Monday, 2021-07-05

T42<萧新婷> Forwarded from 邰明婺: Estimated launch date: July 5, 2021  UTC 12:1013:01
T42<萧新婷> 2021 The most anticipated defi god, there is no one!13:01
T42<萧新婷> Free Britney/BNB13:01
T42<萧新婷> pancakeswap list13:01
T42<萧新婷> Technical team: 0 chips 0 reserved13:01
T42<萧新婷> https://t.me/Britneylegends13:01
T42<萧新婷> Contract address: 0x1A03ac27087b8622E70AD74D864469f47d634CdC13:01
T42<elros34> @b100dian you could be right about that notification bug, could be related to qml. I found something related to opacity and opacityanimator but I do not know why this cause issue. Please test patch, if it works for you then I will upload it to pm2 catalog. https://pastebin.com/sCyutFXd13:23
Thaodanspam from the telegram side again13:24
T42<b100dian> Cool! And I have given up chasing notifcationmanager.cpp for dbus interactions yesterday evening:) Testing as soon as I have ssh13:27
T42<b100dian> very weird, restart lipstick takes me now to crazy ppoll strace and 20-30% cpu, taking a screenshot and waiting for the notification to dismiss makes it plumment back to ~0% @elros34.14:05
T42<elros34> ouh.. could be some another issue too, so best would be to test after restarting device14:08
T42<b100dian> Same after restart. Let me clear the other notifications and test again14:10
T42<b100dian> Same result. So its like you have to make a screenshot now to get back to normal cpu usage14:34
T42<b100dian> Ill unapply the patch and test again..14:34
T42<elros34> you can also remove that animator, that should be fine too14:34
T42<b100dian> Btw, patchmanager needs a new maintainer.. Even the web site needs 4.1.0.24 added at least14:35
T42<elros34> I still use slightly modified v2 version and that one works14:36
T42<b100dian> Built it yourself, right? Did you need to manually enable apply on boot?14:37
T42<elros34> pm2 doesn't have such an option like apply on boot because it change files not pretend to change them like pm314:38
T42<b100dian> oh, I see, that is pm3 that I’ve tried, thanks for pointing this out14:43
T42<b100dian> removed that line altogether, restart no major cpu usage, screenshot no major cpu usage after auto dismissal. I think you’re on to something @elros34!14:45
T42<elros34> ok thanks, let more advanced jolla/omp devs find proper fix now14:48
T42<b100dian> did a test without the patch again and with that line removed, I can re-confirm the results14:49
rinigusI am currently testing a small governor that switches off power-hungry CPUs (big) while screen is off and changes cpufreq policy to conservative on LITTLE CPUs.17:14
riniguson switching screen back on, interactive cpufreq is restored on both policy sets as well as power-CPUs are tunded back on17:14
rinigusif interested in testing, source at https://github.com/rinigus/zgovernor17:15
rinigusidea is to reduce battery consumption in idle and it seems to be working. although, hard to put it into numbers yet17:15
rinigusfor testing, you may have to modify .ini file to take into account your device cpus and how the governors are configured. that setup is based on tama with sdm84517:17
Thaodanrinigus: maybe this could be a part of mce as a plugin18:26
Thaodanor dsme18:27
Thaodanhttps://github.com/sailfishos/dsme/blob/master/rpm/dsme.spce18:27
rinigusThaodan: maybe it should. let's see if it works in practice and then discuss it.18:27
rinigusmce I checked out, dsme - didn't think about it. (there is a typo in url above)18:28
rinigushttps://github.com/sailfishos/dsme18:28
Thaodaneh yeah thx18:28
Thaodanspiiroin: would be a good person to talk to I think18:29
rinigusI am actually confused in dsme/mce difference, probably missed some docs explaining their more specific aims18:30
Thaodanhttps://sailfishos.org/wiki/Architecture18:32
ThaodanThere is something here but not much18:32
ThaodanIts quite old Nokia software similar to ohm18:37
ThaodanIn Meego it was listed under "Device Health"18:40
Thaodanhttp://www.subdude-site.com/WebPages_Local/RefInfo/Computer/Linux/LinuxGuidesOfOthers/linuxProgrammingGuides/pdfs/MobileDeviceProgramming/MOBmee_Meego_whereAreWe_whatNext_Intel_2010jul_slides4linuxFoundatnTalk_31pgs.pdf18:40
rinigusI'd call it probably "mature" rather than "old" :)18:41
ThaodanThat's kinda better - I would say it is a bit hidden and not as documentd18:41
ThaodanI found even more docs on scratchbox2 compared to this18:41
rinigusThaodan: true, it would make sense to merge such functionality with the available HW monitoring/state machines. it is just easier to prototype separately and see if it makes sense to introduce such functionality.18:45
piggzlbt: hi, any closer to 4.1?19:56
lbtno - I've been swamped, sorry23:45

Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!