T42 | <adampigg> Ill fix image building soon now obs is up again | 06:35 |
---|---|---|
T42 | <shouaccount> I am getting error when building libgbinder, https://paste.ubuntu.com/p/BXRqXq9Yb3/ any suggestions? | 08:03 |
T42 | <shouaccount> I think this has something to do with my vm crashing once before I was building | 08:08 |
T42 | <elros34> run git clean -xdf in source directory ad try again | 08:36 |
T42 | <shouaccount> didn't work | 08:58 |
T42 | <Verevka86> comment this line https://github.com/mer-hybris/libgbinder/blob/master/rpm/libgbinder.spec#L46 (re @shouaccount: I am getting error w...) | 09:03 |
T42 | <shouaccount> ok tryinh | 09:03 |
T42 | <shouaccount> ok trying (edited) | 09:03 |
lbt | PSA: some mer/sfos infra will be down as I fix the wiki and update a server | 09:52 |
T42 | <lawless305> There a FAQ as to what all has been ported already etc. etc. | 12:35 |
T42 | <b100dian> This, but it probably doesnt work because of the annoncement above https://wiki.merproject.org/wiki/Adaptations/libhybris | 13:38 |
T42 | <shouaccount> thanks, it worked although still have to figure out why my build does not boot :0 (re @Verevka86: comment this line ht...) | 13:40 |
T42 | <MyNameIsEarlB> is there a particular log to check if sailfish just shuts down randomly? | 16:28 |
T42 | <MyNameIsEarlB> itll just say goodbye and shut down randomly | 16:29 |
T42 | <Shirley Bell> Forwarded from unknown: 𝚃𝚑𝚊𝚗𝚔 𝚢𝚘𝚞 𝚏𝚘𝚛 𝚋𝚎𝚒𝚗𝚐 𝚜𝚘 𝚐𝚎𝚗𝚎𝚛𝚘𝚞𝚜, 𝚝𝚑𝚘𝚞𝚐𝚑𝚝𝚏𝚞𝚕 𝚊𝚗𝚍 𝚜𝚘 𝚝𝚛𝚞𝚎. 𝚆𝚒𝚝𝚑 $1000 𝚜𝚝𝚊𝚛𝚝-𝚞𝚙 𝚒 𝚖𝚊𝚍𝚎 $7000 𝚃𝚑𝚊𝚗𝚔 𝚢𝚘𝚞 𝚏𝚘𝚛 𝚔𝚎𝚎𝚙𝚒𝚗𝚐 𝚢𝚘𝚞𝚛 𝚠𝚘𝚛𝚍𝚜 𝚊𝚗𝚍 𝚋𝚎𝚒𝚗� | 17:11 |
T42 | <elros34> @MyNameIsEarlB power key press detected by mce should be logged in journalctl. Maybe you have faulty powerkey? | 17:27 |
rinigus | piggz: had a chance to think about chum repos? I must say that I like that wild west style approach | 17:32 |
rinigus | and when the wheels will fall off, we can think about lessons learned and what to do better | 17:32 |
piggz | hehe | 17:33 |
rinigus | at least it will start and will not be wasted for long discussions :) | 17:33 |
piggz | there is also the benevolant dicator approach, where only I decide what goes in | 17:33 |
rinigus | yes, if you wish. it just takes time | 17:33 |
rinigus | ... your time I mean | 17:34 |
piggz | sure ... some kind of sensible wild west would be good, where we have a quick discussion maybe? | 17:34 |
rinigus | we can have discussions as well, just not too much :). there is also chum:testing that can be used to see if the package brings whole repo down | 17:35 |
rinigus | so, we could let all add / change at chum:testing and then few of us promote to :chum | 17:36 |
piggz | yeah, lets do that | 17:36 |
T42 | <MyNameIsEarlB> Thanks, I'll check out journalctl. | 17:36 |
T42 | <MyNameIsEarlB> It doesn't happen using any other operating system on my Pinephone. I am using an older version though as I can't get the newer one to dl (re @elros34: @MyNameIsEarlB power...) | 17:36 |
piggz | lbt: first user complaint that they cant install amazifsh on 3.4 (no builds on obs) :P | 17:36 |
rinigus | as we don't have many SFOS versions supported, it is easy to start with chum as well. | 17:37 |
rinigus | lbt: pretty please - can you add a macro for each supported SFOS version that will indicate the version in corresponding SFOS repo at OBS | 17:38 |
rinigus | ? | 17:38 |
piggz | rinigus: I suggested that to mal also ... i guess will have to be packages somewhere | 17:38 |
rinigus | piggz: doesn't have to be. OBS macros can be used for that as well. | 17:38 |
lbt | wasn't that supposed to happen? | 17:39 |
piggz | https://en.opensuse.org/openSUSE:Build_Service_cross_distribution_howto | 17:39 |
lbt | and yes. we can add it to OBS builds | 17:39 |
rinigus | I opened SFO thread on that a long time ago, no official response yet | 17:39 |
rinigus | so, OBS is kind of going around it :) | 17:40 |
piggz | that would be handy .. for amazfish, mal found a workaround, but its a massive fluke | 17:40 |
lbt | mal: why don't we define sailfish version in the sdk ? Any idea? | 17:40 |
lbt | I guess there is the "it's all supposed to be done by features" argument... | 17:41 |
rinigus | lbt: we would suggest with piggz to go for a wild west style with chum and, if possible, allow all interested to submit packages to Chum:testing . from :testing, it could be promoted by pinging Chum maintainers. | 17:43 |
rinigus | as for sfos versioning, still would help with the packaging. for cross-platform apps, it is nice to know if you compile for fedora or sfos, for example | 17:43 |
rinigus | fedora and suse define those macros, sfos should do too | 17:44 |
piggz | i agree ... i had case with kcalendarcore being needed for 4.1, but not in prev versions ... so was very hard to write a spec that worked for all versions | 17:44 |
lbt | I don't want to make chum:testing world writeable as such - people could too easily edit other people's packages | 17:47 |
lbt | however allowing anyone to SR packages there is easy - I just make you two admins | 17:48 |
lbt | and you approve them | 17:48 |
lbt | you may want to track who owns which package | 17:49 |
lbt | what macros did you want? | 17:49 |
rinigus | lbt: OK, let's see how it works. no idea what is SR, but I guess it is something like pull request in OBS lingo | 17:50 |
lbt | yes | 17:50 |
piggz | lbt: maybe macros like in the abot cross distro howto? | 17:51 |
rinigus | macro like `sailfishos_version` and the version encoded. from piggz link, looks like suse is using 1010 for 10.10 version | 17:51 |
piggz | s/above | 17:51 |
rinigus | so, what would it be for 4.1.0.24 - 40100 should be sufficient? | 17:52 |
lbt | mmm | 17:52 |
rinigus | as 4.01.00 - 24 is minor anyway | 17:52 |
piggz | rinigus: maybe 040100 ... to account for sfos 10 :) | 17:53 |
lbt | 4.1.0 should probably be 401000 | 17:53 |
lbt | haha | 17:54 |
lbt | 040100 | 17:54 |
lbt | do you want to lose the .24 ? | 17:55 |
lbt | the other thing we have for every version is like 0.20210714.0.1 | 17:56 |
lbt | essentially it's the date of the snapshot | 17:56 |
lbt | so that is 4.2.0.9 | 17:57 |
piggz | ooo, leaked version number! | 17:57 |
lbt | hehe | 17:57 |
piggz | i think we could live without the minor version ... but, if its easy enough to add, there is probably no down side... | 17:58 |
piggz | i guess we you be writing code such as if sfos_version >= 04010000 .... | 17:58 |
lbt | yes - lots of trailing zeros is easy to get wrong :/ | 17:59 |
rinigus | piggz: but with sfos 10 it will still be larger than 4. so no need to have front zero. that way version 100 can be handled the same way | 18:01 |
rinigus | and indeed, too many numbers are easy to get wrong. as for using the date - then we need to consult the table with the dates | 18:02 |
lbt | ack | 18:03 |
lbt | so 4.0.1.24 => 40100 | 18:03 |
rinigus | that is fine by me. piggz ? | 18:03 |
rinigus | nonono, sorry, that would be 4.0.1.24 -> 4.0001 | 18:04 |
rinigus | 40001 | 18:04 |
rinigus | as in 4.00.01 | 18:04 |
lbt | so 4.1.0.24 => 40100 | 18:05 |
rinigus | yes | 18:05 |
lbt | so 4..0.1.48 => 40001 | 18:05 |
lbt | sorry | 18:05 |
piggz | yeah, i think thats how suse does it after flicking down the list | 18:05 |
rinigus | yes | 18:05 |
rinigus | just in case if we have 4.11.0.23 | 18:05 |
piggz | they have 9.3 and 10.0, as 930 1000 | 18:06 |
rinigus | will be afk for 20-30 mins | 18:06 |
lbt | done | 18:08 |
piggz | lbt: youre on a roll! | 18:12 |
piggz | lbt: rinigus: maybe a is_sailfishos macro too? | 18:12 |
lbt | the existence of sailfishos_version does that | 18:13 |
piggz | ok | 18:13 |
lbt | https://build.sailfishos.org/project/prjconf/sailfishos:4.1.0.24 line 3 fyi | 18:13 |
T42 | <elros34> @MyNameIsEarlB then I am not sure it's worth to debug it. Try to download latest version maybe from https://gitlab.com/sailfishos-porters-ci | 18:14 |
piggz | lbt: cool, that easy | 18:15 |
lbt | yes - need to make it part of my import | 18:15 |
lbt | and of course it won't work in SDKs :( | 18:15 |
piggz | sure ... have you tried to get 3.4 and 3.3 targets? | 18:16 |
piggz | so, is community obs on par with internal obs these days? | 18:17 |
lbt | it's the same codebase | 18:18 |
lbt | actually I should be able to copy all the internal _service files for the oss code to community too | 18:19 |
lbt | it's not that trivial as webhooks works differently internally | 18:20 |
lbt | I think 3.4.0.24 may be coming to life... | 18:26 |
piggz | neat | 18:26 |
lbt | ok - try it and let me know... | 18:27 |
piggz | will do ... | 18:30 |
piggz | just updating volla packages | 18:30 |
rinigus | lbt: thank you very much! | 18:48 |
rinigus | now I can start deleting my own sfos detection hacks (as if not suse or fedora, must be sfos) | 18:49 |
rinigus | lbt: we would probably make chum:testing for submissions - this one is missing so far. as for running many chum subrepos separate for each sfos version, this seems to be overkill. | 18:52 |
piggz | rinigus: except for when packages build against libraries that dont maintain compatibility between versions | 19:12 |
piggz | :looks at libicu | 19:13 |
rinigus | piggz: but those could be handled in the same source still - I mean that we don't need this https://build.merproject.org/project/subprojects/sailfishos:chum | 19:13 |
rinigus | instead, define multiple build targets in the same repo | 19:13 |
piggz | yes | 19:14 |
piggz | i thought that was what you were arguing against :) | 19:14 |
rinigus | at least for the start. there could be packages that don't work with some versions, but it is problem for future | 19:15 |
lbt | like: https://build.sailfishos.org/project/show/home:lbt:sailfishos | 19:47 |
rinigus | lbt: exactly, that would be easier to maintain. at least at the beginning | 19:53 |
lbt | https://build.sailfishos.org/project/show/sailfishos:chum:testing | 20:06 |
lbt | and 3.1.0.12 is looking OK | 20:07 |
rinigus | lbt: nice! I'll start on my side tomorrow. | 20:31 |
rinigus | thank you! | 20:31 |
rinigus | lbt: but those old versions - we don't need aarch64 for those | 20:32 |
rinigus | 3.4.0.24, I mean | 20:32 |
piggz | https://build.merproject.org/project/show/home:piggz:kf5 | 20:43 |
lbt | OK - I don't have a map of what's needed - it was released so I added it | 20:47 |
lbt | want me to remove it ? | 20:47 |
piggz | lbt: for histric purposes, will you go back to 1.0? :P | 20:47 |
lbt | piggz: I might - depends if I automate it ;) | 20:48 |
lbt | or if people ask for specific versions | 20:48 |
lbt | I know slava wants some old ones | 20:48 |
piggz | heh | 20:48 |
lbt | it's a bit tedious doing it by hand ... so on demand is more likely | 20:48 |
lbt | at least now we know it will work | 20:49 |
piggz | lbt: so, shall we assume wednedsays are quiet for you in the future ? ;) | 20:50 |
lbt | sadly not... I've been swamped at work and finally got some time off - was supposed to be doing stuff outside but ended up doing things that had big "wait gaps" so I made myself useful | 20:52 |
lbt | plus I got a handle on the issues: we did miss some packages from some releases and we added them to the prjconf to enable rust builds | 20:53 |
lbt | and now... dinner (late!) | 20:54 |
lbt | enjoy chum, the macros and the releases - let me know how it's going | 20:54 |
mal | lbt: is there some way to un-break projects without editing service files of each package | 20:55 |
mal | for older targets | 20:55 |
lbt | there's a trigger service | 20:56 |
lbt | or hit rebuild in the failed build | 20:57 |
mal | lbt: trigger rebuild is not helping | 20:57 |
mal | lbt: for example https://build.sailfishos.org/package/show/nemo:testing:hw:fairphone:fp2-sibon/droid-config-fp2-sibon | 20:58 |
mal | no indication why it thinks it is broken | 20:58 |
lbt | I think I may need to do a back end prod... | 21:00 |
lbt | it's going | 21:02 |
lbt | if you want to prepare a list I need projects / arch pairs that have problems | 21:02 |
lbt | ie this was: nemo:testing:hw:fairphone:fp2-sibon armv8el | 21:03 |
* lbt is afk now o/ | 21:03 | |
mal | a lot of projects are broken like that | 21:04 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!