| Daanct12 | One thing i realized is that Leste uses less ram than stock Maemo | 04:42 |
|---|---|---|
| stan | i just discovered droid4 can playback 960x540 h265 if encoded with -tune fastdecode | 06:16 |
| stan | at least with some material, low bitrate, far superior quality and filesize to h264. cpu has been 70-80% with mplayer | 06:17 |
| stan | this might not be true for all source material: higher framerates than 30fps and high > 1mbit/s bitrates are not likely achievable | 06:19 |
| stan | however anyone transcoding for droid4 may be advised to try shooting for h265 with -fastdecode | 06:20 |
| mighty17[m] | Isn't video decoding done by the 2 Cortex M3s in Droid4? | 06:45 |
| freemangordon | Daanct12: not really, we don;t have contacts and conversations | 07:38 |
| freemangordon | phone application too | 07:38 |
| Danct12 | freemangordon i see, so these essential stuffs uses more ram | 09:12 |
| parazyd | I'd still expect the N900 to be fine once we have all that. | 09:17 |
| parazyd | Also now we have zram, which also helps. | 09:17 |
| parazyd | freemangordon: btw, if you know someone who needs to practice their English: https://hosted.weblate.org/translate/maemo-leste/osso-tutorial-l10n/bg/ | 09:19 |
| parazyd | :p | 09:19 |
| Danct12 | i'm waiting! been using my N900 as hexchat system for few weeks now | 09:56 |
| parazyd | :) | 09:59 |
| Wizzup | freemangordon: ok, I am reday whenever | 10:49 |
| Wizzup | freemangordon: ah, until tuesday next week, ok | 10:50 |
| Wizzup | enjoy | 10:50 |
| Wizzup | Danct12: one reason it uses a bit less ram is maybe because we have a few less apps running still (phone/sms ui is always running on fremantle) | 10:51 |
| bencoh | we could probably reduce memory impact by using a native app here (vs the browser-based Conversation) | 10:56 |
| Wizzup | yup | 11:16 |
| pere | I just tested maemo leste on my n900 just now. it was upgraded a while back, but now it fail to find any wifi networks. | 21:00 |
| Wizzup | pere: dpkg -l | grep supplicant, what versions does it show? | 21:04 |
| pere | no idea where to find | on the keyboard. clues? | 21:06 |
| Wizzup | raise the special virtual keyboard, perhaps | 21:07 |
| pere | looked to the output from dpkg -l, and I see 2:2.7+git20190128+0c1e29f-6+ (and something I can not see) | 21:08 |
| pere | iwconfig show no wifi capable network card. | 21:09 |
| Wizzup | did you reboot after the upgrade? | 21:09 |
| pere | yes | 21:12 |
| pere | it is a while since the upgrade, but I have not looked at the status until today, when I notice there is no network. | 21:12 |
| pere | messed up kernel? | 21:13 |
| Wizzup | if iwconfig does not show a capable network card, it sounds like a kernel problem | 21:13 |
| pere | wlan0 no wireless extension... | 21:16 |
| uvos | iwconfig just fails often these days | 21:16 |
| Wizzup | problem persists after reboot? | 21:16 |
| uvos | on my laptop it claims that i have no wifi too | 21:17 |
| pere | booted twice so far. | 21:17 |
| uvos | its just unmaintained and obsolete | 21:17 |
| uvos | what dose iw say | 21:17 |
| Wizzup | does '/sbin/ifconfig -a' show anything related to wlan0 | 21:17 |
| pere | no iw, the reason I tested iwconfig. | 21:17 |
| pere | Wizzup: yes, wlan show up with a mac address. | 21:18 |
| uvos | then likely iwconfig is just lieing | 21:19 |
| pere | any ideas for a solution to the missing network? | 21:20 |
| Langoor | I just keep having issues with my N900 not booting... It's stuck on a black screen, and when i unlock it with the button (not slider) I see half a lockscreen :s | 21:21 |
| Langoor | I tried 2 SD cards and all builds that are online... I just DD'd the image | 21:21 |
| uvos | lovely to see everything is working fine | 21:22 |
| uvos | pere: try wpa_cli | 21:22 |
| uvos | and see if it can scann | 21:22 |
| uvos | also just leave the device alone for 10 minutes | 21:22 |
| uvos | there is some bug with wifi scanning | 21:22 |
| uvos | it sometimes fails to scann | 21:23 |
| uvos | but will upon the next autoscann interval | 21:23 |
| Wizzup | pere: yes, knowing the wpa supplicant version and the libicd-network-wpasupplicant versions is key | 21:25 |
| Wizzup | Langoor: does usbnet work on that device? | 21:25 |
| pere | "wpa_cli scan" just return OK | 21:29 |
| Wizzup | don't run it like that, run wpa_cli, then from interactive prompt type "scan", wait a bit, and then type "scan_results" | 21:32 |
| pere | get an scan failed event with ret=-22 | 21:33 |
| pere | and the scan_results seem to only list header names. | 21:33 |
| Langoor | Wizzup: uhh.. I'm not sure! | 21:34 |
| Langoor | how do I check? | 21:34 |
| Langoor | should it just expose an usb network interface? | 21:34 |
| Wizzup | Langoor: if you connect it over usb to a linux pc, you might see a usb interface show up | 21:35 |
| Wizzup | pere: hm, that is strange, can you check if apt/dpkg is happy with the system state? | 21:35 |
| pere | just checked dpkg, and it claim something was interrupted. completed the run now. | 21:36 |
| pere | will try a reboot again. | 21:37 |
| Wizzup | *nod* | 21:37 |
| Langoor | it indeed has an usb network interface | 21:40 |
| Wizzup | ok, perhaps you could help us by inspecting it via terminal, if you feel up for it | 21:40 |
| Wizzup | https://leste.maemo.org/Status/USB_Peripheral | 21:40 |
| Wizzup | I have a work meeting in a bit, but if ssh works over the usb link, maybe we can pry inside the device and see why it's so unhappy | 21:41 |
| Langoor | Aight thanks! I added the IP, and changed the interface to the corrosponding one on my pc | 21:43 |
| Langoor | ssh is giving me a connection refused | 21:43 |
| Wizzup | that means that it's likely stuck in the boot process, which i sperhaps not such a surprise given your bug report | 21:43 |
| Wizzup | I assume you typed this: ssh root@192.168.42.2 | 21:43 |
| Langoor | Aye | 21:44 |
| Wizzup | well, we could make ssh start earlier, but that will require mounting the sd card | 21:44 |
| Wizzup | parazyd: should we just always start ssh much earier? | 21:44 |
| Wizzup | s/earier/earlier/ | 21:44 |
| Langoor | the weird thing is that I do get half a lockscreen, which (Seems to me) that something seems to be running | 21:44 |
| Wizzup | yes, it sounds like hildon-desktop is not starting or something | 21:44 |
| uvos | prv and or hildon is failing | 21:45 |
| Wizzup | did we try to inspect the log files on the sd card before? | 21:45 |
| Langoor | nope | 21:45 |
| Wizzup | parazyd: also, should we enable rc_logger ? | 21:45 |
| Wizzup | parazyd: rc_logger="YES" in /etc/rc.conf | 21:46 |
| Langoor | I had the same issue a while ago, and the weird thing was that someone else mentioned the recent builds working fine on their device | 21:46 |
| Wizzup | Langoor: ok, I guess the next step is either analyzing the log files on the device, or adding more debug info | 21:47 |
| Wizzup | can you make a tar of /var/log/ ? | 21:47 |
| Langoor | What would be the best way to turn it off now? | 21:47 |
| Wizzup | hold the power button for ~6 seconds | 21:47 |
| Langoor | I have the targz, any specific place to upload it? | 21:50 |
| Wizzup | anything that is convenient for you I guess | 21:50 |
| Wizzup | if it's small you can also email it to me | 21:50 |
| pere | look like it was the dpkg state. after boot, wifi networks are found. | 21:51 |
| Wizzup | pere: glad to hear that it works now, too bad too hear that the ke-recv thing was the problem | 21:51 |
| Wizzup | I did build a newer version of ke-recv to maybe debug the reset problems | 21:51 |
| Wizzup | (which cause dpkg interruptions) | 21:52 |
| pere | I'm upgrading again now from the command line | 21:52 |
| Langoor | Oh, e-mail is fine | 21:52 |
| Wizzup | mail to merlijn@wizzup.org please | 21:53 |
| parazyd | Wizzup: Regarding ssh, we could add it to sysinit, but this is a bit inconvenient as it'll then slow down charge-mode and all. | 21:53 |
| Wizzup | slow down by how much? | 21:54 |
| parazyd | Wizzup: rc_logger - sure. I'd also start thinking about rc_parallel | 21:54 |
| Wizzup | I think we already do parallel, don't we? | 21:54 |
| parazyd | We don't | 21:54 |
| parazyd | Slowdown, at least 2 services, possibly more | 21:54 |
| Wizzup | I mean in time, if it's not like 5 seconds it's probably worth it for these cases | 21:55 |
| Wizzup | just my opinion anyway | 21:55 |
| pere | why not? sysvinit in debian with startpar would run in parallel. | 21:55 |
| Wizzup | wrt parallel, I think it would be good to have, but I don't want it to create more headaches | 21:55 |
| pere | systemd too. :) | 21:55 |
| parazyd | Well | 21:56 |
| parazyd | Testing it should show real results :) | 21:56 |
| parazyd | biab | 21:56 |
| Wizzup | pere: the point was *less* headaches :P | 21:56 |
| pere | wow, 532 packages to upgrade on my n900 now. been a while since the last time, apparently. | 21:57 |
| Wizzup | pere: oh yeah, maybe you can check if the proper norwegian stuff is in place now | 21:58 |
| pere | "systemd-analyze blame" might provide useful information. was a similar system for sysvinit, do not remember its name. | 21:58 |
| Wizzup | and if not, we should be prtty close to get there | 21:58 |
| Wizzup | pere: we don't use systemd | 21:58 |
| Wizzup | (you probably already knew, just fyi) | 21:58 |
| pere | benchmarking is simple with systemd. did bench startpar boots in Debian using bootchart or something like that before systemd, and it provided useful feedback on what took so long to boot. :) | 21:59 |
| Wizzup | yes, that's nice indeed | 22:09 |
| uvos | what would be the point of doing ssh in sysinit | 22:13 |
| uvos | as oposed to the start of default | 22:13 |
| uvos | its not like your going to be able to ssh in during charge-mode | 22:13 |
| Wizzup | uvos: for Langoor it would allow him to log before hildon is started | 22:25 |
| Wizzup | Langoor: I received the logs | 22:26 |
| Wizzup | uvos: this looks a bit odd: | 22:28 |
| Wizzup | May 29 00:00:46 localhost mce[2603]: Error sending with reply to com.nokia.system_ui.request.powerkeymenu_close: The name com.nokia.system_ui was not provided by any .service files | 22:28 |
| Wizzup | May 29 00:00:47 localhost mce[2603]: Error sending with reply to com.nokia.system_ui.request.tklock_close: The name com.nokia.system_ui was not provided by any .service files | 22:28 |
| Wizzup | it looks like systemui is not starting perhaps | 22:28 |
| Wizzup | oh, it does start after that | 22:29 |
| Wizzup | uvos: can you fix this typo in your mce tree: x11-ctrl: Dsiplay dose not support DPMS | 22:30 |
| Wizzup | 'Display does' as opposed to 'Dsiplay dose' | 22:30 |
| Wizzup | Langoor: did you have a lock code set in fremantle, or not? | 22:31 |
| Wizzup | uvos: also for example if we have ssh, we can type 'rc-status' | 22:33 |
| Langoor | It's the first boot | 22:33 |
| Wizzup | Langoor: right, but the lock codes are not stored on the sd card but somewhere in the hardware | 22:34 |
| Langoor | Ah right | 22:34 |
| Langoor | No, don't think i had | 22:34 |
| Wizzup | (we should deal with those gracefully now, btw, but just checking) | 22:34 |
| Langoor | Fremantle is broken actually, I have booted Leste before on it sucessfully | 22:34 |
| Wizzup | hmmm | 22:35 |
| Wizzup | can you edit /etc/rc.conf on the sd card and set rc_logger="YES" (instead of #rc_logger="NO") | 22:35 |
| Wizzup | and perhaps 'mv /etc/runlevel/default/ssh /etc/runlevel/sysvinit' (this is a guess, I haven't tried, but basically move the symlink to another runlevel) | 22:36 |
| Langoor | Yes! let me try that | 22:37 |
| Wizzup | great, I need to go outside for a bit. I might not be back for ~1.5 hours, but I'll be around after that, and/or tomorrow | 22:38 |
| Wizzup | Langoor: neither of those will make the system automatically work, but they aid in debuggen | 22:38 |
| Wizzup | debugging* even | 22:38 |
| Wizzup | rc_logger will create /var/log/rc.log and having ssh start earlier can help with logging in over usbnet | 22:39 |
| * Wizzup afk | 22:39 | |
| Langoor | I don't have an /etc/rc.conf | 22:39 |
| Wizzup | ok, in that case just added the rc_logger="YES" line | 22:41 |
| Langoor | My bad, I do have it. I made the changes and am now rebooting the phone :) | 22:42 |
| Wizzup | btw, you said fremantle was broken, can you describe how it is broken? | 22:44 |
| * Wizzup afk | 22:45 | |
| Langoor | ssh is working now! | 22:48 |
| Langoor | booting the internal eMMC results in a lot of mmc1(part 0) is current device, invalid partition 3 (/4/2), file not found uImage, FIle not found boot.src | 22:52 |
| Langoor | I don't actually know what happend to fremantle. I didn't install uboot on this device | 22:52 |
| sicelo | if it wants uImage, then u-boot is installed (or flashed) :) | 22:54 |
| sicelo | if you want to try your fremantle (i'm not saying you must, or mustn't), you could start with just flashing or loading the maemo kernel using flasher-3.5 or 0xFFFF. there's stock kernel and power kernel ... possibly stock kernel would be better option (assuming its modules still exist in your fs ...) | 23:00 |
| Langoor | I mean, I could give it a try! | 23:00 |
| Langoor | how uh, would one go about that? I have 0xFFFF installed | 23:03 |
| parazyd | Wizzup: We could try utilizing some of our initscripts to make ssh start sooner in the default runlevel. | 23:23 |
| parazyd | See https://www.funtoo.org/Openrc for depends() keywords | 23:28 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!