*** faari-ddr has quit IRC | 00:01 | |
*** AsiQue has joined #maemo-ssu | 00:10 | |
AsiQue | Hey all, can i run hildon-desktop as a debug application in esbox? | 00:23 |
---|---|---|
*** rhkfin has joined #maemo-ssu | 00:50 | |
rhkfin | Hi there. How long is the CSSU installation supposed to take? I've had it running for about an hour. Black screen, backlight works. No SSH over WLAN response any more. | 00:51 |
RST38h | It is dead, Jim | 00:51 |
rhkfin | .. | 00:51 |
rhkfin | Is there anything else to do but restart and then.. reflash if it doesn't work..? | 00:52 |
rhkfin | I take flashing the rootfs is enough? | 00:55 |
ThreeM | it takes 10min to install cssu | 01:01 |
rhkfin | ok so it's a fail :/ | 01:01 |
rhkfin | I'll try to restart and see what happens.. | 01:01 |
*** ZogG has quit IRC | 01:01 | |
lcuk | rhkfin, can you run over what happened before the black screen, any anomolies | 01:02 |
*** ZogG has joined #maemo-ssu | 01:02 | |
rhkfin | Nothing I'd concider unexpected. went to wiki page, clicked the install button, started SSU, asked for permissions to add repository, install, warning about SSU being unstable or so. Then it finished, I ran the CSSU app from menu, that started the SSU again to upgrade CSSU. Recommended backups, did that and.. hmm.. don't know/remember how far I got.. | 01:04 |
rhkfin | Hmm.. on the first run, I did press cancel when being asked a question, just to restart and take screenshots of what happens. I think it was the 'are you sure you want to install CSSU....' screen. | 01:05 |
rhkfin | lcuk: Repository was already installed and package downloaded when I pressed cancel so I suppose it was that extra 'are you sure...' question | 01:06 |
DocScrutinizer | rhkfin: aiui CSSU needs a reboot at a particular point n time. | 01:06 |
rhkfin | DocScrutinizer: ok.. and if it doesn't get that it'll fail & kill the system? | 01:07 |
rhkfin | So maybe I now restart. | 01:07 |
DocScrutinizer | for sure waiting another hour won't help anything ;-) | 01:08 |
rhkfin | I pressed the power button shortly and got the menu! | 01:08 |
DocScrutinizer | heh | 01:08 |
DocScrutinizer | :-D | 01:08 |
rhkfin | the power menu, I mean. But that doesn't really help, I take.. | 01:08 |
rhkfin | now i saw the keylock screen ;) | 01:09 |
DocScrutinizer | I'd guess system can't be really messed up completely if you get that powerbutton menu | 01:09 |
rhkfin | right | 01:09 |
DocScrutinizer | sometimes even *nix needs a reboot ;-D | 01:10 |
rhkfin | There's not much useful I can do in the powerbutton menu | 01:10 |
DocScrutinizer | shut down maybe? | 01:10 |
rhkfin | DocScrutinizer: indeed, restart on N900 s sometimes needed.. | 01:10 |
rhkfin | yes, shutdown - and it actually restarts automatically | 01:10 |
rhkfin | no, it didn't.. | 01:11 |
DocScrutinizer | nota bene a plugged charger keeps device from shutting down normally | 01:12 |
rhkfin | ok, I see, that's why. | 01:12 |
rhkfin | Asked for code and started to desktop. NOt bad :) | 01:12 |
rhkfin | yay, seems to work, at least the portrait menu mode.. | 01:13 |
DocScrutinizer | that's why reflashing (actually reboot after flashing) is best done on battery power, not when plugged to any USB | 01:14 |
rhkfin | the scrolling is FAAAAAAST | 01:14 |
rhkfin | right.. But installed recommended to use wall charger.. Maybe this could be added to FAQ? | 01:14 |
DocScrutinizer | It should | 01:15 |
rhkfin | I guess there isn't much that the software could do to make it work better? | 01:15 |
*** andre__ has quit IRC | 01:15 | |
DocScrutinizer | actually it should get added to reboot warning when flashing kernels etc, everything that needs a reboot should notify user to unplug USB prior to boot | 01:16 |
rhkfin | true | 01:16 |
DocScrutinizer | prior to shutdown to be precise | 01:17 |
ThreeM | hmm | 01:19 |
ThreeM | when i use the backup tool, will it restore all configurations include all installed apps? | 01:19 |
lcuk | ThreeM, a backup will write all configuration options for all apps that have them configured | 01:21 |
lcuk | and also a list of applications to install | 01:21 |
rhkfin | Added the boot issue to FAQ: http://wiki.maemo.org/Community_SSU/Installation_FAQ | 01:21 |
ThreeM | that is very nice. thx | 01:21 |
lcuk | when you restore, it first restores the options to the folders, then presents HAM with a ticklist to reinstall the apps | 01:21 |
lcuk | it saves the selected repository lists too | 01:21 |
lcuk | so, generally a very thorough process | 01:22 |
rhkfin | that is indeed nice! | 01:22 |
rhkfin | Didn't know it works that well! | 01:22 |
DocScrutinizer | lcuk: (thorough process) only if the repo still has an installable version of the app | 01:26 |
lcuk | DocScrutinizer, it does the best it possibly can. | 01:26 |
DocScrutinizer | which it doesn't for most of the PR1.2 -testing and -devel apps | 01:27 |
* lcuk finds that admirable | 01:27 | |
DocScrutinizer | and sure it's admirable, just a pity the repo management broke it for PR1.2 | 01:28 |
rhkfin | :D Can't take a screenshot with ctrl+shift+P on portrait mode ;) | 01:29 |
rhkfin | ('cause no portrait if keyboard's shown :) | 01:30 |
*** lardman has joined #maemo-ssu | 01:31 | |
BCMM | rhkfin: just don't open the keyboard all the way | 01:32 |
BCMM | or override with ctrl-alt-blue-l | 01:32 |
rhkfin | uhhhh... ctrl shift blue L - hard stuff for my fingers ;) | 01:32 |
rhkfin | But succeeded :) | 01:33 |
*** ZogG has quit IRC | 01:33 | |
*** lardman_ has quit IRC | 01:34 | |
*** ZogG has joined #maemo-ssu | 01:34 | |
rhkfin | Interesting, so ctrl+shift+blue+L for example on menu doesn't reorganize the icons so many are left out of the screen? | 01:38 |
DocScrutinizer | ctrl-shift-blue-<any-key> is hard stuff for the keyboard matrix as well. There are quite a number of combinations with only two qualifier keys plus regular char key that simply fail as the matrix can't detect them resp can't distinguish which of two combinations resulting in same bit pattern on matrix lines is the one actually pressed | 01:45 |
DocScrutinizer | so whoever invented ctrl-shift-blue-L better makes damn sure there's no collisions on matrix for that particular one | 01:46 |
rhkfin | haha :) | 01:48 |
*** NIN101 has quit IRC | 01:48 | |
DocScrutinizer | see http://wiki.maemo.org/N900_Hardware_Subsystems#Keyboard | 01:49 |
*** mirakels has quit IRC | 01:50 | |
*** lardman has quit IRC | 01:50 | |
rhkfin | Do you remember what was the bug that some process (the lock app?) started to eat resources slowly requiring a restart every two days or so? | 01:54 |
rhkfin | (hmm.. my dad found some weird keyboard layout bug, need to ask what was it about.. I think it was that the layout and the produced characters didn't match on cyrillic.. something..) | 01:56 |
DocScrutinizer | iirc that's been mere swap fragmentation, when the sum of sequential writes hit the end of the 768M partition so following writes had to use scattered blocks | 01:56 |
rhkfin | Has there been a fix / is it fixable? | 01:59 |
DocScrutinizer | this is a hw limitation and not fixable | 02:08 |
DocScrutinizer | ooh, you mean swap frag | 02:08 |
DocScrutinizer | yeah, swapon <new-swap> && swapoff <used-swap> | 02:09 |
rhkfin | ok, I see.. | 02:11 |
rhkfin | is there a way to check the fragmentation rate (to see if that's the one causing the slowing down) | 02:11 |
rhkfin | well, ignore that.. | 02:12 |
DocScrutinizer | shadowjk has a script for that :-) http://enivax.net/jk/n900/reswap.sh.txt | 02:14 |
rhkfin | nice, thanks! | 02:15 |
ThreeM | for what? | 02:15 |
rhkfin | ThreeM: 'reset' swap | 02:15 |
rhkfin | along time it gets fragmented -> slow | 02:15 |
rhkfin | resetting makes it faster again | 02:15 |
ThreeM | öhm, its flashmemory or? | 02:16 |
ThreeM | how can fragmentation cause lower performance?! | 02:16 |
rhkfin | ThreeM: 01:56 < DocScrutinizer> iirc that's been mere swap fragmentation, when the sum of sequential writes hit the end of the 768M partition so following writes had to use scattered blocks | 02:17 |
ThreeM | oh sorry im overread it | 02:17 |
rhkfin | np | 02:18 |
rhkfin | now zzz.. | 02:18 |
DocScrutinizer | ThreeM: on flash fragmentation is a huge problem, as flash can write to ~128k "sectors" only | 02:18 |
DocScrutinizer | so each write of a 512byte block can cause an actual read-modify-write cycle of a new 128k sector | 02:19 |
DocScrutinizer | so, on fragmentation, the write speed may slow down by a factor (based on above figures) of 256 | 02:21 |
DocScrutinizer | that's worst case of course, just to show you there *is* an issue | 02:21 |
*** merlin1991 has joined #maemo-ssu | 02:39 | |
*** wmarone__ has joined #maemo-ssu | 03:03 | |
*** lcuk2 has joined #maemo-ssu | 03:03 | |
*** wmarone_ has quit IRC | 03:04 | |
*** Dhraakellian has quit IRC | 03:04 | |
*** lcuk has quit IRC | 03:06 | |
*** chainsawbike has quit IRC | 03:08 | |
*** chainsawbike has joined #maemo-ssu | 03:11 | |
*** AsiQue has quit IRC | 03:21 | |
*** merlin1991 has quit IRC | 03:22 | |
*** Dhraakellian has joined #maemo-ssu | 03:23 | |
*** BCMM has quit IRC | 03:28 | |
*** trbs has quit IRC | 03:28 | |
*** ZogG has quit IRC | 03:39 | |
*** ZogG has joined #maemo-ssu | 03:39 | |
*** M4rtinK has quit IRC | 04:18 | |
*** ZogG has quit IRC | 04:22 | |
*** ZogG has joined #maemo-ssu | 04:22 | |
*** lcuk2 has quit IRC | 04:25 | |
*** ZogG has quit IRC | 04:57 | |
*** ZogG has joined #maemo-ssu | 04:57 | |
*** lcuk has joined #maemo-ssu | 05:46 | |
*** lcuk has quit IRC | 05:46 | |
*** lcuk has joined #maemo-ssu | 05:46 | |
*** psycho_oreos has joined #maemo-ssu | 06:00 | |
*** ZogG has quit IRC | 07:29 | |
*** ZogG has joined #maemo-ssu | 07:29 | |
*** DocScrutinizer has quit IRC | 07:31 | |
*** DocScrutinizer has joined #maemo-ssu | 07:31 | |
*** fiberspeed has quit IRC | 09:18 | |
*** mirr0r has joined #maemo-ssu | 09:18 | |
*** mirr0r is now known as fiberspeed | 09:48 | |
*** AsiQue has joined #maemo-ssu | 10:39 | |
*** NIN101 has joined #maemo-ssu | 10:42 | |
*** psycho_oreos has quit IRC | 11:00 | |
MohammadAG | isn't this all off topic here :p | 11:25 |
Lava_Croft | and isnt it solved by just rebooting | 11:27 |
Lava_Croft | non issue | 11:27 |
Jaffa | Morning, all | 11:40 |
Lava_Croft | mo | 11:43 |
thp | hey ho | 12:01 |
thp | can we get rid of the system-upgrade-type updates for "small" cssu cssu updatse? so we don't always need to restart and close all apps on each update? | 12:01 |
Jaffa | thp: I don't see how - you don't know how big a jump the user is taking | 12:10 |
Lava_Croft | that could be determined during the install process | 12:13 |
Jaffa | Lava_Croft: But the reboot flag for HAM is on the package description. So it'd need a new bit of code writing to simulate it | 12:13 |
Jaffa | ...dynamically at runtime. | 12:14 |
Jaffa | Patches welcome :-) | 12:14 |
Lava_Croft | :) | 12:15 |
rhkfin | MohammadAG: what is on-topic here? SSU yes but.. how to limit it.. I this case I started talking about the bug (that's ropably the swap thing as DocScrutinizer wrote) thinking if it is/could/will be included in CSSU.. | 12:15 |
rhkfin | Lava_Croft: also UI freezes are solved by just rebooting -> not issues..? | 12:15 |
MohammadAG | rhkfin, ah, nevermind, thought it was just random discussion | 12:16 |
rhkfin | ok :) | 12:16 |
rhkfin | Other than that, CSSU is great. | 12:16 |
rhkfin | It just gives an extra boost to the UI speed making it nice smooth | 12:17 |
MohammadAG | Jaffa, do you mind testing if changing the 0 to 1 in /etc/powervr.d/hildon-desktop.ini breaks anything? | 12:20 |
MohammadAG | had it set for 3 days, no problems, it used to crash the whole device before | 12:21 |
MohammadAG | so I'm guessing it was fixed upstream, somewhere | 12:21 |
Lava_Croft | same, still no crash | 12:23 |
Lava_Croft | non-cssu friend srtill no crash too | 12:23 |
Lava_Croft | -r | 12:23 |
Jaffa | MohammadAG: And what's the effect? | 12:24 |
MohammadAG | Jaffa, smoother transitions | 12:25 |
Jaffa | Looks shitter when panning desktop | 12:26 |
Jaffa | (to me: more bunching up, like vertical interlacing on a non-interlaced display) | 12:27 |
MohammadAG | i'm guessing you restarted the desktop? | 12:27 |
Jaffa | Rebooted | 12:27 |
Jaffa | Transitions into and out of task switcher/launcher are smoother but slower. | 12:27 |
Jaffa | Default transitions.ini with zaxisrotation | 12:27 |
* MohammadAG has blurless set to 1 | 12:28 | |
*** marcels has joined #maemo-ssu | 13:41 | |
*** BCMM has joined #maemo-ssu | 13:47 | |
*** Venemo has joined #maemo-ssu | 13:58 | |
*** obironbo has quit IRC | 14:06 | |
*** M4rtinK has joined #maemo-ssu | 14:06 | |
*** _NIN has joined #maemo-ssu | 14:08 | |
*** NIN101 has quit IRC | 14:09 | |
*** psycho_oreos has joined #maemo-ssu | 14:10 | |
thp | Jaffa: hmm as a power-user, i could just dist-upgrade from the command line and have the same effect, right? | 14:16 |
thp | MohammadAG: so vsync works now? | 14:17 |
*** merlin1991 has joined #maemo-ssu | 14:19 | |
MohammadAG | thp, it's not vsync | 14:19 |
Jaffa | thp: Yes, but as a power user you can deal with the fall out ;-) | 14:19 |
MohammadAG | vsync is missing on the kernel level :p | 14:20 |
Jaffa | thp: Detecting whether a reboot is necessary in mp-fremantle-community-pr and prompting in postinst seems sensible though | 14:20 |
MohammadAG | thp, upgrade will be enough | 14:20 |
thp | MohammadAG: so what's that setting in that file in powervr.d then? | 14:20 |
Jaffa | thp: Setting hwsync=1 degrades performance when using blurless=0 | 14:20 |
MohammadAG | thp, it just makes sure all frames take the same time afaik | 14:20 |
MohammadAG | or something like that | 14:20 |
thp | Jaffa: just thought it'd be a good idea, as we have such a high update frequency now | 14:20 |
Jaffa | And I'm seeing a small graphics corruption (about 4x3 pixels) on notifications | 14:21 |
Jaffa | thp: Agreed. Just a question of doing the work :-) | 14:21 |
MohammadAG | Jaffa, that's a bug in hildon-desktop | 14:21 |
MohammadAG | it corrupts graphics randomly | 14:21 |
thp | Jaffa: for me, doing it with apt-get in a terminal is fine, so no need to implement it for me ;) | 14:21 |
Jaffa | MohammadAG: Yeah, but I'd not seen it until the reboot with hwsync=1 | 14:21 |
MohammadAG | reboot again, it'll be gone :p | 14:22 |
MohammadAG | it's a coincidence I guess | 14:22 |
MohammadAG | then again... | 14:22 |
thp | Jaffa: maybe a side effect of OCing? | 14:22 |
MohammadAG | Jaffa, highlight me | 14:22 |
Jaffa | thp: Not overclocked, but looks similar to the effect you get with undervoltage | 14:28 |
Jaffa | MohammadAG: ? | 14:28 |
thp | Jaffa: (offtopic) for the qt quick compile problem: add #include <QtDeclarative> at the top - that could help :) | 14:30 |
*** trbs has joined #maemo-ssu | 14:40 | |
*** ^24seven_ is now known as ^24seven | 14:55 | |
*** crashanddie_ has joined #maemo-ssu | 15:16 | |
*** crashanddie_ has joined #maemo-ssu | 15:16 | |
*** andre__ has joined #maemo-ssu | 15:31 | |
*** andre__ has joined #maemo-ssu | 15:31 | |
Jaffa | thp: It did indeed. Thanks :-) | 15:37 |
Jaffa | thp: Now I've found the SIS files for Symbian and can install them over BT, I'm hoping to see it running ton the N8 too | 15:37 |
Jaffa | thp: Is there a constant in Qt/QML for "gravity" e.g. 9.8... | 15:38 |
Jaffa | ? | 15:38 |
rhkfin | Jaffa: it's not constant ;) | 15:39 |
Jaffa | rhkfin: No, but... | 15:39 |
rhkfin | You need to use the location of the user and then correlate that to... ;D | 15:40 |
Lava_Croft | Maybe anyone here knows, but if you make a parallel between PC games | 15:40 |
rhkfin | (no, I have no answer) | 15:40 |
Lava_Croft | what would be the prettiest PC game that would run on the N900 | 15:40 |
Lava_Croft | if ported properly | 15:40 |
Lava_Croft | I know that Quake3 era stuff runs fine | 15:40 |
Lava_Croft | A friend of mine is creating http://osirion.org/ and wonder if N900 would be powerful enough | 15:40 |
Lava_Croft | Just to stay completely on topic ;) | 15:41 |
rhkfin | Lava_Croft: something with an easy UI to play.. | 15:41 |
Lava_Croft | rhkfin: i mean performance wise | 15:41 |
rhkfin | right | 15:41 |
* psycho_oreos wonders how is this even remotely related to the use of CSSU packages | 15:42 | |
psycho_oreos | and no, its been tested before that N900 can struggle when its running stuff made for PCs | 15:43 |
rhkfin | psycho_oreos: hmm.. give us 10 minutes and we'll come up with an explanation ;) | 15:43 |
Jaffa | More people in #maemo; use that :-p | 15:43 |
Lava_Croft | oreo is a lovely kind of cookie, but the answer is a bit odd | 15:43 |
psycho_oreos | rhkfin, more like a boot :) | 15:43 |
Lava_Croft | Jaffa: true, im just lazy | 15:43 |
Lava_Croft | which is why i already mentioned how it was so on topic :) | 15:43 |
psycho_oreos | s/on/off/ | 15:44 |
Lava_Croft | 1441 < Lava_Croft> Just to stay completely on topic ;) | 15:44 |
Lava_Croft | on :) | 15:44 |
Lava_Croft | anyhow, this is silly | 15:44 |
psycho_oreos | run his game on N900 and then see the performance, no need to go pondering | 15:44 |
Lava_Croft | not everything is a simple matter of just running it on the N900 | 15:45 |
Lava_Croft | but thanks for the advice | 15:45 |
psycho_oreos | indeed not everything is as easy as making it to run on N900 but different platforms serves different purposes. N900 is classed as embedded device, PC on the other hand you have people talking about x86 arch not the arm arch | 15:46 |
Lava_Croft | fantastic news, but how is this useful | 15:47 |
Lava_Croft | thanks for the advice anyway | 15:47 |
psycho_oreos | it wasn't meant to be useful, it was a mere comment | 15:48 |
AsiQue | MohammadAG: how do i install and debug athe hildon-desktop under esbox? is it possible? i can't even install the .deb file | 15:53 |
AsiQue | i hope i'm not off-topic here... :-) had my wrist slapped once or twice... | 15:55 |
AsiQue | anyone else? | 15:57 |
merlin1991 | AsiQue, you can't go more ot than Lava_Croft just did :D | 16:06 |
Lava_Croft | That's not even remotely hard to accomplish, really! :D | 16:07 |
AsiQue | sorry, didn't understand that... | 16:09 |
* rhkfin is a bit surprised how strictly on-topicness (?) is required here | 16:09 | |
AsiQue | and now i did :-) | 16:10 |
rhkfin | (not that it'd be wrong but it surprises every now and then :) | 16:10 |
merlin1991 | I dunno, I never saw anything happening about ot in here | 16:11 |
rhkfin | ot being on or off? | 16:15 |
AsiQue | anyway... i'm trying to debug hildon-desktop but the console tells me that i need to be root (on the emulator) so is it possible? | 16:16 |
*** rhkfin has left #maemo-ssu | 16:29 | |
*** AsiQue has left #maemo-ssu | 17:04 | |
*** AsiQue has joined #maemo-ssu | 17:04 | |
*** AsiQue has quit IRC | 17:45 | |
*** AsiQue has joined #maemo-ssu | 17:52 | |
*** normaltusker has joined #maemo-ssu | 17:52 | |
*** _NIN has quit IRC | 19:12 | |
*** NIN101 has joined #maemo-ssu | 19:13 | |
*** merlin1991 has quit IRC | 19:25 | |
*** normaltusker has quit IRC | 19:26 | |
*** fiberspeed has quit IRC | 19:28 | |
*** crashanddie_ has quit IRC | 20:01 | |
*** AsiQue has quit IRC | 20:20 | |
*** marcels has quit IRC | 20:31 | |
*** Sc0rpius has quit IRC | 20:36 | |
*** Brownout_ has joined #maemo-ssu | 20:46 | |
*** w00t___ is now known as w00t_ | 20:46 | |
*** w00t_ has joined #maemo-ssu | 20:46 | |
*** Brownout has quit IRC | 20:47 | |
*** Sc0rpius has joined #maemo-ssu | 20:49 | |
*** BCMM has quit IRC | 21:02 | |
*** marcels has joined #maemo-ssu | 21:06 | |
*** andre__ has quit IRC | 21:13 | |
*** andre__ has joined #maemo-ssu | 21:24 | |
*** BCMM has joined #maemo-ssu | 21:31 | |
*** trumee has quit IRC | 22:34 | |
*** trumee has joined #maemo-ssu | 22:36 | |
*** marcels has quit IRC | 23:05 | |
*** jpala16 has joined #maemo-ssu | 23:25 | |
*** arcean has joined #maemo-ssu | 23:47 | |
*** jpala16 has quit IRC | 23:50 | |
*** s1gk1ll has quit IRC | 23:51 | |
*** s1gk1ll has joined #maemo-ssu | 23:51 | |
*** ZogG_laptop has joined #maemo-ssu | 23:58 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!