2 Aug 2021 |
εΉΈη« (πππΎπ/πππΎπ) | Licaon_Kter[xmpp]: I've just updated "Jiten [Online]" to v1.0.3, which adds the option to choose another server :) | 00:23:19 |
Licaon_Kter[xmpp] | εΉΈη« (πππΎπ/πππΎπ): π | 05:26:38 |
Passiing | In reply to @obfusk:matrix.org sent an image. Yeah that's much better(contrast wise) | 08:53:32 |
jochensp | FestplattenSchnitzel: do you need help with squashing? | 12:17:00 |
FestplattenSchnitzel | Fortunately I figured it out, but it still requires a little fix and I'll be AFK for a few hours β¦ | 12:30:29 |
FestplattenSchnitzel | But thanks for asking! :) | 12:30:47 |
| Licaon_Kter[xmpp] left the room. | 13:26:45 |
| Licaon_Kter[xmpp] joined the room. | 13:27:06 |
FestplattenSchnitzel | jochensp: how can I replace the commits at my fork with those I have locally (and squashed)? git push --force ? | 16:46:23 |
jochensp | FestplattenSchnitzel: yes, or better --force-with-lease | 16:47:25 |
FestplattenSchnitzel | Thanks, that worked! :tada: | 16:50:24 |
FestplattenSchnitzel | * Thanks, that worked! π | 16:50:37 |
FestplattenSchnitzel | Feel free to merge! | 17:06:27 |
jochensp | FestplattenSchnitzel: lgtm, feel free to self merge (I don't have my credentials at hand) | 17:22:50 |
εΉΈη« (πππΎπ/πππΎπ) | there's finally a kotlin package in Debian's NEW queue. | 17:39:37 |
dazinism | In reply to @obfusk:matrix.org I've seen the vagrant source code. I'm not touching anything made by hashicorp if I have a choice. Seen https://darkcrystal.pw/ ? | 20:34:54 |
_hc | proletarius101: Sylvia as far as I know, the signing process is almost never the problem. Its the build process, and that the signing process waits for the build process to finish. If you want faster builds, I recommend working on making it possible for build and signing to be run on their own scheduiles. | 21:25:27 |
Licaon_Kter[xmpp] | _hc: the build ended 2-3 days earlier... Yes, sometimes that's the issue, but many times it is not. | 21:27:15 |
Sylvia | _hc: Really though? Because if I look at https://f-droid.org/wiki/index.php?title=Special:RecentChanges&days=14&from=&hidebots=0&limit=500, the build ended on 27-07-2021 14:12 and deploy didn't start until 31-07-2021 15:51 | 21:28:06 |
Sylvia | From my experience, the whole build process takes a bit over 24 hours and then there is at least 2 to 3 days of waiting for signing. So I really disagree the bottleneck is in the builds tbh | 21:28:47 |
Sylvia | In fact, looking at https://monitor.f-droid.org/builds/running it seems that roughly 20 hours passed already since the build completed (which took 1 day and 4 hours) so... we're already pretty close to waiting for signing to take long again | 21:31:40 |
Licaon_Kter[xmpp] | Sylvia: build is limited to 36h, that limit is hit sometimes. | 21:34:05 |
Licaon_Kter[xmpp] | After the last app is built it takes 5-6 to end, I guess it parses apps that don't need to be build, but not sure. | 21:34:06 |
jochensp | I agree with Sylvia that currently signing is delayed and the build is reasonably fast. On the other hand the signing bottleneck is known for a long time and the only thing that would help is someone working on the proposed solution, complaining here will likely lead to the opposite. | 22:21:25 |
jochensp | Licaon_Kter[xmpp]: yes, it parses the remaining apps in that time | 22:21:45 |
3 Aug 2021 |
Izzy | Just wondering what "fdroid update" wants to do with git if there's a .git directory on the top level of a repo? Currently setting up a repo for a customer and ran into that issue: "CRITICAL: Unknown exception found!" with a stack trace ending on "ValueError: Reference at 'refs/heads/master' does not exist" (the repo was just initialized, nothing committed yet β and the intention was to have metadata maintained in git). | 08:48:56 |
Izzy | Now I'm worried that "fdroid update" will interfere with something whenever I run "fdroid update"β¦ | 08:49:35 |
Licaon_Kter[xmpp] | Izzy: with git? My fdroiddata is always from git so .git always there | 08:57:28 |
Izzy | Not in my repo, so I never encountered that before. I've just set up a new repo for a customer and we decided to maintain metadata via git, so I ran "git init" on the root of that repo. Just ran "fdroid update" after that, and got that error. Looks like it wants to check in something, as that crash happens after both indices (repo+archive) were created and signed. | 09:06:40 |
jochensp | Izzy: does it work if you commit the metadata before running fdroid update? | 09:08:11 |