| sicelo | Wizzup: yay, https://paste.debian.net/1323104/ | 00:23 |
|---|---|---|
| sicelo | quick and dirty ... will cleanup tomorrow. midnight now | 00:23 |
| sicelo | side effect - this reveals that the mount matrix i calculated for N900 is wrong for the Z axis :p | 00:24 |
| sicelo | oh, i should mention that i didn't (yet) solve the issue shown at https://paste.debian.net/1323081/ ... at least on the Droid 4. my current test/patch was done from N900 running postmarketOS. that corruption of values doesn't happen there (pmOS) | 00:47 |
| sicelo | will check why it happens under Leste. the patch itself should apply with no change at all under Leste, since the file it touches has not been modified since iio-sp 3.0 (version in Leste) | 00:49 |
| sicelo | ah, seems to only be mangled by gdb ... weird. looks to be working fine when run outside gdb | 08:03 |
| sicelo | uvos: https://github.com/maemo-leste/mce/blob/master/src/modules/iio-accelerometer.c#L1 ... what should we use instead? | 08:06 |
| uvos | iio-s-p directly, the module is 1. just proxying iio-s-p with no value add and 2. dosent fit into mce because this is the only external interface to a sensor of any kind mce has so its wierd that mce plays hal for just this one sensor | 08:44 |
| sicelo | understood | 08:51 |
| sicelo | btw how to do more verbose logging for mce | 08:51 |
| uvos | -v -v | 08:51 |
| sicelo | ty | 08:52 |
| sicelo | Wizzup: it works, https://paste.debian.net/1323147/ | 09:38 |
| sicelo | alarm silences perfectly with face down now ... but not sphone (yet) | 09:59 |
| uvos | we really need a way to disable this now | 10:01 |
| sicelo | is it really a big problem? i've never had it interfere unintentionally with calls/alarms under Fremantle or even Android | 10:21 |
| Wizzup | sicelo: sweet :) | 10:23 |
| uvos | Its a massive problem to me | 10:30 |
| Wizzup | like every fremantle default feature :DF | 10:32 |
| Wizzup | :D | 10:32 |
| uvos | if you use the alarm just as a alarm its ok, but if you use it in a position where the alarm is likely to fire while you are walking/working out/moving in general it makes it liable to silence the alarm and those kinds of alarms tend to be most curital (ie medication timers) | 10:32 |
| Wizzup | well it only fires when the position changes rightg | 10:33 |
| Wizzup | it doesn't just silence the alarm immediately when the phone is (already) face down | 10:33 |
| uvos | and yes that happens on occasion | 10:33 |
| Wizzup | I haven't had that on my n900 with fremantle really tbh | 10:33 |
| uvos | and once is enough to be dangerous | 10:33 |
| Wizzup | arno11: https://news.ycombinator.com/item?id=40951472 | 10:39 |
| Wizzup | not sure if we'd want it, big picture doesn't make sense for us :D | 10:39 |
| arno11 | Wizzup: (for big picture) yes indeed :D | 12:23 |
| Wizzup | it also seems to be ps2 I think, so maybe we can't do that anyway | 12:24 |
| arno11 | yes it is ps2 | 12:24 |
| arno11 | awesome emulator btw | 12:25 |
| arno11 | but nightmare to run and configure, even on a good desktop machine | 12:25 |
| Wizzup | fair enough | 12:26 |
| Wizzup | :D | 12:26 |
| Wizzup | I need go head out for a bit, I will be back later today, can I help with the drnoksnes stuff? | 12:27 |
| arno11 | it already works :D | 12:27 |
| Wizzup | oooh | 12:27 |
| Wizzup | I will test it on the d4 then | 12:27 |
| arno11 | cool | 12:27 |
| arno11 | hope there is no issue in 24bpp | 12:28 |
| arno11 | keyboard scancodes could be an issue on D4 but easy to fix from config file | 12:29 |
| Wizzup | I was thinking about the different res too | 12:35 |
| arno11 | not sure what will happen, but should work using default armx2 in fullscreen | 12:38 |
| arno11 | on N900 it is already small in window mode | 12:39 |
| arno11 | *armx2 scaling in 4:3 | 12:39 |
| arno11 | if it is too small in fullscreen on D4 then we will need sdlhaa | 12:41 |
| Wizzup | uvos: how to force turn off my mz617 | 12:43 |
| Wizzup | it's stuck in some dmesg not starting X and I cannot figure out how to reset it | 12:43 |
| arno11 | Wizzup: btw i opened an issue for swap stuff with 6.6 | 12:47 |
| Wizzup | ok | 12:49 |
| Wizzup | uvos: vol + and power or vol - and power do not work | 12:49 |
| Wizzup | now the screen is just on and I cna't find my serial :( | 12:49 |
| Wizzup | that's so stupid | 12:49 |
| Wizzup | vol - worked somehow | 12:59 |
| uvos | mz617 has the same reset claw as d4 | 13:17 |
| arno11 | btw sd card swap (u3) seems to work far better on N900 for multitask: with irssi, calendar, conversations, hamsterfiler, audacious, htop, HAM and maep opened, still able to play picodrive @60fps @600MHz max :D | 13:46 |
| arno11 | 6.6 seems really stable | 13:53 |
| Wizzup | uvos: yeah somehow it worked after the 10th time | 14:55 |
| Wizzup | maybe some contact issue | 14:55 |
| Wizzup | arno11 you said you filed an issue for swap, where? | 14:56 |
| uvos | https://github.com/maemo-leste/bugtracker/issues/741 | 14:57 |
| uvos | obvious place | 14:57 |
| Wizzup | strange, because I didn't get it in my email | 14:58 |
| Wizzup | normally I get an email for every new issue an dcomment | 14:58 |
| Wizzup | arno11: do you have a desktop icon doe drnoksnes? | 15:00 |
| Wizzup | I don't see it | 15:01 |
| Wizzup | s/doe/for/ | 15:02 |
| Wizzup | freemangordon: btw re: your haze merge request, we have contact now with the guy so let's try to use it as well at some point :) | 15:22 |
| arno11 | Wizzup: yes there is an icon but remember, you have to use command line to launch a game | 16:23 |
| arno11 | /usr/bin/drnoksnes -c /usr/games/drnoksnes.conf /game_path | 16:24 |
| arno11 | if scancodes are ok: space for start, a s z x for buttons, f for fullscreen (by default) | 16:29 |
| arno11 | e for exit | 16:30 |
| arno11 | if scancodes are different on D4: the console shows the corresponding scancode when you press a button. we have just to add them in the conf file | 16:34 |
| arno11 | for the icon: on n900 it appears immediately after installation (there is postinstall stuff to refresh icon cache) | 16:40 |
| arno11 | i should add a wiki btw | 16:44 |
| gnarface | hey arno11 just out of curiosity what does the output of "dmesg |grep mmc" say? | 16:44 |
| gnarface | (listed speed ratings might explain why you're getting better results with swap on microsd) | 16:44 |
| arno11 | i already try but didn't see any relevant stuff unfortunately :( | 16:46 |
| arno11 | for better results: ok | 16:46 |
| gnarface | hmm, i was sure it'd say some speed ratings | 16:46 |
| gnarface | for each mmc bus | 16:46 |
| gnarface | it might not actually look like a speed rating | 16:46 |
| gnarface | they have super generic looking names | 16:47 |
| arno11 | ah ok | 16:47 |
| gnarface | it might say something like "high speed" which just looks like descriptive fluff but is actually an official rating (25MB/s) | 16:48 |
| gnarface | whereas "default speed" is 12.5MB/s | 16:48 |
| gnarface | UHS, UHS-II, etc are higher | 16:48 |
| gnarface | i don't know if it'd actually say "UHS" or say "ultra high speed" | 16:49 |
| gnarface | but it might give you some ideas of where your physical bus limits are | 16:49 |
| gnarface | afaik it won't actually say a number though | 16:49 |
| gnarface | you have to look up a chart | 16:50 |
| gnarface | i found this, dunno if it's an exhaustive list or not: https://www.sdcard.org/developers/sd-standard-overview/bus-speed-default-speed-high-speed-uhs-sd-express/ | 16:50 |
| arno11 | ok, it says high speed on my device btw | 16:50 |
| gnarface | ok, both of them? | 16:50 |
| gnarface | i assume the emmc and the microsd each get their own bus | 16:51 |
| gnarface | so that means if you're already saturating 25MB/s on the emmc with other traffic then it would make sense that moving swap to the microSD would give better performance | 16:51 |
| gnarface | also, from what i've heard, this is no guarantee that the hardware can't do faster, it's just what the .dts is set at | 16:52 |
| gnarface | (i don't know N900 specifics though other than that it's old, so maybe high speed is the most that can be expected out of it) | 16:52 |
| arno11 | ok (and yes both emmc and sd are 'high speed' in dmesg) | 16:53 |
| gnarface | that also would mean that there's little point in buying microSD cards rated at faster speeds than that; you can opt for stuff rated for high durability instead | 16:55 |
| arno11 | yes absolutely | 16:56 |
| arno11 | but usually the build quality of vry fast cards makes the diff | 16:56 |
| arno11 | *very | 16:56 |
| gnarface | sure, i just know that there's specific brands/models that are designed to handle a higher amount of total write cycles before wearing out, and those are usually also not as fast as the top-shelf stuff | 16:58 |
| arno11 | ok | 16:58 |
| gnarface | i'm thinking of some specific white-label Samsung ones, but i know there's others | 16:58 |
| gnarface | stuff for industrial purposes and such | 16:59 |
| arno11 | ah yes ok | 16:59 |
| arno11 | btw samsung evo plus work very well on N900, better than a faster (theorically) sandisk i have | 17:00 |
| gnarface | interesting | 17:01 |
| arno11 | yes indeed | 17:01 |
| Wizzup | arno11: I don't see an icon, and why does the icon not have that command line? :D | 17:13 |
| uvos | the emmc bus can do 50MB/s because its 8bits wide | 17:13 |
| Wizzup | maybe I am just blind | 17:13 |
| uvos | but the maximum speed is totally irrelivant | 17:13 |
| uvos | as no sdcard or emmc chip can do 25MB/s very small random reads | 17:14 |
| sicelo | Wizzup: tion_t orit) | 17:14 |
| sicelo | { | 17:14 |
| uvos | but very small random reads is all that matters for swap | 17:14 |
| sicelo | ugh | 17:14 |
| arno11 | Wizzup: i can't add the command line to the icon because you have to specify the game you want to use | 17:15 |
| uvos | modern flash chips are mutch much much much better at this than 2000s era ones | 17:15 |
| sicelo | Wizzup: https://github.com/maemo-leste-upstream-forks/iio-sensor-proxy/pull/2 , https://github.com/maemo-leste/mce/pull/61 ... maybe you can test when you have a moment ;-) | 17:15 |
| arno11 | Wizzup: btw i made a yad gui for drnoksnes linked to the icon and it is not too bad | 17:17 |
| uvos | sicelo: did you fix the nan issue? | 17:17 |
| sicelo | which one? | 17:17 |
| uvos | you repoted a nan issue on d4 with this patch | 17:17 |
| sicelo | ah, that was just gdb, as weird as it sounds | 17:18 |
| sicelo | and that was without any patching at all | 17:18 |
| uvos | ok | 17:20 |
| sicelo | btw, what is now the correct kernel branch to submit patches against? | 17:24 |
| uvos | 6.6.y | 17:25 |
| sicelo | want to submit fix for N900 z-axis. tyvm | 17:25 |
| sicelo | will also upstream it, of course | 17:25 |
| arno11 | Wizzup: i have to go but let me know for drnoksnes | 17:25 |
| Wizzup | arno11: huh, drnoksnes comes with a gtk gui | 17:32 |
| arno11 | Wizzup: ?? you mean THE gui or a kind of generic one ? | 18:32 |
| arno11 | gui was supposed to be removed but indeed it seems still there lol but returns maemo-invoker error on N900. Maybe not on your D4 ? interesting anyway | 18:46 |
| arno11 | curious to see if it works | 18:47 |
| sicelo | uvos: thanks, i see mce PR is merged | 18:56 |
| sicelo | please also take https://github.com/maemo-leste/mce/pull/62 ... otherwise on face up, for example, device will force itself to landscape. apologies for that regression | 18:57 |
| sicelo | and then, of course, this stuff depends on the iio-s-p patch to actually work | 18:57 |
| Wizzup | arno11: why remove the gtk gui was my question? | 18:58 |
| arno11 | because of buggy deps but i finally removed them | 19:00 |
| arno11 | is it working or ? | 19:00 |
| arno11 | info please :D | 19:00 |
| Wizzup | sorry, I am getting pulled away from the computer | 19:03 |
| Wizzup | it looks like I need to take a bit more time to test this | 19:03 |
| Wizzup | I have only ever used the drnoksnes gui on fremantle | 19:03 |
| arno11 | ok no probs | 19:03 |
| Wizzup | when you mean the buggy deps, do you mean the hildon game libs? | 19:03 |
| arno11 | i meant xsp stuff, sdlhaa which was finally not needed/related | 19:04 |
| arno11 | (understanding what android-808 and clort did was a bit complicated) | 19:05 |
| arno11 | hildon libs are fine iirc | 19:06 |
| uvos | sicelo: ok, i need a bit more time to look at the iio-s-p changes | 19:13 |
| sicelo | sure, no rush | 19:13 |
| uvos | its a bit hard to understand how these changes fit in when the pr contains only a patch | 19:13 |
| uvos | so i need to apply it and look at it first | 19:14 |
| uvos | (not your fault ofc) | 19:14 |
| sicelo | i'll upstream the iio-s-p changes. initial discussions with them seemed positive | 19:15 |
| arno11 | Wizzup: lol, looking @logs, big misunderstanding @gui, sorry | 19:20 |
| sicelo | < uvos> its a bit hard to understand how these changes fit in when the pr contains only a patch ___ ah yes, i get what you mean. the patch workflow sucks, but yeah, it is what it is | 19:21 |
| arno11 | Wizzup: got the gui working locally :D | 21:31 |
| arno11 | it works amazingly well: the issue was 2 different config files (gui and emulator) with the same name and same path...so it was not possible to run both | 21:42 |
| arno11 | keeping the same name but moving the gui conf file elsewhere with the correct path in desktop file solved the issue | 21:43 |
| arno11 | so i need to make a new release with just 2 small changes and the emulator will work like in fremantle :D | 21:46 |
| arno11 | (at least on N900 atm) | 21:46 |
| Wizzup | arno11: sweet!\ | 21:53 |
| arno11 | yes that's really cool ! | 21:56 |
| sicelo | nice progress arno11 ;-) | 22:15 |
| arno11 | ty :) | 22:19 |
| arno11 | i always forget to thank you all guys, being always happy to help and available to explain things to a non dev guy like me. | 22:38 |
| Wizzup | well, you contribute a lot too :) | 22:49 |
| arno11 | yes i try (at least) :) | 23:05 |
| Wizzup | latest 6.6 also works on mz617 | 23:15 |
| arno11 | ah cool | 23:17 |
| Wizzup | the race in the ts also seemingly is gone, at least 2/2 tries have been ok | 23:17 |
| sicelo | uvos: Wizzup: https://github.com/maemo-leste/droid4-linux/pull/11 | 23:31 |
| Wizzup | ty merged | 23:39 |
| Wizzup | will you send it upstream too? | 23:39 |
| Wizzup | maybe also for fixes | 23:39 |
| sicelo | yes, will upstream :-) | 23:40 |
| sicelo | the one for iio-s-p already submitted, https://gitlab.freedesktop.org/hadess/iio-sensor-proxy/-/merge_requests/384 | 23:41 |
| Wizzup | I think we have forked this so we can pull this in now too | 23:41 |
| sicelo | yeah. uvos wanted to have a look first, | 23:44 |
| sicelo | Commits 1 | 23:44 |
| sicelo | Checks 0 | 23:44 |
| sicelo | https://github.com/maemo-leste-upstream-forks/iio-sensor-proxy/pull/2 | 23:44 |
| sicelo | wrong pastes ... sorry | 23:44 |
| Wizzup | ok, I'll leave that one for him | 23:48 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!