*** Daaanct12 is now known as Daanct12 | 04:51 | |
T42 | <adampigg> Rinigus: sure, im away today, feel free to ship | 09:29 |
---|---|---|
rinigus | @adampigg: will do | 09:29 |
*** Mister_Magister_ is now known as Mister_Magister | 10:05 | |
mal | PSA: anyone building 5.0 adaptations in testing projects on OBS please add sailfishos_5.0 target to your adaptation project, I switched the repos to use major.minor versioning to remove the need to have separate target for each small release | 13:14 |
mal | an example https://build.sailfishos.org/project/show/nemo:testing:hw:fairphone:fp5 | 13:14 |
mal | so next time someone updates their devices from testing project the adaptation and common repo will change to use that 5.0 target | 13:15 |
Mister_Magister | mal: what about my mce PR ;-; | 13:22 |
Mister_Magister | time to abuse obs | 13:23 |
* Mister_Magister opening literally 6 projects | 13:24 | |
Mister_Magister | also mal what about 3.4 fix? | 13:25 |
Mister_Magister | its been months | 13:31 |
Mister_Magister | Might just build the fix you didn't like cause i'd like not to be stuck at 4.6 :P | 13:32 |
mal | well the fix didn't work on all devices which is the problem | 13:33 |
mal | or if it works on any 3.4 devices | 13:34 |
Mister_Magister | I see | 13:34 |
Mister_Magister | well I believe in you <3 | 13:34 |
Mister_Magister | I've tested the mce PR on two phones and it works great | 13:35 |
Mister_Magister | it could use some debounce but other than that its great | 13:35 |
mal | the brightness value changes are something I wonder | 13:49 |
Mister_Magister | what do you mean? | 13:50 |
mal | isn't new_hbm_level now 1? | 13:50 |
Mister_Magister | could you elaborate on what do you mean? out of context i don't know what you mean | 13:50 |
mal | https://github.com/sailfishos/mce/pull/41/files#diff-c54a5a546ff09891bf77db63660f655be34e29666b9d646bbd82838d455f538aR3578 | 13:51 |
Mister_Magister | new_hbm_level is 1 when brightness is 100 yes | 13:51 |
Mister_Magister | and if its less than maximum brightness its 0 | 13:51 |
mal | but before it was something else | 13:51 |
Mister_Magister | before it was always 0 | 13:51 |
Mister_Magister | which is explained in description | 13:52 |
Mister_Magister | second paragraph | 13:52 |
Mister_Magister | before code made it so that it will never be anything but 0 | 13:53 |
Mister_Magister | I've debugged it through and through i can explain you everything to tiniest detail if need be :D | 13:57 |
mal | some comments: there is indentation mismatch in line with "MCE_CONF_HBM_CONTROL_PATH, 0);", the comment before that part is for next code block so the new code is in wrong place, I wonder if where the new code for config loading should be because now there is tiny chance of memory leak if that hbm path is configured and it ends up in the condition "ss(DISPLAY_BACKLIGHT_PATH DISPLAY_DISPLAY0, W_OK) | 14:18 |
mal | == 0)" | 14:18 |
mal | one option would be to have it before display_type_get_from_config call in mdy_display_type_get and adding a null check to that DISPLAY_DISPLAY0 case | 14:20 |
Mister_Magister | mal: yeah I wasn't sure where to put loading of the config | 14:22 |
mal | or it could be near the end of mdy_display_type_get with maybe check that display_type != DISPLAY_TYPE_NONE | 14:23 |
Mister_Magister | alright i'll see about that | 14:23 |
rinigus | mal: regarding your PSA - I guess this is the corresponding commit: https://github.com/mlehtima/droid-config-fp5/commit/8740bf43f9325d8fb9eadfd236a3b32bd8062af0 | 14:23 |
mal | rinigus: that is special for fp4 and fp5, those have extra repo hosted on my server | 14:24 |
mal | no need to do anything to droid-config for normal devices | 14:24 |
mal | unless you have some manually added custom repo | 14:24 |
mal | rinigus: this is the change for the normal repos https://github.com/mer-hybris/community-adaptation/pull/6 | 14:25 |
rinigus | thanks! I do have custom repos for tama, will have to look into it | 14:26 |
mal | can you point me to those? | 14:26 |
rinigus | but this change makes sense - no need to race for each small change | 14:26 |
mal | yeah, we got annoyed with the amount of targets on community obs, both in common and even more in chum | 14:27 |
mal | those chum builds take a lot of space from the server | 14:27 |
mal | and also makes chum rebuilds much slower | 14:28 |
rinigus | mal: having huge chum is, in principle, sign that it works. but trimming it to MajorMinor does make sense | 14:30 |
mal | Mister_Magister: if you put it to the end make sure you free the old hbm path | 14:30 |
Mister_Magister | thank | 14:30 |
mal | assuming it's set, only that one case sets it | 14:30 |
Mister_Magister | I think setting it at the top makes most sense | 14:31 |
mal | well the it would try to load the config even if there is no display, doesn't matter much | 14:32 |
Mister_Magister | yeah we don't really need display to load config | 14:32 |
mal | but the config would be useless anyway | 14:33 |
Mister_Magister | but it causes no harm :P | 14:42 |
Mister_Magister | mal: https://paste.opensuse.org/c77b8e7087ba is that okay? | 23:12 |
*** Mister_Magister_ is now known as Mister_Magister | 23:31 | |
mal | seems reasonable | 23:31 |
Mister_Magister | oki | 23:33 |
mal | Mister_Magister: a small style comment, check the formatting of if statements in other parts, your code doesn't match | 23:33 |
mal | looks like you have "if (something)" when code in mce uses "if( something )" | 23:33 |
Mister_Magister | oh because you guys waste the bytes on useless spaces | 23:34 |
Mister_Magister | yeah | 23:34 |
Mister_Magister | the wrong kind of formatting | 23:34 |
Mister_Magister | i'll chaaaaange iiiit | 23:34 |
mal | it's old nokia legacy and we just keep using it in that project so code is consistent | 23:34 |
Mister_Magister | i know i know | 23:35 |
mal | I also think that code style is very odd | 23:35 |
Mister_Magister | mal: the ifs literally above don't follow it either xd | 23:35 |
mal | it seems some do and some don't | 23:36 |
Mister_Magister | https://github.com/sailfishos/mce/pull/41/files#diff-c54a5a546ff09891bf77db63660f655be34e29666b9d646bbd82838d455f538aR5534 | 23:36 |
Mister_Magister | if does, else doesn't | 23:36 |
mal | I have been considering removing those old legacy platform things from mce | 23:37 |
Mister_Magister | pushed the changes | 23:37 |
Mister_Magister | oh shiet | 23:37 |
Mister_Magister | now we good | 23:38 |
mal | wondering if that if( display_type == DISPLAY_TYPE_DISPLAY0 ) is good, it does work but would it be cleaner to just check if mdy_high_brightness_mode_output.path is non-null | 23:38 |
Mister_Magister | mdy_high_brightness_mode_output.path == NULL or nullptr? | 23:39 |
Mister_Magister | it doesn't seem to be initialized | 23:39 |
Mister_Magister | so I don't think you can check if its null | 23:39 |
mal | pretty sure it has to be null, otherwise g_free((void*)mdy_high_brightness_mode_output.path); which is done always would do bad things | 23:40 |
Mister_Magister | but which null | 23:41 |
mal | NULL seems to be used in other parts of the code | 23:41 |
Mister_Magister | aight | 23:41 |
Mister_Magister | pushed, now i'll rebuild it and check | 23:42 |
Mister_Magister | btw are there any docs that need updating? | 23:42 |
Mister_Magister | mmm don't think so | 23:42 |
mal | hmm, actually free does work on NULL so the if is useless in a way, no null check here either https://github.com/sailfishos/mce/blob/master/modules/display.c#L12400 | 23:43 |
Mister_Magister | okay i broke something cause it doesn't build xd | 23:43 |
Mister_Magister | ahh | 23:44 |
mal | if you need to fix something remove that null check | 23:44 |
Mister_Magister | i did | 23:45 |
Mister_Magister | don't you worry | 23:45 |
Mister_Magister | i needed to move the hbmdir variable | 23:45 |
Mister_Magister | aand it built | 23:45 |
mal | well hbmdir could be in the if also, some examples in the code elsewhere with variable inside if | 23:46 |
mal | since it's only used there anyway | 23:47 |
Mister_Magister | still works, didn't destroy my phone | 23:49 |
mal | also I think you don't release the hbmdir variable, check in mdy_display_type_get_from_config what is done to the return value of mce_conf_get_string_list | 23:49 |
mal | g_strfreev seems to be used | 23:49 |
Mister_Magister | you're entirely correct | 23:50 |
Mister_Magister | fixed now | 23:51 |
Mister_Magister | still works | 23:52 |
Mister_Magister | I guess I can add instructions to hadk faq after it gets merged | 23:53 |
Mister_Magister | if it gets merged* | 23:53 |
mal | sorry to nitpick but that is a bit ugly to have gchar **hbmdir = 0; and the on next line set a new value when it could be just one line, remember to fix indentation of next line when you have that | 23:53 |
Mister_Magister | and btw link to hadk faq is still broken | 23:53 |
Mister_Magister | mal: thought it was more according to the code style | 23:54 |
mal | https://github.com/sailfishos/mce/blob/master/modules/display.c#L5292 just an example | 23:54 |
Mister_Magister | ye ye i know | 23:54 |
Mister_Magister | I'm like professional or something xd | 23:54 |
mal | and I'm always complaining about style :) | 23:55 |
mal | well in this case also a bit more than just style | 23:56 |
Mister_Magister | as long as it gets merged ill suffer through nitpicking :P | 23:56 |
mal | wondering why those links are broken | 23:57 |
mal | it's not like the content is anything bad | 23:57 |
Mister_Magister | maybe use google's shortener | 23:58 |
Mister_Magister | most of those shorteners used are blocked by adblockers/malware blockers anyway | 23:59 |
mal | also that always-grep-irc-logs is broken, wondering where it pointed | 23:59 |
Mister_Magister | https://piggz.co.uk/sailfishos-porters-archive/ | 23:59 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!