | rigo, what does it say in cli ? | 10:27 |
| djszapiN9 | Venemo, also, whois would be nice to have as a supported command | 10:28 |
| rigo | Can't delete | 10:28 |
| Venemo | djszapiN9, okay :) | 10:28 |
| rigo | Can't uninstall Facebook | 10:28 |
| Venemo | rigo, just don't use it. | 10:28 |
| djszapiN9 | paste the full cli output pls. | 10:29 |
| *** leinir has quit IRC | 10:29 | |
| djszapiN9 | Venemo, well it is still weird you cannot nuke it. | 10:29 |
| rigo | djszapiN9, this is in the application manager and the application manager doesn't show more output. What should I try to get more output? | 10:29 |
| *** leinir has joined #harmattan | 10:30 | |
| *** leinir has joined #harmattan | 10:30 | |
| djszapiN9 | rigo, cli | 10:30 |
| djszapiN9 | in developer mode | 10:30 |
| djszapiN9 | unless someone knows the exact reason by off hand, we need to get more information :) | 10:32 |
| rigo | no cli here, apt-cache search "cli" gives man things but nothing immediately tangible | 10:34 |
| Venemo | cli = command line interface | 10:35 |
| *** nebulon has joined #harmattan | 10:37 | |
| spenap | just wondering: if you don't mind having it installed, maybe you can simply remove the .desktop file to have it hidden from the application grid? | 10:39 |
| spenap | rigo: ^ | 10:39 |
| spenap | (not sure if aegis would mind) | 10:39 |
| jonni | you cannot uninstall since its depecy in mp-harmattan-001-pr, you would uninstall whole harmattan on it. But yes you can remove desktop file. | 10:39 |
| Venemo | jonni +1 | 10:40 |
| spenap | but I'm personally doing is storing all the apps I don't use in a folder | 10:40 |
| spenap | works quite well for hidding them from me | 10:40 |
| rigo | dam' it, should have asked the acronymfinder. Ok, did dpkg -r | 10:40 |
| rigo | rather apt-get remove facebook | 10:41 |
| denism | dpkg -r --force-depends | 10:41 |
| rigo | and now it wants to remove facebook mp-harmattan-001-pr which makes the entire system obsolete | 10:41 |
| rigo | so if I remove facebook, my phone is bricked | 10:42 |
| rigo | nice | 10:42 |
| jonni | dont remove, just hide | 10:42 |
| rigo | jonni, I know, this is not the goal. I don't want to fix a bug here | 10:42 |
| rigo | I want to understand, as Brussels might be interested in such behavior. This is really weird | 10:43 |
| rigo | so I think I understood now that the facebook application is hardwired into the basic binaries when the phone is flashed | 10:44 |
| jonni | its just a core app that you cannot remove, nothing weird in it. | 10:44 |
| *** jreznik has joined #harmattan | 10:44 | |
| rigo | I know, but see, MS media player and IE also were just core apps that you cannot remove and yet we now have more choice | 10:45 |
| rigo | but just consider my issue solved. :) | 10:45 |
| jonni | you can always install 3rd party facebook client if you dont like the native one | 10:46 |
| rigo | my point is another one. You don't have a core client for google+ or status.net that you can't remove | 10:46 |
| ieatlint | that argument is the equiv of proving a negative | 10:47 |
| rigo | exactly | 10:47 |
| phako | because google+ didn't exist when the features for that phone were planned? | 10:47 |
| rigo | ieatlint, and by turning it upside down, you will see a positive discrimination in favor of the market leader | 10:47 |
| ieatlint | incorrect | 10:48 |
| ieatlint | you lack sufficient knowledge to draw any conclusions at all | 10:48 |
| rigo | ah, ok | 10:48 |
| ieatlint | for all you know, facebook paid nokia for it, and it's a working business relationship | 10:48 |
| ieatlint | which is a standard business model, and no reason to believe it was exclusive | 10:49 |
| rigo | ok | 10:49 |
| ieatlint | simply put, you don't know, and nokia does not prevent 3rd parties from adding things like google+ support | 10:49 |
| rigo | they don't need to | 10:50 |
| flux | isn't mp-harmattan-001-pr a virtual package? so basically removing it only results in potential future upgrades not working. | 10:52 |
| flux | perhaps in that case you can re-install it before upgrading. | 10:52 |
| itsnotabigtruck | flux: except that everything in harmattan is marked auto | 10:53 |
| rigo | flux, thanks, but I just wanted to understand the bundling | 10:53 |
| itsnotabigtruck | removing mp-harmattan-x-pr makes everything in the system autoremovable | 10:53 |
| flux | itsnotabigtruck, and there's a system that automatically removes non-dependend auto-packaes? | 10:53 |
| itsnotabigtruck | and autoremoved it will be | 10:53 |
| itsnotabigtruck | completely hosing your system | 10:53 |
| flux | well, apt-get install the rest of the packages first then :) | 10:53 |
| itsnotabigtruck | yeah, the package management backend (package-manager) autoremoves after every operation | 10:53 |
| itsnotabigtruck | also the apt-mark utility for managing the auto flag doesn't work without installing a bunch of stuff from the sdk repo | 10:54 |
| itsnotabigtruck | you can safely get rid of it if you apt-mark everything as manual | 10:54 |
| itsnotabigtruck | and don't ever want to SSU in the future | 10:54 |
| flux | nice, I didn't know about apt-mark | 10:54 |
| flux | I was thinking though that when you apt-get install an installed auto-package, it just marks it non-auto | 10:55 |
| itsnotabigtruck | i think that works too | 10:56 |
| *** aheinecke has joined #harmattan | 10:56 | |
| djszapiN9 | it is quite weird you cannot remove facebook | 10:58 |
| djszapiN9 | so that it is also hard wired... | 10:58 |
| djszapiN9 | rigo, did that work what denism wrote ? | 11:00 |
| rigo | djszapiN9, it would have wrecked my phone, so I did not do the --force-depends | 11:01 |
| djszapiN9 | damn facebook :) | 11:02 |
| rigo | I just did apt-get remove facebook and apt offered to remove facebook mp-harmattan-001-pr and autoremove everything else | 11:02 |
| rigo | at that point, I selected option "n" | 11:02 |
| *** snowpong has joined #harmattan | 11:02 | |
| djszapiN9 | yes, of course, if it is also hard wired | 11:02 |
| rigo | just wanted to understand, now I understood. Thanks also to ieatlint for explaining me the business aspect of it | 11:03 |
| denism | rigo: "dpkg -r --force-depends facebook" will just remove the facebook package, that's all | 11:03 |
| rigo | with potential consequences for the SSU | 11:04 |
| denism | yes :) | 11:04 |
| rigo | not keen on doing that | 11:04 |
| *** aquarius has joined #harmattan | 11:05 | |
| itsnotabigtruck | denism: that will also hose up apt | 11:08 |
| itsnotabigtruck | don't do that | 11:08 |
| itsnotabigtruck | go with the apt-mark approach, you could also tamper with /var/lib/dpkg/status | 11:09 |
| itsnotabigtruck | and change the dependencies of mp-harmattan-001-pr | 11:09 |
| djszapiN9 | rigo i do not personally advice the desktop file removal | 11:09 |
| djszapiN9 | the less hassle, if you just ignore it as me. :) | 11:11 |
| djszapiN9 | advise* | 11:11 |
| flux | maybe you can chmod a-r the desktop file | 11:14 |
| flux | not sure what will happen though :) | 11:14 |
| rigo | practically, that's what I'll do. | 11:14 |
| flux | reminded by this, I'll create a folder 'Crap' and put all the, well, crap inside it | 11:15 |
| djszapiN9 | rigo life is full of compromises :) | 11:15 |
| djszapiN9 | ...and pants :) | 11:15 |
| mece | itsnotabigtruck, how do you mess with the meta package dependencies? | 11:16 |
| djszapiN9 | flux that folder is gonna run out of space in no time :p | 11:16 |
| flux | djszapin9, they don't scroll? | 11:16 |
| djszapiN9 | just joking :) | 11:17 |
| qronic | rigo, just add NotShowIn=X-MeeGo; into facebook .desktop file | 11:17 |
| rigo | hey!! | 11:17 |
| rigo | there is always a workaround in IT :) | 11:18 |
| *** leinir has quit IRC | 11:18 | |
| *** leinir_ has joined #harmattan | 11:18 | |
| *** leinir_ has joined #harmattan | 11:18 | |
| djszapiN9 | point is, why do you need to modify a core package content for no real reasons. | 11:19 |
| djszapiN9 | to not see one icon out of the about 40. | 11:20 |
| rigo | there is a reason, and this reason is created by business and I have understood that now :) | 11:20 |
| rigo | and a very onerous reason, I assume | 11:20 |
| djszapiN9 | it is created by business reason to hide one icon ? :) | 11:21 |
| rigo | you can't imagine how much noise it created when Jobs refused to have flash on his devices ... So there are entire industries depending on such deals... | 11:21 |
| *** leinir_ is now known as leinir | 11:23 | |
| auenf | time to hide djszapiN9 from the world for a business reason | 11:26 |
| Venemo | heh? | 11:27 |
| itsnotabigtruck | djszapiN9: sounds like a corporate policy to baninate social networking anything? | 11:27 |
| itsnotabigtruck | also o.O | 11:27 |
| itsnotabigtruck | mece: well, you go edit the file and change them to whatever you want | 11:29 |
| itsnotabigtruck | probably best to kick up the version a notch too | 11:29 |
| itsnotabigtruck | like add a +hacked1 or something to the end | 11:29 |
| djszapi | itsnotabigtruck: no, that is not what I meant. | 11:32 |
| djszapi | I asked why bother with altering a core package content for no real gains. | 11:32 |
| *** djszapiN9 has quit IRC | 11:32 | |
| itsnotabigtruck | djszapi: i meant more as in rigo's objective | 11:35 |
| itsnotabigtruck | it sounds like the goal is to actually eliminate facebook from the phone | 11:35 |
| djszapi | I could not mention any real gains in comparison with the mess altering a core package content. | 11:36 |
| jonni | as its integrated into too many places, most likely it cannot be removed, but you can block it from working quite easily, etc/hosts, iptables, etc | 11:37 |
| jonni | (if your corporate policy forbids facebook) | 11:38 |
| djszapi | haha I would be surprised, if it is for real you cannot use the phone you bought :D | 11:38 |
| rigo | I will never ever again raise such an issue here :) | 11:38 |
| djszapi | at your company. | 11:39 |
| *** aquarius has quit IRC | 11:39 | |
| djszapi | and even more, if they actually check technically you nuke "facebook" out. | 11:40 |
| *** aquarius has joined #harmattan | 11:40 | |
| itsnotabigtruck | rigo: tweaking the system metapackage and expunging anything that looks like facebook seems like the most elegant solution | 11:40 |
| itsnotabigtruck | that should at least remove things like the account plugin and facebook app | 11:41 |
| itsnotabigtruck | you could of course also easily break your system doing tricks like that | 11:41 |
| djszapi | rigo: have you company asked for nuking facebook on your phone ? | 11:41 |
| djszapi | your* | 11:41 |
| rigo | 1/ my phone is a company phone | 11:41 |
| rigo | 2/ my company is veeeery tolerant ;) | 11:42 |
| djszapi | both points confirm that, you do not really need to hack the phone out. | 11:42 |
| djszapi | for no real gains. | 11:42 |
| *** arcean has joined #harmattan | 11:42 | |
| rigo | djszapi, that wasn't my intent. I wanted to understand the bundling for non-technical solutions | 11:43 |
| *** oberling_ is now known as oberling | 11:43 | |
| rigo | s/solutions/conclusions/ | 11:43 |
| infobot | rigo meant: djszapi, that wasn't my intent. I wanted to understand the bundling for non-technical conclusions | 11:43 |
| rigo | good bot | 11:43 |
| djszapi | indeed, and that is why I said, it is just better to ignore one icon, no ? :) | 11:44 |
| *** merlin1991 has quit IRC | 11:44 | |
| rigo | and that's what I'll do | 11:45 |
| djszapi | nice | 11:45 |
| *** dm8tbr has joined #harmattan | 11:45 | |
| *** merlin1991 has joined #harmattan | 11:45 | |
| *** M4rtinK has joined #harmattan | 11:45 | |
| rigo | djszapi, but see, that actually gives me hope | 11:46 |
| rigo | the guys @facebook are everything but stupid | 11:46 |
| rigo | so if they bet money on having their app integrated into the N9, that means the have given this phone a potential | 11:47 |
| djszapi | tehdely: do you have a qt5 build on Mac ? | 11:47 |
| jonni | well actually facebook.provider is not in refhash list, so if you remove that file, then there is noway to make facebook accout, and your device would be 'without facebook' | 11:48 |
| jonni | if someone would want to make device without facebook.... | 11:48 |
| djszapi | jonni except that if it blows unexpectedly in the future | 11:48 |
| djszapi | they add it to the refhashlist etc | 11:48 |
| djszapi | it is really not worth it ... | 11:48 |
| jonni | but nobody wants to do that anyways :) | 11:48 |
| djszapi | or even a simple regression | 11:49 |
| jonni | yep that profider can appear back on next update. | 11:49 |
| djszapi | nah, really... | 11:49 |
| djszapi | you do not wanna risk this by any means. | 11:49 |
| djszapi | there are already enough troubles, to have yet another potential one. | 11:49 |
| rigo | it's really impossible to stop, seems too intriguing :)) | 11:50 |
| djszapi | jonni: actually, you gave me an idea for a report :) | 11:52 |
| djszapi | since it is a problem imo... | 11:52 |
| djszapi | also, I am surprised about the desktop file modifications too | 11:53 |
| djszapi | since if it is this benevolent, you can replace the exe line with anything.... | 11:53 |
| djszapi | which is an attack surface... | 11:54 |
| *** DocScrutinizer has quit IRC | 11:59 | |
| *** DocScrutinizer has joined #harmattan | 12:01 | |
| *** dm8tbr has quit IRC | 12:02 | |
| *** longthen has quit IRC | 12:03 | |
| *** longthen has joined #harmattan | 12:04 | |
| *** vihta has left #harmattan | 12:05 | |
| *** Venemo has quit IRC | 12:10 | |
| radiofree | djszapi: pong | 12:14 |
| djszapi | radiofree: sorry, wanted to find someone with native English for checking my email back then, but already figured out. Thanks. :) | 12:14 |
| *** dfaure has joined #harmattan | 12:14 | |
| radiofree | heh, ok ;) | 12:15 |
| radiofree | my irc bouncer isn't too great, I didn't see the PM until I got in today | 12:15 |
| djszapi | does http://codereview.qt-project.org/ load for others ? | 12:15 |
| djszapi | ok, it now loads. | 12:16 |
| dfaure | itsnotabigtruck: love you, man ;) | 12:17 |
| dfaure | itsnotabigtruck: my N9 boots again, and I didn't lose my calendar/contacts | 12:17 |
| dfaure | just had to flash the .bin you gave me the url for | 12:17 |
| *** nid0 has joined #harmattan | 12:18 | |
| *** lizardo has joined #harmattan | 12:29 | |
| *** longthen1 has joined #harmattan | 12:33 | |
| *** longthen has quit IRC | 12:34 | |
| *** djszapi has left #harmattan | 12:35 | |
| *** jaywink has quit IRC | 12:35 | |
| *** djszapi has joined #harmattan | 12:50 | |
| djszapi | any arch users here ? | 12:50 |
| gabriel9|work | here | 12:55 |
| gabriel9|work | i use one and only arch :) | 12:55 |
| djszapi | gabriel9|work: http://paste.kde.org/433568/ | 12:56 |
| djszapi | and http://paste.kde.org/433586/ | 12:56 |
| gabriel9|work | you have problems with keys? | 12:57 |
| gabriel9|work | importing them? | 12:57 |
| djszapi | gabriel9|work: not sure. | 12:57 |
| gabriel9|work | did you update your pacman.conf? | 12:57 |
| gabriel9|work | you need to use new one | 12:57 |
| gabriel9|work | i think new one is named pacman.conf.new | 12:58 |
| gabriel9|work | or something like that | 12:58 |
| djszapi | gabriel9|work: I fixed it: http://paste.kde.org/433592/ | 12:58 |
| gabriel9|work | ok | 12:58 |
| gabriel9|work | i'm not at home and ar work i have win | 12:59 |
| gabriel9|work | so i can't check that commands | 12:59 |
| gabriel9|work | but if it works great :) | 12:59 |
| djszapi | the thingie is that... | 13:00 |
| djszapi | the arch keyserver (default) is unrelaiable. | 13:00 |
| djszapi | unreliable* | 13:00 |
| djszapi | and the second got also messy | 13:00 |
| djszapi | and I now need to use a third one .. | 13:00 |
| *** jluisn has joined #harmattan | 13:00 | |
| *** djszapiN9 has joined #harmattan | 13:08 | |
| *** leinir has quit IRC | 13:11 | |
| *** leinir has joined #harmattan | 13:15 | |
| *** djszapiN9 has quit IRC | 13:18 | |
| *** faenil has joined #harmattan | 13:19 | |
| gabriel9|work | for me that new system with keys is too much | 13:24 |
| gabriel9|work | things worked good before | 13:24 |
| djszapi | +1 | 13:24 |
| djszapi | if you make things complicated, it is a general rule to make it automated for users to switch without hassles. | 13:25 |
| ab | itsnotabigtruck, regarding bug 1062, I've got promise from the head of imaging team it will be fixed. | 13:38 |
| _MeeGoBot_ | Bug https://harmattan-bugs.nokia.com/show_bug.cgi?id=1062 nor, ---, ---, marcell.lengyel, ASSI, libfcam-dev (camera control API) not available in Platform SDK | 13:39 |
| ab | by way of the bug update you can also see there will be SDK update very soon. | 13:41 |
| *** Necrosporus has quit IRC | 13:48 | |
| *** Necrosporus has joined #harmattan | 13:50 | |
| *** psycho_oreos has quit IRC | 13:51 | |
| *** M4rtinK has quit IRC | 13:58 | |
| *** faenil has quit IRC | 14:13 | |
| *** tsenyk has joined #harmattan | 14:19 | |
| tsenyk | hmpf ... i kind of bricked my device :) ... same as 2 others in my office :) | 14:25 |
| tsenyk | we were brave/dumb enought to try PR1.3 .. it seams that the lock-screen and the SIM-stack is broken since 40.2012.9-x | 14:26 |
| tsenyk | n9 retail version | 14:26 |
| djszapi | tsenyk: you might wanna use internal irc channel for PR1.3 | 14:26 |
| tsenyk | hm ok, your right :) | 14:26 |
| djszapi | since that is not released yet. | 14:26 |
| tsenyk | sorry :) | 14:26 |
| djszapi | I can help with the same nickname there. | 14:26 |
| phako | tsenyk: ping me(jgeorg) on PMO | 14:28 |
| *** deimos has joined #harmattan | 14:31 | |
| jreznik | wouldn't be easier open source the whole harmattan and than use this channel? :))) | 14:52 |
| djszapi | jreznik: open source does not mean there are no confidential information in a company, itself. | 14:53 |
| jreznik | djszapi: yeah, I understand and I hate it - everything should be open in open company :) | 14:54 |
| djszapi | agree | 14:54 |
| djszapi | that is why I disliked MeeGo, too. | 14:55 |
| djszapi | or Tizen. | 14:55 |
| djszapi | on an interesting note, it seems Tizen dropped WAC2.0 | 14:55 |
| djszapi | they might go for some jquery mobile or what ? :) | 14:56 |
| jreznik | lol | 14:59 |
| jreznik | the intel guys are crazy, every half yer a new technology... | 15:00 |
| djszapi | I would attend to the Tizen conference for fun. | 15:00 |
| jreznik | would be great to have billions dollars and just spend them depending on your mood :) | 15:00 |
| djszapi | but I would not be able to get sponsorship. | 15:00 |
| djszapi | hehehe | 15:00 |
| jreznik | djszapi: that could be funny :) | 15:00 |
| *** faenil has joined #harmattan | 15:06 | |
| *** etrunko has joined #harmattan | 15:12 | |
| *** tbf_ has joined #harmattan | 15:15 | |
| *** beford has joined #harmattan | 15:16 | |
| *** tbf has quit IRC | 15:19 | |
| *** dm8tbr has joined #harmattan | 15:22 | |
| djszapi | http://harmattan-dev.nokia.com/docs/library/html/guide/html/Developer_Library_Developing_for_Harmattan_Developer_tools_Performance_testing_tools_Using_PowerTOP.html -> Does powertop work for anybody | 15:25 |
| djszapi | I tried to install the power analyzes that way, but I am not getting "powertop" in the command line. | 15:26 |
| djszapi | "devtools-power" and "energy-profiler-plugin-power" packages I get on ly. | 15:26 |
| *** hardaker has joined #harmattan | 15:31 | |
| djszapi | seems like a buggy documentation | 15:44 |
| * djszapi is opening a bug report... | 15:44 | |
| *** xarcass has quit IRC | 15:46 | |
| *** tbf has joined #harmattan | 15:48 | |
| *** tbf_ has quit IRC | 15:48 | |
| djszapi | https://harmattan-bugs.nokia.com/show_bug.cgi?id=1119 -> feel free to vote ;-) | 15:52 |
| _MeeGoBot_ | Bug 1119 maj, ---, ---, ext-risto.lahti, NEW, PowerTOP is not available for developers despite of the documentation | 15:52 |
| itsnotabigtruck | ab: awesome :) and yeah, the promise of an SDK update of some kind incoming is probably the most important thing | 15:54 |
| *** danielcbit has joined #harmattan | 15:54 | |
| *** mece has quit IRC | 15:55 | |
| danielcbit | Hi Everybody. Does some one know how to erase user files with postrm deb script? | 15:55 |
| itsnotabigtruck | danielcbit: what kind of user files? if they're data files it's not normally something you're supposed to do | 15:56 |
| itsnotabigtruck | anyway, you would write a case statement and when removing, you'd nuke the files with rm, simple as that | 15:56 |
| danielcbit | files created in .config and .local by my app. | 15:57 |
| merlin1991 | anbody got a working ntp client on the n9? I don't want to install the full ntpd | 15:57 |
| danielcbit | itsnotabigtruck: I did this but I always get Permission denied | 15:57 |
| merlin1991 | just make an udpate once in a while | 15:57 |
| djszapi | danielcbit: I can show an example | 15:58 |
| djszapi | if you are interested in that. | 15:58 |
| itsnotabigtruck | oh yeah, the files are owned by user | 15:58 |
| itsnotabigtruck | your postrm script runs as root | 15:58 |
| djszapi | it is very simple. | 15:58 |
| itsnotabigtruck | try asserting <request context="INSTALL"><credential name="UID::user" /></request> | 15:58 |
| itsnotabigtruck | that makes your postrm script run as user, i think | 15:59 |
| djszapi | danielcbit: for my cups package: http://paste.kde.org/433772/ | 15:59 |
| danielcbit | itsnotabigtruck: Ok, I'll try it. | 15:59 |
| djszapi | danielcbit: http://paste.kde.org/433778/ -> and this is the aegis file for that | 15:59 |
| djszapi | danielcbit: you can remove the first two sections, and customize the last according to your needs (user files: user access). | 16:00 |
| danielcbit | djszapi: but i still need the capability described by itsnotabigtruck. Right? | 16:00 |
| djszapi | and I gave you a full fledged example :) | 16:00 |
| djszapi | so it is only copy/paste :) | 16:00 |
| djszapi | Also, note that: you cannot fine tune the request only for the postrm script. | 16:01 |
| danielcbit | Thanks djszapi! This should be in the nokia Docs, although it is quite trivial. | 16:01 |
| djszapi | I wanted to implement that granularity back then, but did not find the time. | 16:01 |
| djszapi | danielcbit: it is in the Nokia docs ;-) | 16:01 |
| danielcbit | djszapi: I didn't managed to find. And I looked really hard. | 16:02 |
| itsnotabigtruck | djszapi: hm | 16:03 |
| djszapi | danielcbit: "The context attribute" here: http://harmattan-dev.nokia.com/docs/library/html/guide/html/Developer_Library_Developing_for_Harmattan_Harmattan_security_Security_guide_Aegis_manifest_syntax.html | 16:03 |
| itsnotabigtruck | so preinst and postrm always run as UID::root and no caps? | 16:03 |
| djszapi | that is false. | 16:03 |
| djszapi | what I referred to is that you cannot have different cred set for those separately. | 16:04 |
| djszapi | as in, having A for 1, B for 2, etc | 16:04 |
| itsnotabigtruck | djszapi: that doc wouldn't help if he didn't already know how to solve it | 16:04 |
| itsnotabigtruck | oh ok | 16:04 |
| djszapi | where 1,2 .. are the maintainer scripts. | 16:04 |
| itsnotabigtruck | right, i think the 4 scripts always run with the same caps | 16:04 |
| djszapi | well, that docs helped many people :) | 16:04 |
| djszapi | I do not just think, I know :) | 16:05 |
| djszapi | unfortunately, I had not the time for making it nicer. oh well. | 16:05 |
| djszapi | Also, I wonder why the SDK team did not pubish the many examples I wrote back then. | 16:06 |
| djszapi | I think they would be useful for noobs about Harmattan security :/ | 16:07 |
| danielcbit | djszapi: Ok, I see it. But it is not related to remove user files. It seem kind of obscure info. The user has to make the connections to understand and use it. | 16:07 |
| djszapi | yes of course. | 16:07 |
| djszapi | we cannot provide copy/paste examples for everybody | 16:07 |
| djszapi | but the theory is there how to request creds for the maintainer scripts. | 16:08 |
| itsnotabigtruck | keep in mind that a lot of people still don't realize that being root != having dac_override god-mode filesystem access | 16:09 |
| djszapi | the documentation is already too long to make it even longer and more useful, so a minimal thinking is expected :) | 16:09 |
| itsnotabigtruck | and on normal debian you can rm anything anywhere from your maintainer scripts | 16:09 |
| danielcbit | djszapi: A nice note where would make things much clear: http://harmattan-dev.nokia.com/docs/library/html/guide/html/Developer_Library_Alternative_development_environments_Platform_SDK_user_guide_Packaging_your_application_with_Harmattan_Platform_SDK.html | 16:09 |
| djszapi | danielcbit: what do you mean ? | 16:09 |
| *** adlan has joined #harmattan | 16:10 | |
| danielcbit | djszapi: A note about the requirements for running maintainer scripts and a link to the aegis doc for deeper info | 16:11 |
| djszapi | there are two types of users: 1) Hate reading long docs 2) Hate short docs -> we have to find a balance in that, and I do not think this special case would be any hard. If you need to remove something, you need to be able to do that, so you need to request a credential. How to request a credential ? Manifest syntax guide. | 16:11 |
| djszapi | if we enumeate every single use case word by word, without requiring even a simple logic from /developers/, the documentation would go haywire. | 16:12 |
| djszapi | It is already very long according to many. | 16:13 |
| danielcbit | djszapi: everything is easy after you solved it. Bear in mind that. But thanks for all the help. I'll try to look harder next time | 16:13 |
| danielcbit | djszapi: I'm not saying about making it longer, but making it more connected. This would make novice developers feel more supported | 16:15 |
| djszapi | I do not really understand the problem. | 16:15 |
| djszapi | I need to do something. Do I have the creds for that ? No, mmh... I need to request it. How can I request it ? Go to the manifest syntax guide. | 16:15 |
| djszapi | it sounds natural logic to me. | 16:16 |
| djszapi | I am unsure in this context what "connected" menas. | 16:16 |
| djszapi | means* | 16:16 |
| djszapi | Please note that, the security team is not responsible for making the SDK documentation "connections", if you need such a thing | 16:17 |
| djszapi | we have never seen such a packaging documentation internally. | 16:17 |
| djszapi | Feel free to file a bugreport against the SDK team, if it really makes you happy. We need people happy in the end :) | 16:21 |
| danielcbit | djszapi: Like I was saying. Everything is much easier after you solved It. But a small not in harmattan-dev.nokia.com/docs/library/html/guide/html/Developer_Library_Alternative_development_environments_Platform_SDK_user_guide_Packaging_your_application_with_Harmattan_Platform_SDK.html would certainly help lots of users. | 16:22 |
| *** faenil has quit IRC | 16:22 | |
| danielcbit | djszapi: But thanks for the help and for the example! | 16:22 |
| djszapi | your are the first one I heard so far :) | 16:23 |
| djszapi | but really, if it bugs you, create a bugreport | 16:23 |
| djszapi | I will not mind :) | 16:23 |
| djszapi | They can probably add a link to the security documentation. | 16:23 |
| danielcbit | djszapi: Ok. | 16:24 |
| itsnotabigtruck | i think the issue here is that this is the sort of situation where no one realizes they need caps | 16:25 |
| *** M4rtinK has joined #harmattan | 16:25 | |
| *** ab has quit IRC | 16:25 | |
| djszapi | itsnotabigtruck: and that is how, if operation permitted ? | 16:25 |
| itsnotabigtruck | also qt creator works hard to hide aegis from the dev, which means that if you need any deviation from that then people don't know what to do | 16:25 |
| *** Appiah has quit IRC | 16:26 | |
| djszapi | nah.. | 16:26 |
| djszapi | operation permitted is operation permitted everywhere. | 16:26 |
| djszapi | and even the SDK says that, if you need something that it does not autogenerate, go to the security guide | 16:26 |
| djszapi | "For further details". | 16:26 |
| itsnotabigtruck | doesn't mean people will fully understand the security guide and know they need more caps | 16:27 |
| itsnotabigtruck | well, creds, i should call them creds | 16:27 |
| djszapi | you can use any of those, I will understand :) | 16:29 |
| djszapi | well, if operation is not permitted, it is always a good thing to ask yourself: are you allowed to do that ? | 16:30 |
| djszapi | which is, if you are not sure, go to the security guide. | 16:30 |
| danielcbit | djszapi: Another problem is to know that caps are needed and which ones are really needed. What every caps provide is not documented. | 16:31 |
| djszapi | danielcbit: most of them are upstream linux | 16:31 |
| djszapi | /usr/include/linux/capabilities.h | 16:31 |
| dfaure | itsnotabigtruck: saw my note this morning? I want to thank you for your howto and for the download link, my N9 is working again. | 16:31 |
| djszapi | this is also documented, and also how you get the available creds that you can request. | 16:32 |
| djszapi | for the latter: accli is a good helper util. | 16:32 |
| djszapi | but there are other ways, too. | 16:32 |
| itsnotabigtruck | dfaure: glad you got your N9 going again :) | 16:33 |
| dfaure | itsnotabigtruck: and I didn't lose my calendar nor contacts. You saved me man ;) | 16:34 |
| danielcbit | djszapi:"/usr/include/linux/capabilities.h" from where? My distro? Harmattan? | 16:34 |
| djszapi | danielcbit: /usr/include/linux/capability.h on Harmattan of course :) | 16:35 |
| djszapi | there is a simple mapping from there into CAP::X stuff. | 16:35 |
| itsnotabigtruck | yeah, CAP::x is all linux capabilities, which are poorly documented but exist in all linux | 16:35 |
| djszapi | danielcbit: you know what ? I will put up a wikipage about the several policies | 16:36 |
| djszapi | to make it simpler to grab. | 16:36 |
| itsnotabigtruck | usually they're used so you can e.g. have a daemon that's allowed to listen on low ports but not do any other root stuf | 16:36 |
| danielcbit | djszapi: My N9 doesn't this file anywhere in the system. | 16:38 |
| itsnotabigtruck | danielcbit: it's a kernel header file, it would be in your scratchbox | 16:38 |
| djszapi | danielcbit: it is bundled during the build I guess. | 16:38 |
| itsnotabigtruck | but you can find info about caps on the net | 16:38 |
| * itsnotabigtruck has to run | 16:38 | |
| djszapi | danielcbit: the dpk wrapper "imports" the mapping in during the build time. | 16:38 |
| djszapi | dpkg* | 16:38 |
| danielcbit | djszapi: This wiki would certainly help lots of people. | 16:39 |
| djszapi | danielcbit: you can grab this file on the kernel lxr page | 16:39 |
| djszapi | with the relevant kernel version | 16:39 |
| djszapi | I do not think we have local alterings in comparison with upstream | 16:39 |
| djszapi | wrt that file. | 16:39 |
| itsnotabigtruck | basically UID = your process' uid, GID = your process' gid, GRP = adds more groups beyond the gid, CAP = partial root access, tcb = lets you control aegis, everything else = can do anything, most of them have to do with dbus services | 16:40 |
| itsnotabigtruck | danielcbit ^ | 16:40 |
| danielcbit | djszapi: Ok. But you see. It is an enormous amount of work for such information. The wiki would be great! | 16:40 |
| djszapi | tcb is mostly irrelevant | 16:41 |
| djszapi | and there are way more creds than that :) | 16:41 |
| djszapi | danielcbit: many people have been complaining about documentation since last summer, but nobody put up a wikipage yet even a minimal one, ehh ? :) | 16:41 |
| itsnotabigtruck | djszapi: well yeah, there's way more creds, but they aren't hardwired into the system like those are | 16:42 |
| *** faenil has joined #harmattan | 16:43 | |
| djszapi | huh ? | 16:43 |
| *** M4rtinK has quit IRC | 16:43 | |
| djszapi | what do you mean ? Could you please clarify ? | 16:43 |
| itsnotabigtruck | djszapi: i mean, the UID/GID/GRP/CAP/tcb ones have special meaning | 16:44 |
| itsnotabigtruck | the other ones only do something if an app checks for them | 16:44 |
| itsnotabigtruck | e.g. the package manager ones that package-manager or dbus checks for | 16:44 |
| djszapi | still not sure what is your point | 16:44 |
| *** Guest82265 has joined #harmattan | 16:44 | |
| djszapi | UID is needed for A | 16:44 |
| djszapi | other cap is needed for B | 16:45 |
| itsnotabigtruck | so it makes sense to lump them together as "everything else" | 16:45 |
| djszapi | Nokia specific token is needed for C | 16:45 |
| djszapi | Community specific token is needed for C | 16:45 |
| djszapi | D* | 16:45 |
| djszapi | they all have their purposes. | 16:45 |
| djszapi | mmmh, it is gonna be a hard job, I am used to the Nokia style wiki pages...not the meego style.. | 16:46 |
| itsnotabigtruck | bbl | 16:46 |
| *** briglia has joined #harmattan | 16:46 | |
| *** rm_work has joined #harmattan | 16:46 | |
| *** rm_work has joined #harmattan | 16:46 | |
| djszapi | danielcbit: http://wiki.meego.com/Aegis_Policies Not much information yet, but as a start ... | 16:51 |
| djszapi | I am soon off to the Plasma Active sprint, so I might not be able to extend it any soon :/ | 16:51 |
| djszapi | http://wiki.meego.com/N950_landing_page#Developing -> and I put a link in here for getting it easily accessible. | 16:52 |
| danielcbit | djszapi: But this is a great start! | 16:52 |
| djszapi | if I found my examples I wrote, I will publish those, too. | 16:53 |
| djszapi | would* | 16:53 |
| djszapi | but I do not have access to my Nokia email anymore. | 16:53 |
| djszapi | so cannot guarantee. | 16:53 |
| *** benares_98 has joined #harmattan | 16:53 | |
| *** lmoura has quit IRC | 16:53 | |
| djszapi | oh noes, I nuked everything aegis related while leaving... | 16:53 |
| djszapi | even what was public stuff, meh. | 16:53 |
| djszapi | danielcbit: that was it: http://paste.kde.org/433814/ Please check, if it is useful, I can put up to a public wiki page. | 16:56 |
| danielcbit | djszapi: This would certainly help, Put also an note about "/usr/include/linux/capabilities.h" | 16:59 |
| danielcbit | for other capabilities not described in this doc. | 17:00 |
| djszapi | they are, but we could link the lxr kernel source seeker engine | 17:00 |
| djszapi | for online browsing without checking it inside the build environment. | 17:01 |
| djszapi | though it would be better to link against the other page I have just created and describe it there more explicitely than the header file | 17:01 |
| djszapi | and if anybody (for some reasons, I do not know why) would need more information, he can go to the capability.h header file. | 17:01 |
| djszapi | radiofree: ping, are you satisfied with Marble ? | 17:03 |
| *** dfaure has left #harmattan | 17:03 | |
| danielcbit | djszapi: This would be great! | 17:04 |
| pa | is it possible to delete all messages (SMS) via command line? | 17:10 |
| *** benares_98 has quit IRC | 17:10 | |
| *** arcean_ has joined #harmattan | 17:10 | |
| *** risca has joined #harmattan | 17:12 | |
| *** arcean has quit IRC | 17:14 | |
| *** benares_98 has joined #harmattan | 17:15 | |
| *** Aard has quit IRC | 17:28 | |
| *** Saviq_ has joined #harmattan | 17:34 | |
| *** risca has quit IRC | 17:34 | |
| *** risca has joined #harmattan | 17:34 | |
| *** risca has quit IRC | 17:34 | |
| *** Saviq_ has quit IRC | 17:36 | |
| *** Aard has joined #harmattan | 17:41 | |
| *** etrunko has quit IRC | 17:42 | |
| *** etrunko has joined #harmattan | 17:44 | |
| djszapi | anybody experiencing that, while connected to a wifi network, it writes that "you need to connect to a wifi network for downloading" inside the maps application ? | 17:45 |
| *** mairas has joined #harmattan | 17:52 | |
| beford | new nokia publisher page has some rendering problems :/ | 18:01 |
| *** kozzi has quit IRC | 18:04 | |
| *** gabriel9|work has quit IRC | 18:05 | |
| djszapi | does anybody know how to align my Column in a page right above the Toolbar with some spacing like 5-10 px ? | 18:14 |
| djszapi | anchor.bottom: mytoolbarid.top; did not work out quite well. | 18:14 |
| spenap | i think you have to put parent.bottom, regardless of the toolbar being there | 18:18 |
| spenap | and then parent.bottomMargin | 18:18 |
| *** decibyte has quit IRC | 18:18 | |
| *** merlin1991 is now known as newnick | 18:19 | |
| djszapi | okay, let me try. | 18:19 |
| *** newnick is now known as merlin1991 | 18:19 | |
| djszapi | (I think I already tried that, but I may have messed that up) | 18:19 |
| *** decibyte has joined #harmattan | 18:20 | |
| *** risca has joined #harmattan | 18:29 | |
| djszapi | spenap: ok that really helped, thanks. | 18:30 |
| *** Guest82265 is now known as ab | 18:35 | |
| *** ab has joined #harmattan | 18:35 | |
| *** beford has quit IRC | 18:36 | |
| *** pawky|2 has quit IRC | 18:41 | |
| *** lmoura has joined #harmattan | 18:53 | |
| *** beford has joined #harmattan | 18:53 | |
| *** nebulon has quit IRC | 18:55 | |
| itsnotabigtruck | anyone on atm? | 18:56 |
| beford | no | 18:57 |
| *** NIN101 has joined #harmattan | 18:57 | |
| itsnotabigtruck | lol | 18:57 |
| njsf | I am also off | 18:58 |
| djszapi | beford: xD | 18:58 |
| *** risca has quit IRC | 18:58 | |
| djszapi | ieatlint: pants | 18:59 |
| djszapi | ieatlint: do I need to make any activities my qt5 patch was accepted, or will the approvers/maintainers manage that ? | 18:59 |
| *** blueslee has joined #harmattan | 19:00 | |
| *** risca has joined #harmattan | 19:01 | |
| *** djszapiN9 has joined #harmattan | 19:04 | |
| *** leinir has quit IRC | 19:08 | |
| *** etrunko has quit IRC | 19:11 | |
| npm | winning! "meegolem-38-~> cp -p /media/n950-user/home/user/MyDocs/DCIM/1203*.jpg /media/n9-user/home/user/MyDocs/DCIM/. | 19:20 |
| *** Venemo has joined #harmattan | 19:27 | |
| *** Venemo has joined #harmattan | 19:27 | |
| *** risca has quit IRC | 19:29 | |
| *** hardaker has quit IRC | 19:30 | |
| *** jreznik has quit IRC | 19:38 | |
| npm | wtf is a "tweet content error"... (status of a failed attempt to share a photo via twitter) | 19:39 |
| flux | btw, the pictures from my camera seem to have different kind of names than before | 19:42 |
| *** etrunko has joined #harmattan | 19:43 | |
| flux | probably changed at pr1.2.. is there a way to revert? they used to be yyyymmdd_nnn.jpg but now they are yymmnnnn.jpg | 19:43 |
| *** hag has joined #harmattan | 19:43 | |
| *** hag is now known as Guest14075 | 19:44 | |
| *** Guest14075 has quit IRC | 19:44 | |
| *** faenil has quit IRC | 19:48 | |
| *** npm_ has joined #harmattan | 19:48 | |
| *** npm has quit IRC | 19:49 | |
| *** npm_ has quit IRC | 19:50 | |
| *** npm_ has joined #harmattan | 19:53 | |
| *** leinir has joined #harmattan | 19:56 | |
| *** Free-MG has joined #harmattan | 19:56 | |
| *** jluisn has quit IRC | 20:02 | |
| *** M4rtinK has joined #harmattan | 20:04 | |
| *** jluisn has joined #harmattan | 20:05 | |
| *** hardaker has joined #harmattan | 20:07 | |
| *** piggz_ has joined #harmattan | 20:08 | |
| *** blueslee has quit IRC | 20:11 | |
| *** djszapiN9 has quit IRC | 20:16 | |
| *** Venemo has quit IRC | 20:25 | |
| *** jluisn has quit IRC | 20:26 | |
| *** npm_ is now known as npm | 20:26 | |
| *** jluisn has joined #harmattan | 20:28 | |
| *** kimju has quit IRC | 20:38 | |
| *** hardaker has quit IRC | 20:47 | |
| veli | hello, any idea how to locate the correct (as in same that qml uses) offlinestorage file from c++ side? Wonder if the md5 is stored somewhere. | 20:49 |
| *** ZogG_laptop has joined #harmattan | 20:55 | |
| *** admiral0 has joined #harmattan | 20:59 | |
| admiral0 | hello | 21:00 |
| admiral0 | ~update | 21:00 |
| infobot | update is, like, http://www.developer.nokia.com/Community/Blogs/blog/n9-developer/2012/01/25/harmattan-1.2-beta-now-available-as-ocf-for-nokia-n950 | 21:00 |
| admiral0 | DocScrutinizer: update ~update please | 21:00 |
| ZogG_laptop | admiral0: no OTA here, DocScrutinizer didn't want to flash bot =P | 21:02 |
| djszapi | any ideas how to disable the button for clicks, but enable them all back in a positioner for an event ? https://projects.kde.org/projects/kde/kdeedu/khangman/repository/revisions/master/entry/src/harmattan/GamePage.qml#L260 | 21:07 |
| *** hardaker has joined #harmattan | 21:08 | |
| ZogG_laptop | djszapi: enabled=true? | 21:13 |
| djszapi | that is not gonna solve it for all. | 21:14 |
| ZogG_laptop | than you make script | 21:14 |
| ZogG_laptop | or enabled:(something===1)?false :true | 21:15 |
| ZogG_laptop | and than you make the event to change it | 21:15 |
| ZogG_laptop | fro all buttons | 21:15 |
| djszapi | I tried to make the value conditional, depending on a value, but it did not work out for me. | 21:17 |
| djszapi | as far as I set it, once false, it remains. | 21:17 |
| ZogG_laptop | works for me | 21:17 |
| ZogG_laptop | as wall you can use states | 21:18 |
| ZogG_laptop | well* | 21:22 |
| *** lmoura_ has joined #harmattan | 21:23 | |
| *** lmoura has quit IRC | 21:23 | |
| *** lmoura_ is now known as lmoura | 21:23 | |
| djszapi | mmh, it seems it works what I did in kanagram | 21:24 |
| *** jaywink has joined #harmattan | 21:26 | |
| ZogG_laptop | kanagram? | 21:26 |
| *** kimju has joined #harmattan | 21:27 | |
| djszapi | mmmh, nope..I was wrong. | 21:30 |
| djszapi | well, no idea how to solve it. | 21:30 |
| djszapi | the most simply. | 21:30 |
| ZogG_laptop | show how you use it? | 21:30 |
| djszapi | nah, it is still not clear even in theory. | 21:30 |
| ZogG_laptop | okay it works for me anyway | 21:31 |
| djszapi | meanwhile I have achieved same in kanagram according to the model data. | 21:31 |
| djszapi | that is not gonna work here. | 21:31 |
| ZogG_laptop | using it in imgrup | 21:31 |
| djszapi | nabled:(something===1)?false :true -> it is not okay for my use case | 21:32 |
| djszapi | since my use case is not black and white | 21:32 |
| djszapi | my use case is black and colorful | 21:32 |
| djszapi | so any combination can occur during the game. | 21:32 |
| djszapi | but enabled by default. | 21:32 |
| ZogG_laptop | djszapi: yeah,that's why you change something on and off | 21:33 |
| ZogG_laptop | 1 or 0 i mean | 21:33 |
| djszapi | nope | 21:33 |
| djszapi | with the above line, everything is either enabled or disabled. | 21:33 |
| djszapi | and that is not my scenario. | 21:34 |
| ZogG_laptop | than just use in cript button1.enabled=true; | 21:34 |
| djszapi | there is no button1 | 21:34 |
| djszapi | all have the same id | 21:35 |
| ZogG_laptop | whateverbuttonname.enabled=true | 21:35 |
| djszapi | nah | 21:35 |
| djszapi | that is no separation | 21:35 |
| djszapi | for individual buttons | 21:35 |
| veli | Well if anyone else wonders how that offline storage can be found from c++: http://qt.gitorious.org/qt/qt/blobs/4.8/src/declarative/qml/qdeclarativesqldatabase.cpp#line371 | 21:36 |
| ZogG_laptop | than do it | 21:36 |
| djszapi | do what ? | 21:36 |
| djszapi | anyway, others too: https://projects.kde.org/projects/kde/kdeedu/kanagram/repository/revisions/master/entry/src/harmattan/GamePage.qml#L297 This is something that I would like to achieve. | 21:36 |
| djszapi | but the problem is that with khangman, it does not depend on the model data. | 21:36 |
| djszapi | so it is a bit more trouble.. actually I have no ideas how to do that. | 21:37 |
| ZogG_laptop | just load the name for every button in loader there are ids | 21:38 |
| djszapi | I was thinking of (status == init) ? true : true; | 21:40 |
| djszapi | so they are true for each status change. | 21:41 |
| djszapi | and if I flip out of the init state once, and then false immediately, I am good to go | 21:41 |
| djszapi | and when I turn it back to init, everything gets true. | 21:41 |
| djszapi | if the flip does not happen for status change in qml, I am not sure then ... | 21:42 |
| *** gabriel9 has joined #harmattan | 21:47 | |
| *** pipacs has joined #harmattan | 21:48 | |
| *** pipacs has quit IRC | 21:51 | |
| *** pipacs has joined #harmattan | 21:52 | |
| *** pipacs has left #harmattan | 21:56 | |
| *** pipacs has joined #harmattan | 21:57 | |
| *** rantom is now known as rantom_ | 22:02 | |
| *** rantom has joined #harmattan | 22:04 | |
| *** gabriel9 has quit IRC | 22:05 | |
| *** knee has joined #harmattan | 22:18 | |
| *** npm has quit IRC | 22:38 | |
| *** admiral0 has quit IRC | 22:38 | |
| *** npm has joined #harmattan | 22:38 | |
| *** etrunko has quit IRC | 22:53 | |
| * djszapi still waiting for ideas, if any :) | 22:54 | |
| pa | anyone knowledeable of the HW of N9? if i plug it to a usb port that is capable of up to 2A (instead of the standard 500mA), does the N9 drains 2A, or it sticks to 500mA? | 22:56 |
| *** Free-MG has quit IRC | 22:56 | |
| *** tbf has quit IRC | 22:57 | |
| djszapi | SpeedEvil: ^ | 22:58 |
| *** lizardo has quit IRC | 23:02 | |
| *** jaywink_ has joined #harmattan | 23:07 | |
| *** jluisn has quit IRC | 23:08 | |
| *** DrGrov has joined #harmattan | 23:13 | |
| *** danielcbit has quit IRC | 23:13 | |
| DrGrov | Evening | 23:13 |
| *** risca has joined #harmattan | 23:14 | |
| DrGrov | Safe to keep the N9 charging over right, 12-15 hours or so without removing charger once it has fully charged? | 23:14 |
| DrGrov | *night | 23:14 |
| djszapi | yes | 23:14 |
| DrGrov | Thanks djszapi | 23:14 |
| djszapi | even with wall charger imho | 23:14 |
| DrGrov | Yes, I am using the wall charger | 23:14 |
| djszapi | one of my device is on the charger for many weeks xD | 23:14 |
| DrGrov | :) | 23:15 |
| djszapi | I forgot to actually remove now you say... | 23:15 |
| DrGrov | I just thought to verify since I probably will forget to remove it anyhow later tonight before I get to bed | 23:15 |
| * djszapi does not go to bed tonigh | 23:15 | |
| djszapi | t | 23:15 |
| djszapi | frals ping | 23:16 |
| DrGrov | What kind of times are we talking about when it eventually becomes harmful for the phone battery to be charging constantly? | 23:16 |
| djszapi | frals you are good at Qml, do you have any ideas for the button enable/disable issues ? | 23:16 |
| *** jaywink_ has quit IRC | 23:17 | |
| itsnotabigtruck | any idea why: sudo chmod -R a=rX,u+w dir | 23:19 |
| itsnotabigtruck | seems to be skipping files in the directory tree? | 23:19 |
| itsnotabigtruck | (this isn't harmattan, but close enough) | 23:19 |
| djszapi | interesting you use capital x | 23:21 |
| djszapi | what precisely do you mean by "skipping files" ? | 23:21 |
| itsnotabigtruck | i'm under the impression it's supposed to make directories +x and files -x | 23:21 |
| *** snowpong has quit IRC | 23:22 | |
| itsnotabigtruck | so the idea is to normalize a directory tree to 0644 on files and 0755 on directories | 23:22 |
| djszapi | spenap: do you have any ideas for that very qml issue ? | 23:23 |
| *** Swerige has joined #harmattan | 23:25 | |
| *** gabriel9 has joined #harmattan | 23:27 | |
| knee | Anyone else having problems with wlan connections in pr1.2?? Connection 'seems' to be ok, but I can't download anything / open webpages.. Reconnect to the network helps.. at least for a while (problem occurs every few minutes) | 23:28 |
| knee | I tried to reflash my n9 (which didn't help), I'm now clearing the device to see if that helps.. But I'm open to any other solution (/or test..) | 23:29 |
| *** jaywink_ has joined #harmattan | 23:30 | |
| njsf | knee no, my current problem is with gmail syncing not happening on Fenix | 23:31 |
| njsf | but Wlan connections work fine | 23:31 |
| *** longthen has joined #harmattan | 23:32 | |
| *** longthen1 has quit IRC | 23:33 | |
| *** DrGrov has left #harmattan | 23:37 | |
| *** jaywink_ has quit IRC | 23:42 | |
| itsnotabigtruck | btw it looks like my problem was that X doesn't do quite what i thought it did, in chmod | 23:50 |
| itsnotabigtruck | i thought it was +x for directories only, it's actually +x for directories /and/ files that have any of their +x bits set | 23:51 |
| itsnotabigtruck | that is, it'll set +x on files but it won't clear it | 23:51 |
| *** piggz_ has quit IRC | 23:57 | |
| *** niqt has joined #harmattan | 23:58 | |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!