!lZLfSUtSOVjwYTmPbU:nixos.org

nixpkgs-update

150 Members
Keeping nixpkgs up to date. r-ryantm bot. https://github.com/ryantm/nixpkgs-update and https://github.com/nix-community/infra/blob/master/build02/nixpkgs-update.nix42 Servers

Load older messages


SenderMessageTime
22 Jul 2025
@asymmetric:matrix.dapp.org.ukasymmetric hexa: due to the amount of packages you maintain, the bot is hitting this issue 😉 20:01:14
@hexa:lossy.networkhexa🙂 20:01:19
@hexa:lossy.networkhexathat one issue that will never be fixed20:01:34
@hexa:lossy.networkhexaloving it20:01:35
@hexa:lossy.networkhexamight want to escape package names with backticks, if possible20:01:56
@hexa:lossy.networkhexaimage.png
Download image.png
20:02:34
@hexa:lossy.networkhexaalso, does it update the list over time?20:02:55
@asymmetric:matrix.dapp.org.ukasymmetricno, not atm. a new follow will sub to any newly maintained packages, but won't remove any. need to work on proper sync/reconciliation at some point 😅20:27:05
@asymmetric:matrix.dapp.org.ukasymmetric

ok now it handles cases with lots of subs/unsubs more gracefully:

  • prints something if M_TOO_LARGE happens
  • notifies user if a follow is taking > 10s
21:00:54
@sandro:supersandro.deSandro 🐧same yesterday. Now it seems to work somewhat.21:27:01
23 Jul 2025
@kuflierl:matrix.orgkuflierlAlright, it also works for me now09:08:39
@qaspr:matrix.orgJacoMalan1Is there a way to manually have r-ryantm create a PR for a package update, or do you need to wait for the package to bubble up to the top of the queue? Also, how long does it usually take for a package update ?11:52:43
@sandro:supersandro.deSandro 🐧nope, you need to wait11:55:09
@sandro:supersandro.deSandro 🐧depends, some days, sometimes longer11:55:18
@qaspr:matrix.orgJacoMalan1Damn11:55:25
@sandro:supersandro.deSandro 🐧if you want the PR immediately, just open it yourself 😅11:55:39
@qaspr:matrix.orgJacoMalan1If I wait, though, and the bot opens it, then I can merge it myself. Otherwise I have to clog up the already clogged up PR review queue 😂11:56:40
@sandro:supersandro.deSandro 🐧If it is a straight forward update and no build failures and you tick the PR template, it can probably be merged immeadetely 12:08:12
@sandro:supersandro.deSandro 🐧* If it is a straight forward update and no build failures and you tick the PR template, it can probably be merged immediately12:08:23
@kuflierl:matrix.orgkuflierlHonestly same issue here. I always tend to wait a bit for the update queue to trickle down12:09:42
@kuflierl:matrix.orgkuflierl I also have to ask... Why do we have a queue and why is it so long? 12:10:35
@sandro:supersandro.deSandro 🐧the bot just iterates through everything IIRC12:24:04
@sandro:supersandro.deSandro 🐧there is no real "queue"12:24:10
@kuflierl:matrix.orgkuflierl

I would beg to differ

https://nixpkgs-update-logs.nix-community.org/~supervisor/queue.html

12:26:19
@qaspr:matrix.orgJacoMalan1Yeah... It also seems like it does the queue from back to front12:27:22
@qaspr:matrix.orgJacoMalan1So 1st in line is actually last12:27:30
24 Jul 2025
@zowoq:matrix.orgzowoqUpdates are moved to the bottom of the queue when the bot starts processing them. There are multiple workers running package updates in parallel so the queue can advance quite a bit while an update is still in progress. Because of the time it takes to build a package and run nixpkgs-review it can take several hours for the bot to actually open a PR for an update.03:08:19
@sorimaxuki:matrix.orgsorimaxuki set a profile picture.07:16:51
@sorimaxuki:matrix.orgsorimaxuki removed their profile picture.07:19:23
@codgician:matrix.codgician.mecodgician changed their profile picture.16:58:08

Show newer messages


Back to Room ListRoom Version: 9