| poetaster | piggz, weird thing. went to make pizza, took a break, GS5 had gone 'down'. Restarted, it was like 'reset'. Checked what os, said 4.4. It's been running 4.5 for over a month. | 15:34 |
|---|---|---|
| poetaster | piggz, I don't get it. I had done some opt-kf5 install stuff, but nothing remotely related to the system itself. And i had turned off chum testing again. Just to not spoil my day. | 15:35 |
| poetaster | piggz, but the phone is like reset. what can I do? | 15:35 |
| poetaster | piggz, ignore me. Wrong phone. I gotta get a day job. | 15:36 |
| sskras | :) | 15:43 |
| poetaster | I have this bad habit of making all my phones look the same. bad idea. | 16:49 |
| *** amccarthy is now known as Guest1691 | 17:02 | |
| *** amccarthy_ is now known as amccarthy | 17:02 | |
| rinigus | poetaster: I have sent kaidan back to you in chum. according to the build results (rpm package description), few requirements are not provided and not found by that build. | 17:06 |
| rinigus | please fix and resubmit again | 17:06 |
| piggz | poetaster: lol -ETOOMANYPHONES | 17:24 |
| direc85[m] | @piggz: Me, too :') | 17:43 |
| poetaster | rinigus, I'm not sure what you mean by sent it back? | 18:05 |
| piggz | poetaster: rejected | 18:06 |
| poetaster | rinigus, it's also a bit immaterial. v0.8.0 is not really getting us anywhere. I'm just testing it to see if I can get omemo, qca and qxmpp to a certain 'state' (1.5) | 18:06 |
| poetaster | piggz, that's not very usefull. I don't want it in chum, I want it in testing. I'll probably throw it away anyway. I cann't maintain it. | 18:07 |
| poetaster | piggz, rinigus the supporting libraries, however, I can get to a certain point. even with all of your efforts, which are great, kaidan with v0.9.1 we will not nbe abe to build on obs. | 18:08 |
| piggz | poetaster: why? i think it would be a great addition | 18:09 |
| poetaster | piggz, v0.8.0 is castrated vs. MUC and omemo. The reason I submitted it to testing is to see what actually happens when resolution takes place on a device. Several errors I only spotted by doing local pkcon installs having installed the requirements on device. | 18:11 |
| poetaster | rinigus, are you referring to: kaidan.i486: W: desktopfile-without-binary /usr/share/applications/im.kaidan.kaidan.desktop qt-runner ? | 18:12 |
| rinigus | poetaster: I am referring to https://build.sailfishos.org/package/binary/home:poetaster:qt5.15/kaidan?arch=i586&filename=kaidan-0.8.0-1.61.1.jolla.i486.rpm&repository=sailfish_latest_i486 . See column on the right (Requires) | 18:13 |
| poetaster | rinigus, or: /usr/share/knotifications5/kaidan.notifyrc which I could not even using the method from kwallet to install in /opt. | 18:13 |
| rinigus | every empty cell means that there is no package that provides it | 18:14 |
| rinigus | there are also wrong requirements in terms of packages. | 18:14 |
| rinigus | poetaster: chum:testing is not for sending broken packages but as a final step before getting to chum | 18:14 |
| rinigus | there is no point in taking into chum:testing if there is no plan to send it to chum proper | 18:15 |
| rinigus | you can always play at your home repo | 18:15 |
| poetaster | rinigus, opt-kf5 is an experiment in hacking around the usuaal use of rpms. | 18:15 |
| rinigus | poetaster: I don't follow opt-kf5 statement. please explain | 18:16 |
| poetaster | rinigus, as you have yourself said quite clearly, we cannot depend on requires in the conventional sense but have to hard code everything. to discern how to do that is laborious. | 18:17 |
| poetaster | rinigus, to find a lower level issue (libmemo-c) I had to do installs on devices since nothing shows up on obs. | 18:17 |
| poetaster | rinigus, but I can 'fix' the kaidan issues if you wish, but I will not maintain it. it's not worth it. I'll help work on an alternative. 5 days of work to realize that the g++ toolchain won't get me farther an I'm burnt. | 18:19 |
| rinigus | poetaster: yes, it is laborious. to work around it, I am usually building using my docker images and check on them all the requirements using rpm. (don't have a history on this PC right now, so cannot provide commands). that way, most is sorted before I upload to OBS | 18:21 |
| rinigus | as for kaidan, then it makes sense to stop until someone will step up as a maintainer | 18:21 |
| rinigus | re gcc - we may have to make /opt/gcc again. although, I hope we can avoid it if jolla gets new gcc released. no idea if they even work on it | 18:22 |
| poetaster | rinigus, I will get it to the point that v0.8.0 builds cleanly and will advertise that fact on the forum. As for forther work, I'm going to see if I can support https://openrepos.net/content/ron282/shmong | 18:23 |
| poetaster | rinigus, to that end I have done qca, omemo and qxmpp builds that will support that. | 18:23 |
| poetaster | rinigus, c++20 is becomming a necessity. Although a lot of the reasons for that are questionable. | 18:24 |
| poetaster | rinigus, in any case, sorry to make extra work for you. | 18:24 |
| piggz | you cant blame devs for wanting to use newer/easier c++ features | 18:25 |
| poetaster | piggz, the devs of kaidan implemented class operator== in the most lazy way and continue to use bad form just because 20++ might let you get away with it. suck. | 18:26 |
| poetaster | piggz, I rewrote all the code in Message.h with macros (c++ 17 conform) and it worked. but that's a can of worms that, though I enjoyed the dive, it was clear to me I was not going to take to them. | 18:27 |
| poetaster | piggz, I've been close reading the kpublictransport code for more than a year now and I also realized that I'm not going there. too many disagreements about how it's designed to engage in that. | 18:28 |
| poetaster | and all my code sucks, so better do some cleanup at home :) | 18:28 |
| poetaster | rinigus, how do you get to that view of kaidan you sent here? I cna't find that? | 18:32 |
| poetaster | rinigus, sorry, found it. never seen it before. | 18:33 |
| rinigus | poetaster: its a very handy view and I was referring to it when we were discussing requirements earlier (saturday?). | 18:35 |
| poetaster | rinigus, sigh. I had never seen that before. | 18:36 |
| rinigus | poetaster: well, now you have seen it :) . | 18:36 |
| poetaster | rinigus, which does not make me the wiser :_ ) opt-kf5-kquickimageeditor looks like it's also still /wrong/ | 18:37 |
| poetaster | ah great. shows different things on differnt archs. | 18:39 |
| poetaster | never mind. rebuilding. | 18:41 |
| poetaster | %{_opt_kf5_datadir}/knotifications5/ sigh. | 19:02 |
| poetaster | rinigus having gcc in a newer variant would actaully make it 'easy' to get kaidan (also qxmpp 1.5.5) building. without that, kaidan is castrated. I mena, the whole chat wars have been won by idiots anyway, but, just saying. | 19:08 |
| poetaster | rinigus, where else is opt-qca being used? | 19:26 |
| rinigus | poetaster: kf5-kwallet | 19:29 |
| poetaster | rinigus, ah, ok. cool. that's where I took the hint for installing notifications configs to. but that bombed. | 19:39 |
| poetaster | rinigus, libQt5Core.so.5(Qt_5.15) why is that an isuse when libQt5Core.so.5(Qt_5) qt5-qtcore ... what does that mean? | 19:42 |
| poetaster | rinigus, I'm guessing that an over explicit define ? %global qt_version 5.15.9 | 19:43 |
| *** phlixi is now known as Guest1702 | 22:56 | |
| *** phlixi_ is now known as phlixi | 22:56 | |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!