| rafael2k | hi all | 09:48 |
|---|---|---|
| stano | hi rafael2k | 09:49 |
| rafael2k | the latest kernel upgrade for pinephone was a mixed bag for X behavior... the behavior with accelerated OpenGL improved, still a bit strange, but improved | 09:50 |
| rafael2k | but the NoAccel, which was the stable setup for me... got a bit unstable, with X crashing sometimes | 09:50 |
| stano | is anyone lending out pinephones to test apps on | 09:50 |
| rafael2k | something to ask, definitelly | 09:52 |
| stano | would you be interested in testing some sdl apps? | 09:53 |
| rafael2k | I can test, sure | 09:55 |
| stano | there's a backlog of stuff i have here that doesn't need much to get in repo, but testing on other devices is important | 09:56 |
| rafael2k | cool | 09:58 |
| rafael2k | point me which apps | 09:58 |
| stano | you can compile them? | 09:59 |
| stano | the debian build tools and "dpkg-buildpackage -rfakeroot -b -uc -us" should work to create a deb | 10:00 |
| rafael2k | yes, I can compile with dpkg-buildpackage | 10:01 |
| stano | then i'll try to get picodrive ready. ETA tomorrow | 10:03 |
| Wizzup | morn | 10:05 |
| Wizzup | rafael2k: opengl improved, that's good, is the flickering gone? | 10:06 |
| Wizzup | could also try to switch from GL and GL|ES and see which one is better | 10:06 |
| parazyd | rafael2k: Is this the updates I made yesterday in -devel? | 10:11 |
| parazyd | Kernel is 5.14 and mesa is 21.2.0 | 10:22 |
| parazyd | I also updated crust and uboot which are complementary to this kernel | 10:22 |
| Wizzup | upgrading my pp to see | 10:26 |
| parazyd | <Wizzup> could also try to switch from GL and GL|ES and see which one is better | 10:57 |
| parazyd | Do you mean with the hildon-desktop.env file we have in etc? | 10:57 |
| Wizzup | yes | 11:00 |
| Wizzup | rafael2k: mostly feels the same to me (gl) | 11:13 |
| Wizzup | my pp doesn't rotate again it seems | 11:13 |
| stano | there's just something about the 'feel' of the n900 that's very satisfying | 11:40 |
| stano | hardware wise. the animations on fremantle are jerky compared to leste on droid4 though | 11:41 |
| bencoh | stano: I think that feeling comes from the keyboard, and maybe the quality of the lcd display | 11:42 |
| Wizzup | uvos: the 11 tablets arrived | 11:42 |
| stano | for me it's the case material and slider mechanism | 11:43 |
| bencoh | ah | 11:43 |
| Wizzup | uvos: I could try to check some today and maybe send one your way | 11:43 |
| uvos__ | stano: bencoh: i think its the terrible case quality/ slider design on the d4 particularly the unsuported display back plate that inevitably bends and causes the display to not sit flush with the body when closed | 11:48 |
| uvos__ | the n900 or something like the d1 is mutch better better | 11:48 |
| uvos__ | Wizzup: allready? sweet | 11:48 |
| uvos__ | Wizzup: monitor-sensor on pp? | 11:49 |
| stano | oh i have an old pop3 email password on this n900 but it's all asterisks | 11:53 |
| Wizzup | uvos__: will debug later @ pp | 11:56 |
| Wizzup | uvos__: yeah already, I'm leaving home tomorrow though, so need to act soon on posting if useful | 11:57 |
| Wizzup | s/posting/sending/ | 11:57 |
| stano | will n900 and droid4 continue to work with german phone network? i heard something about 3g being deprecated | 12:04 |
| inetdragon | Regarding: "<Wizzup> could also try to switch from GL and GL|ES and see which one is better", as far as I know, GLX doesn't work on N900. | 12:07 |
| stano | the phone in question was pine64 | 12:07 |
| uvos__ | stano: some operators shut down umts on the 1st of this moth, the last 3g network wil be gone on the 1 jan 2022 in ger | 12:11 |
| stano | but gsm continues? or edge whatever it's called? | 12:13 |
| uvos__ | stano: the d4 can sill use gsm after that for phone calls but effectively mobile internet is gohne, as several operators depreciated EDGE too | 12:13 |
| stano | ok | 12:13 |
| uvos__ | so you get GPRS | 12:13 |
| uvos__ | with dialup modem speed | 12:13 |
| uvos__ | (not usefull for anything except maybe ip messaging) | 12:13 |
| stano | pop3 email perhaps | 12:13 |
| uvos__ | sure | 12:13 |
| uvos__ | email is just old ip messaging :) | 12:14 |
| uvos__ | Wizzup: obv if you can send before then that would be nice, but there is no rush | 12:15 |
| uvos__ | and im pretty swamped undtill mid aug anyhow | 12:16 |
| Wizzup | check | 12:24 |
| uvos__ | stano: lcm2.0 probubly could do lte band 28 that is used by voafone in germany, but dont hold your breath of anyone ever get it going sw wise as its locked and the sw is proriatary | 12:25 |
| uvos__ | also would be illegal and the pentalies for operating sutch a device are stiff | 12:26 |
| inky | hey folks | 12:57 |
| inky | so i confirm, on my pinephone too, with latest image | 12:58 |
| uvos__ | hi | 12:58 |
| inky | no battery indicator | 12:58 |
| uvos__ | no verb | 12:58 |
| inky | i got it yesterday and flushed leshe to sdcard | 12:58 |
| inky | will also try to flush to emmc | 12:58 |
| uvos__ | what do you meen flush | 12:58 |
| uvos__ | what device | 12:58 |
| uvos__ | *mean | 12:59 |
| inky | looks very cool on a big screen | 12:59 |
| inky | pinephone | 12:59 |
| uvos__ | ah flashed i guess | 12:59 |
| inky | flush or flash? i mix em up always. dd, i mean dd | 12:59 |
| uvos__ | inky: ok | 12:59 |
| uvos__ | hmm | 12:59 |
| inky | i reported that when tried memo leste on friends' device | 13:00 |
| inky | now i got mine and the same issue | 13:00 |
| uvos__ | dose upower show the battery | 13:00 |
| uvos__ | ? | 13:00 |
| uvos__ | is the battery in /sys/class/power_supply/* | 13:00 |
| inky | upower? is that console utility? | 13:00 |
| inky | let me check | 13:00 |
| inky | upower shows no output | 13:02 |
| inky | willzcheck /sys/class | 13:02 |
| uvos__ | it might not be upower excatly that might be the deamon not sure (can check atm) | 13:02 |
| uvos__ | upower -d | 13:03 |
| uvos__ | i think | 13:03 |
| inky | sys class contains axp20x-battery | 13:04 |
| inky | probably a dir | 13:04 |
| uvos__ | thats fine | 13:04 |
| uvos__ | upower -d? | 13:04 |
| inky | mmm, a lot of output | 13:05 |
| inky | let me try to wgetpaste it | 13:05 |
| inky | it says discharging | 13:05 |
| inky | battery present | 13:05 |
| uvos__ | sounds fine | 13:05 |
| inky | percentage 65 | 13:05 |
| uvos__ | maybe the battery status applet is not installed | 13:06 |
| uvos__ | check that | 13:06 |
| uvos__ | (idk what the package is called atm) | 13:06 |
| inky | can i start battery nidicator from console to see if it crashes or somuthing | 13:06 |
| inky | let me make sure with dpkg | 13:06 |
| inky | status-area-applet-battery present | 13:08 |
| inky | let me find the binary | 13:08 |
| inky | no binary, it contains sdared object | 13:09 |
| inky | i cannot run it to check | 13:09 |
| inky | shared* | 13:09 |
| uvos__ | you need to run it with meamo-summoner | 13:10 |
| inky | maemo-summoner [space] sdared object? | 13:11 |
| uvos__ | iirc yes | 13:11 |
| Wizzup | I don't see a technology file | 13:11 |
| Wizzup | that's very likely the problem | 13:11 |
| uvos__ | Wizzup: create one | 13:11 |
| Wizzup | it looks like my patch still applies but needs to be changed | 13:11 |
| Wizzup | https://github.com/maemo-leste/pine64-kernel/blob/c62a8aeab344523190c7f539de60d94ba6c9cc5f/drivers/power/supply/axp20x_battery.c#L183 | 13:11 |
| Wizzup | parazyd: what is our most recent branch | 13:12 |
| Wizzup | https://github.com/maemo-leste/pine64-kernel/tree/maemo-5.14 ? | 13:12 |
| parazyd | This + patches in beowulf-devel branch | 13:13 |
| Wizzup | well, the patch is not working, there is no technology file it seems | 13:14 |
| Wizzup | is that is not there, everything in maemo ignores it | 13:14 |
| Wizzup | it is missing a return | 13:15 |
| Wizzup | return 0; | 13:15 |
| Wizzup | the flow of the function changed | 13:15 |
| Wizzup | so it sets the value but then returns -EINVAL | 13:15 |
| Wizzup | mystery solved :) | 13:16 |
| inky | uvos, i get undefined symbol | 13:29 |
| Wizzup | inky: no need to try status applet for battery | 13:30 |
| Wizzup | the problem is the sysfs entry | 13:31 |
| Wizzup | but I think my above solution might not work, since it does seem to return 0 | 13:31 |
| Wizzup | but the problem for sure is that the technology sysfs file is missing | 13:31 |
| uvos__ | why would the status applet care about technology anyhow | 13:31 |
| uvos__ | upower is picking up the battery it seams | 13:32 |
| uvos__ | type should be enough to ensure its a battery no? | 13:32 |
| inky | https://pasteboard.co/KaYntg4.jpg | 13:32 |
| inky | anyway that's what i get | 13:32 |
| Wizzup | uvos__: because the technology matters, it's used to filter out bogus batteries | 13:34 |
| Wizzup | the n900 has several, for example | 13:34 |
| uvos__ | that sounds like a hack though | 13:35 |
| uvos__ | i know n900 has 2 charger chips | 13:35 |
| uvos__ | one falsely reports the connected battery as not li-ion? | 13:35 |
| Wizzup | well, it's how it currently always worked, and the pinephone not specifying the battery tech is also wrong | 13:35 |
| uvos__ | not really its fine by spec | 13:35 |
| Wizzup | uvos__: something like that, yeah | 13:35 |
| uvos__ | the circut might not know | 13:35 |
| inky | and output of ubattery: https://pasteboard.co/KaYp9mg.jpg | 13:36 |
| uvos__ | like if they are regular AA batteries or whatever | 13:36 |
| uvos__ | cant we just blacklist one of the n900 chips by name? | 13:37 |
| inky | another question (again) have you heard of megi's patches? | 13:37 |
| Wizzup | they are there. | 13:37 |
| inky | tdat inccease battery life | 13:37 |
| inky | are they in mainline? or you apply to tde kernel neparately? | 13:37 |
| inky | separately | 13:37 |
| Wizzup | https://github.com/maemo-leste/pine64-kernel/branches | 13:39 |
| Wizzup | uvos__: that sounds more ugly than properly exporting the technology of the battery of the device, which we know | 13:39 |
| Wizzup | https://github.com/maemo-leste/mce/blob/master/modules/battery-upower.c#L353 | 13:40 |
| Wizzup | + https://github.com/maemo-leste/status-area-applet-battery/blob/master/batmon.c#L64 | 13:42 |
| uvos__ | hmm relying on some unrelated (broken) implmentation detail of the n900's chip to detect it that can be valid on other deivces (like ones with primary cells, ok rare i know) seams way more ugly than a blacklist based on the chipname | 13:43 |
| uvos__ | yes we should also fix the pp kernel to show the right tech | 13:43 |
| parazyd | inky: We're using megi's 5.14 kernel branch already | 13:45 |
| parazyd | In devel | 13:45 |
| uvos__ | mce also uses a blacklist by chipname | 13:45 |
| uvos__ | just pull the more sane code from there into the status applet | 13:45 |
| uvos__ | https://github.com/maemo-leste/mce/blob/81c3c41bef4d0d21924138e450d279105b3ad10f/modules/battery-upower.c#L68 | 13:45 |
| Wizzup | yup, possible | 14:00 |
| Wizzup | still, mce also checks for tech | 14:01 |
| Wizzup | like I said | 14:01 |
| inky | i have added /leste beowulf-devel | 15:56 |
| inky | and i am doing upgrade from it | 15:56 |
| inky | it brings me linux-image 5.14 | 15:57 |
| inky | and uboot | 15:57 |
| inky | no battery indicator with this either | 16:01 |
| inky | Wizzup do you have the indicator with devel kernel? | 16:02 |
| parazyd | It's known to be broken on pinephone, we'll fix it. | 16:03 |
| inky | okay (: | 16:04 |
| rafael2k | btw | 16:25 |
| rafael2k | I have no battery indicator | 16:25 |
| rafael2k | since a long time | 16:25 |
| rafael2k | anything I should try? | 16:26 |
| rafael2k | parazyd: updated to -devel, including kernel 5.14 | 16:26 |
| rafael2k | stano: OK, I already built firefox, so I think the pp will be fine | 16:27 |
| rafael2k | btw, when X crashed, it seems the system loses the network configuration, and I can not ssh to it | 16:29 |
| rafael2k | parazyd: how to switch to opengl ES? | 16:30 |
| rafael2k | Wizzup: behavior in X, for example, the vkb keeps going black, with half the widgets in the lower part going black, also in firefox, the flickring continues, but reduces | 16:31 |
| rafael2k | *reduced | 16:31 |
| rafael2k | indeed, almost useable with GL enabled, but not yet (at least to me) | 16:32 |
| parazyd | fwiw I think in portrait it's pretty usable | 16:34 |
| parazyd | Need to still try the updates I made yesterday | 16:35 |
| uvos | btw i can confirm that https://www.ebay.de/itm/173307495965?hash=item2859ee861d:g:rwMAAOSw1fJa8agr work really well with mapphones | 16:36 |
| uvos | in mode 1 its powerd by cpcap and powers the downstream ports from it too and plug detection is reliable | 16:37 |
| parazyd | Nice | 16:37 |
| uvos | in mode 3 it charges the maphone | 16:37 |
| parazyd | So both otg and charge at the same time or? | 16:37 |
| uvos | and downstream usb device work great too | 16:37 |
| uvos | parazyd: yeah | 16:37 |
| parazyd | Sounds ideal :) | 16:37 |
| uvos | you can choose to charge the d4 or take its power for downstream devces :) | 16:37 |
| uvos | it really is | 16:37 |
| uvos | i can use it for the bonic now to have a kbd for debugging :) | 16:38 |
| parazyd | :) | 16:38 |
| parazyd | Yeah and perhaps the tablets | 16:38 |
| uvos | yeah | 16:38 |
| buZz | uvos: does it output 5V? | 16:38 |
| buZz | or 3.3V? | 16:38 |
| uvos | 5v, cpcap supplys 5v @ 200mah max to the hub | 16:38 |
| buZz | nice! | 16:38 |
| uvos | and that powers the downstream devices | 16:38 |
| uvos | if you switch the litte switch on the hub to 3 and insert a charger to the thing soccet with a lightning bolt it supplys 500mah to downstream devices and also charges the d4 | 16:39 |
| parazyd | Great to make some sort of desktop dock too | 16:40 |
| uvos | yeah | 16:40 |
| parazyd | I haven't played much with the hdmi lately | 16:40 |
| parazyd | Maybe 2 years ago | 16:40 |
| parazyd | Then I managed to only duplicate output | 16:41 |
| buZz | i really want those official motorola desktop docks | 16:41 |
| parazyd | And had to keep same resolution | 16:41 |
| uvos | worked "fine" 4 months ago | 16:41 |
| buZz | they look so solid | 16:41 |
| uvos | just hildon freaks out | 16:41 |
| parazyd | Plus internal ts went blank | 16:41 |
| buZz | but 5usd item + 30usd shipping? nooo thanks | 16:41 |
| parazyd | Yeah, I know | 16:41 |
| uvos | you have to kill it | 16:41 |
| parazyd | But saying I didn't look why | 16:41 |
| uvos | then it restarts and works | 16:41 |
| uvos | parazyd: ok | 16:41 |
| parazyd | *nod* | 16:41 |
| uvos | buZz: its a made in china item | 16:42 |
| uvos | thats just german ebay | 16:42 |
| uvos | go find it elsewhere | 16:42 |
| buZz | uvos: that isnt the original motorola one, is it? | 16:42 |
| uvos | no ofc not i thought "5usd item + 30usd shipping" refered to the chinese hub | 16:43 |
| uvos | yeah the moto ones are expensive usually | 16:43 |
| uvos | the lapdocks even more so | 16:43 |
| uvos | i gues you could 3d print something for the chinese hub internals ;) | 16:44 |
| buZz | the moto ones are -cheap- | 16:44 |
| buZz | if you live in US :P | 16:44 |
| uvos | sure eyah | 16:44 |
| uvos | *yeah | 16:44 |
| uvos | in us you got them free for a while | 16:45 |
| uvos | (with the devices) | 16:45 |
| Wizzup | rafael2k: yeah so still some xdamage issues | 16:48 |
| sicelo | stano: pop3 asterisks ... if you were looking for a way to recover the password, it's in gconf | 16:59 |
| rafael2k | also, with the new updated, I "lost" the modem | 17:40 |
| rafael2k | ofono scripts and AT device file are not there also | 17:44 |
| Wizzup | uvos: well one of the boots up at least | 17:52 |
| Wizzup | the rest show leds, maybe they're charging | 17:52 |
| uvos | thats something :) | 17:54 |
| uvos | how is the physical apperiance? | 17:54 |
| rafael2k | # Userspace crypto wrapper cannot use aes-xts-plain64 (-95) | 17:57 |
| rafael2k | :( | 17:57 |
| rafael2k | /sys/class/gpio directory does not exist... and eg25 init script seems to need them to bring up the modem | 18:01 |
| rafael2k | # Misc devices | 18:29 |
| rafael2k | # CONFIG_MODEM_POWER is not set | 18:29 |
| rafael2k | somethings are missing in megis 5.14 kernel which was rolled to the repo | 18:29 |
| parazyd | ah thanks | 18:36 |
| parazyd | Yeah I ported our 5.10 config to 5.14 so missed some things it seems | 18:36 |
| rafael2k | thanks for bringing the megi's kernel to the repo! | 18:39 |
| rafael2k | ; ) | 18:39 |
| rafael2k | modem_power is something I like, easy to enable / disable / restart the modem | 18:39 |
| uvos | you really should never have a reason to do that outside of ofono | 18:45 |
| rafael2k | I don't think eg25 script will go away... | 18:46 |
| rafael2k | this is before ofono kicking in | 18:46 |
| rafael2k | ofono init depends on eg25 | 18:47 |
| uvos | sure | 18:47 |
| uvos | but userpsace messing with gpios is not a solution. | 18:47 |
| uvos | long term anyhow | 18:47 |
| uvos | (just guessing what it dose by what you wrote above i dont have a pp) | 18:48 |
| Wizzup | uvos: more are starting to work after charging | 18:48 |
| uvos | Wizzup: yay :) | 18:48 |
| uvos | Wizzup: so how is the cosmetic condition, they looked kinda rough on the ebay listing | 18:51 |
| uvos | also are all mz617 or did some other variants slip though | 18:51 |
| rafael2k | uvos: should be echo 1 > /sys/class/modem-power/modem-power/device/powered for start, and echo 0 > /sys/class/modem-power/modem-power/device/powered for stop() | 18:51 |
| uvos | sounds fine (ish) but why is runtime pm not used for this | 18:52 |
| uvos | oh well | 18:52 |
| uvos | /sys/class/gpio whatever it dose with that is bad and needs to be implmented in the driver | 18:52 |
| uvos | really | 18:52 |
| Wizzup | uvos: most looks good imho | 18:58 |
| Wizzup | look* | 18:58 |
| rafael2k | this is why megi wrote modem-power driver... so now you don't need to echo to "gpio" | 19:01 |
| rafael2k | the current eg25 in maemo will not work with recent megi's kernel, as now the specific driver does the lower level calls | 19:02 |
| rafael2k | now the userland just need to poke with /sys/class/modem-power/modem-power/device/powered | 19:03 |
| Wizzup | uvos: so far all seem to work/boot | 19:06 |
| uvos | rafael2k: ok that sounds ok ish but still the modem driver should just use the kernel pm framework, likely. | 19:08 |
| uvos | there might be a reason not to | 19:09 |
| uvos | i never even looked at the pp hardware | 19:09 |
| rafael2k | uvos: true, I dunno why to use a separate mechanism... anyway - it is indeed an easy access "kill switch" for telephony - and the reasoning behind this could be security - as it seems the default state of the modem in megi's kernel is disabled | 19:13 |
| parazyd | But there's also the dip switch | 19:15 |
| parazyd | If you want it off | 19:16 |
| parazyd | Otherwise, imho it should be on normally | 19:16 |
| parazyd | Except if in flight mode | 19:16 |
| rafael2k | this will never go upstream like this... | 19:16 |
| parazyd | Doesn't | 19:17 |
| parazyd | matter | 19:17 |
| rafael2k | yeap | 19:17 |
| rafael2k | I want to upgrade its firmware and I have it back | 19:17 |
| rafael2k | run this firmware: https://github.com/Biktorgj/pinephone_modem_sdk | 19:18 |
| rafael2k | (*when I have the modem back) | 19:19 |
| rafael2k | hi quality VoLTE calls are possible now | 19:19 |
| rafael2k | (with 48kHz audio mode) | 19:19 |
| rafael2k | also wanna test calls recording playing with pulseaudio / alsa | 19:20 |
| parazyd | Which reminds me | 19:20 |
| parazyd | We should have a dialer feature which lets you record calls | 19:20 |
| rafael2k | and at some point finally finalize my dialer using ofone | 19:20 |
| uvos | parazyd: sure | 19:21 |
| rafael2k | yeap, that is the idea | 19:21 |
| Wizzup | uvos: I think all work | 19:21 |
| uvos | Wizzup: nice :) | 19:21 |
| Wizzup | cool..l | 19:21 |
| uvos | parazyd: all devices we have support call recording | 19:21 |
| parazyd | Great | 19:21 |
| uvos | but we have to deal with ones who dont too | 19:21 |
| parazyd | I guess you could loopback pulse in a similar fashion like you can do with alsa | 19:22 |
| uvos | no | 19:22 |
| parazyd | oh? | 19:22 |
| rafael2k | modem has two audio routing options (at least) in the pp | 19:22 |
| uvos | like xt1602 you cant route the modem audio back to the cpu afaik | 19:22 |
| uvos | on any device where the audio isent handeld by the cpu at all | 19:23 |
| parazyd | ok | 19:23 |
| uvos | like mapphones, pinephone, xt1602 (i think most modern devices really) | 19:23 |
| uvos | you need hw support to get the audio | 19:23 |
| rafael2k | in the pp you can still route direct, not going throug the cpu | 19:23 |
| parazyd | I wonder how we can then do audio filtering on those | 19:24 |
| uvos | rafael2k: right thats what im talking about, not all devices support routing non direct | 19:24 |
| rafael2k | uvos: correct | 19:24 |
| rafael2k | parazyd: one of the audio routing options can route the audio through the cpu | 19:24 |
| uvos | parazyd: well in cellular calls the modem filters for us | 19:24 |
| uvos | in other calls we do it normaly | 19:24 |
| uvos | if you want to filter in some custom way | 19:25 |
| uvos | your sol on some deives | 19:25 |
| parazyd | Yeah I'm thinking custom filters as well | 19:25 |
| parazyd | Changing pitch or whatever | 19:25 |
| uvos | im not sure you can on mapphones | 19:25 |
| uvos | id have to look at the registers | 19:25 |
| uvos | you can get call audio to cpu | 19:26 |
| uvos | not sure you can get cpu audio to the call | 19:26 |
| parazyd | I see | 19:26 |
| parazyd | Yeah this might be possible only on non-modem calls then | 19:26 |
| uvos | certenly android dosent do this so acerning the registers needed might be non trivial, but not sure sre might have figured it out allready | 19:28 |
| sicelo | android call recording applications don't work on droid 4 running android? | 19:33 |
| uvos | sicelo: ? | 19:33 |
| uvos | sicelo: yes they do | 19:33 |
| sicelo | xt1602? | 19:34 |
| uvos | oh yeah | 19:34 |
| uvos | there they dont | 19:34 |
| sicelo | ok | 19:34 |
| uvos | thats what im basing the "like xt1602 you cant route the modem audio back to the cpu afaik" | 19:34 |
| uvos | on | 19:34 |
| uvos | its possibe it really can in hw | 19:34 |
| tmlind | Wizzup, uvos: looks like the tc358775 (with a linux driver) is pin and register compatible with earlier tc358765, no luck yet with display output on mz617 though, good to hear the cheapo devices are booting :) | 20:50 |
| uvos | tmlind: good to know, dosent sound like to should be tooo hard, we are likely just missing some gpio to power up the pannel regulator or something like that. or maybe EXTCLK on the tc gated | 21:03 |
| tmlind | uvos: yeah so tc358775 has just the extra stby gpio compared to tc358765 it seems, i think the remaining issues are in omapdrm drive | 21:08 |
| tmlind | r | 21:09 |
| uvos | tmlind: o | 21:09 |
| uvos | tmlind: ok | 21:09 |
| tmlind | uvos: the panel backlight seems to work with the lp8550 driver | 21:09 |
| uvos | tmlind: ok | 21:10 |
| uvos | thats sounds more nasty then :P, are you suspecting somethings wrong with the dsi input fro the tc? | 21:10 |
| tmlind | so should be fairly close except for some mystery bugs | 21:10 |
| uvos | ok | 21:10 |
| uvos | *for | 21:10 |
| uvos | dsi is pretty easy to decode with a logic analyzer | 21:11 |
| tmlind | uvos: so i added panel-simple support too but getting; Cannot find any crtc or sizes | 21:11 |
| uvos | maybe i can dump android and mainline output | 21:11 |
| uvos | do you have some high res photos of the board? | 21:11 |
| uvos | (if you think it might be usefull) | 21:11 |
| tmlind | no | 21:11 |
| uvos | ok | 21:12 |
| tmlind | i think it's just some bug in omapdrm remaining somewhere | 21:12 |
| uvos | ok | 21:12 |
| tmlind | it's like the enable functions for bridge or display never get called | 21:12 |
| uvos | oh ok | 21:13 |
| uvos | if you have anything specific you want me to investigate, ill check in whenever i have one :) | 21:13 |
| tmlind | uvos: so i just set up motorola-mapphone-tablet-common.dtsi as most peripherals are different | 21:13 |
| uvos | tmlind: ok sounds sane | 21:14 |
| uvos | just shares the cpcap dtsi then i assume | 21:14 |
| tmlind | uvos: i'll send you my hacks when you get one, still nothing worth posting it seems | 21:14 |
| tmlind | uvos: yeah cpcap seems to work for charging and usb, have not tried audio | 21:15 |
| tmlind | uvos: also modem gpio pin wirign is slightly different, but that's about it i think | 21:16 |
| uvos | i mean that makes sense its a diff modem after all | 21:16 |
| tmlind | yeah | 21:17 |
| tmlind | there's an existing bridge driver for am57xx-idk-common.dtsi | 21:18 |
| tmlind | that's parallel to dsi panel, so a bit different but i guess it works | 21:18 |
| tmlind | oh and wireless is still flakey and won't come up automatically for some reason | 21:19 |
| tmlind | restarting it manually works fine though | 21:19 |
| uvos | same chip? | 21:20 |
| tmlind | yeah | 21:20 |
| uvos | (as mapphones) | 21:20 |
| uvos | ok | 21:20 |
| uvos | wierd | 21:20 |
| tmlind | yeah, i calibrated it too with your scripts | 21:20 |
| uvos | tmlind: needs different fem parameters maybe | 21:20 |
| uvos | the pa might be different | 21:20 |
| uvos | or ant is in different place | 21:20 |
| uvos | etc | 21:21 |
| tmlind | yeah | 21:22 |
| inky | very hard to use pinephone (without hw kbd) with leste: have no backspace key, cannot delete something that is already written | 21:22 |
| tmlind | but it basically boots with droid4 dts more or less | 21:22 |
| uvos | inky: yeah backspace cant work | 21:22 |
| uvos | inky: because of the architecture of him | 21:23 |
| uvos | inky: i hope i can change this at some point | 21:23 |
| tmlind | oh is the musb still not working for pinephone? | 21:23 |
| inky | good that we have a volume up tdat is able to bring keyboard. | 21:23 |
| uvos | tmlind: no he is complaining about the virtual keyboard | 21:23 |
| tmlind | oh right | 21:23 |
| inky | i'll try to connect bh kbd | 21:24 |
| inky | bt | 21:24 |
| tmlind | the m-l virtual kb sucks big time still :) | 21:24 |
| uvos | yeah | 21:24 |
| uvos | you will get no argument from me | 21:24 |
| tmlind | have you noticed how the cursor moves back to the previous key briefly after typing a new key? :) | 21:25 |
| uvos | no not sure what you mean | 21:25 |
| uvos | oh you mean the grey preview? | 21:26 |
| tmlind | so if you look at the virtual kb cursor while typing, it blinks between the last new character and the previous character briefly | 21:26 |
| uvos | thats on purpose | 21:26 |
| tmlind | eh | 21:26 |
| uvos | its whats currently pressed | 21:26 |
| uvos | so you can correct it by sliding your finger to the right key | 21:26 |
| uvos | if you mis | 21:26 |
| tmlind | it's a feature :) | 21:26 |
| uvos | it really is semi usefull | 21:27 |
| uvos | like if you want to press i but land on o | 21:27 |
| uvos | it will show o in gray but you can slide up to i and then release and get an i | 21:27 |
| uvos | and the grey key gives you feedback | 21:27 |
| tmlind | heh, i made a script called upd to do apt-get dist-upgraded to avoid typing more characters on bionic.. i guess i should rename it to u :) | 21:27 |
| uvos | *grey letter | 21:27 |
| uvos | :P | 21:28 |
| tmlind | hmm i guess i should try the sliding though | 21:29 |
| inky | uvos this is great, i can use armenian kbd in debian's pidgin without doing setxkbmap by hand | 21:29 |
| inky | what did you do? | 21:29 |
| uvos | um | 21:29 |
| uvos | nothing | 21:29 |
| inky | (but i cannot pcess enter to send the message) | 21:29 |
| uvos | inky: enter sometimes dosent work the first time | 21:30 |
| uvos | not sure why | 21:30 |
| inky | it writes in ghk app from debian | 21:30 |
| uvos | open the vkb again press enter a few times | 21:30 |
| uvos | and close it again | 21:30 |
| inky | i was doing setxkbmap by hand to make it write | 21:30 |
| inky | yes, it wocked necond time with surf, but doesn'h work with pidgin. :/ | 21:32 |
| inky | worked, second | 21:32 |
| uvos | ok no idea than atm | 21:32 |
| uvos | *then | 21:32 |
| inky | surf is in maemo repos, may be adapted for hildon? | 21:32 |
| inky | and pidgin is from debian. | 21:33 |
| inky | what do you think i can modprobe to test bluetooth? | 21:42 |
| sicelo | no backspace on him vkb? :-/ | 22:06 |
| uvos | the way him works fundamentaly precludes a backspace key that works beyond the text you just enterd | 22:08 |
| sicelo | ah, so not that backspace isn't there (i was getting surprised!), but text editing is not 'normal' | 22:09 |
| uvos | right | 22:09 |
| sicelo | the cursor grey/movement is also used when there's word completion (i don't think this works on Leste yet, or does it now?) | 22:12 |
| uvos | no | 22:12 |
| sicelo | Wizzup: sorry, can't test Leste + N900 for a while | 22:14 |
| inky | so osso mahjong was open source | 22:30 |
| inky | does anyone know if osso tetris was open source? | 22:30 |
| Wizzup | ok yeah all 11 boot to android desktop | 22:41 |
| uvos | Wizzup: great | 22:42 |
| uvos | are all mz617? | 22:42 |
| uvos | or are some wifi variant etc | 22:42 |
| Wizzup | I haven't done much beyond charging and starting them | 22:44 |
| Wizzup | if you have some instructions I can do that in a bit | 22:45 |
| uvos | Wizzup: its on the back of the device | 22:45 |
| uvos | MODEL: blabla | 22:45 |
| Wizzup | all are mz617-16 | 22:47 |
| uvos | Wizzup: ok | 22:47 |
| uvos | to bad | 22:47 |
| Wizzup | remind me? | 22:49 |
| uvos | mz617 is the us wifi variant there is also mz615 (euro version) mz614(euro wifi) mz613? (us wifi) etc | 22:50 |
| uvos | if we had some that where not mz617 we could claim support for these other variants | 22:50 |
| uvos | but its not a big deal | 22:50 |
| Wizzup | I ordered from the us, so to be expected | 22:51 |
| uvos | well you could have goten a wifi one ;) | 22:51 |
| uvos | i gues they wher not popular | 22:51 |
| stano | Desktop Command Execution Widget seems to run the command constantly? http://maemo.org/downloads/product/Maemo5/desktop-cmd-exec/ | 23:34 |
| stano | what if i want a widget that just executes when tapped | 23:34 |
| sicelo | doesn't run constantly, unless you set it to. execution config can be controlled by either long-pressing or tapping the desktop then clicking the cog wheel, then pressing the setting button on the widget | 23:36 |
| stano | after install Desktop Command Execution Wiget should appear under Desktop menu -> Add shortcut? | 23:39 |
| sicelo | yes. you may have to reboot or restart hildon desktop, as there's some sort of bug atm, which causes new widgets to not be detected first time | 23:40 |
| stano | /usr/sbin/dsmetool -k /usr/bin/hildon-desktop and restarting didn't do it | 23:42 |
| stano | i want four widgets to set screen brightness | 23:43 |
| sicelo | reboot, and yes, dcew should be able to do what you want | 23:43 |
| stano | reboot hangs at [ 10.989...] asoc-audio-graph-card soundcard: No GPIO consumer pa found | 23:45 |
| stano | good thing i have that torx screwdriver always handy | 23:48 |
| sicelo | vol button + power works all the time | 23:49 |
| stano | if your vol button works :) | 23:49 |
| sicelo | right | 23:49 |
| stano | now device can't boot past kexec. reboot loop | 23:51 |
| stano | pulling battery again got me to booting | 23:54 |
| parazyd | ok, I see a few things are gone in the devel kernel on pinephone. Will sort it out. | 23:56 |
| parazyd | Supposedly battery life should be better too | 23:56 |
| stano | Desktop now has "Add widget" menu item | 23:58 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
