12 Jul 2024 |
| @valconius:matrix.org left the room. | 01:16:22 |
14 Jul 2024 |
Gaétan Lepage | Redacted or Malformed Event | 15:16:59 |
Gaétan Lepage | Redacted or Malformed Event | 15:17:05 |
Gaétan Lepage | Redacted or Malformed Event | 15:17:38 |
Gaétan Lepage | Redacted or Malformed Event | 15:18:05 |
Gaétan Lepage | Redacted or Malformed Event | 15:18:19 |
Gaétan Lepage | Redacted or Malformed Event | 15:18:22 |
16 Jul 2024 |
| @nairou:matrix.org joined the room. | 00:23:50 |
| @nairou:matrix.org left the room. | 13:41:39 |
18 Jul 2024 |
| BigBabyThor joined the room. | 00:52:54 |
23 Jul 2024 |
| Ezzobir Bezziou joined the room. | 08:23:23 |
24 Jul 2024 |
| Redstone changed their display name from redstone-menace to Redstone. | 10:14:41 |
27 Jul 2024 |
| dave joined the room. | 23:24:38 |
1 Aug 2024 |
asymmetric | hey, i'm planning on building a simple external service that:
- exposes a matrix bot for people to subscribe to nixpkgs-update build failures for a package
- periodically (and infrequently, e.g. once a day) parses new logs from https://nixpkgs-update-logs.nix-community.org/
- if there are new errors for a package, notifies people who have subscribed
i think this is the best way to validate the idea, and maybe one day it can be better integrated/upstreamed into nixpkgs-update, if need be
| 08:14:01 |
asymmetric | * hey, i'm planning on building a simple service that:
- exposes a matrix bot for people to subscribe to nixpkgs-update build failures for a package
- periodically (and infrequently, e.g. once a day) parses new logs from https://nixpkgs-update-logs.nix-community.org/
- if there are new errors for a package, notifies people who have subscribed
i think this is the best way to validate the idea, and maybe one day it can be better integrated/upstreamed into nixpkgs-update, if need be
| 08:14:09 |
asymmetric | i'll be using the user agent "asymmetric", in case you want to track usage | 12:19:55 |
Philip Taron (UTC-8) | I'm quite interested. Thanks asymmetric. Please feel free to use my handle as an assigned tester. | 16:17:19 |
Mindavi | I'm interested in that too | 17:02:54 |
hexa | In reply to @asymmetric:matrix.dapp.org.uk i'll be using the user agent "asymmetric", in case you want to track usage better yet, the url to the project repo 🙂 | 22:39:12 |
2 Aug 2024 |
asymmetric | In reply to @hexa:lossy.network better yet, the url to the project repo 🙂 but naming is so hard 🙈 nixpkgs-update-notify-bot? 😅 | 07:38:26 |
asymmetric | In reply to @hexa:lossy.network better yet, the url to the project repo 🙂 * but naming is so hard 🙈 nixpkgs-update-notify-bot? nun (nixpkgs-update-notifier)? 😅 | 07:38:56 |
shawn8901 | Is there a way how to disable r-ryantm to update a specific package? I was trying to disable vmagent for being updated (as we build it out of VictoriaMetrics package via override). I was trying to add # nixpkgs-update: no auto update being added as comment. https://github.com/r-ryantm/nixpkgs/blob/a9e94032782cc6f89c5fabcf5d9433b45578cad1/pkgs/by-name/vm/vmagent/package.nix#L5 Thus it did still create a PR ( https://github.com/NixOS/nixpkgs/pull/331673 ). I would have assumed that I would just create a PR for VictoriaMetrics now, does somebody know what is wrong? | 10:40:08 |
shawn8901 | * Is there a way how to disable r-ryantm to update a specific package if it's build out of another? I was trying to disable vmagent for being updated (as we build it out of VictoriaMetrics package via override). I was trying to add # nixpkgs-update: no auto update being added as comment. https://github.com/r-ryantm/nixpkgs/blob/a9e94032782cc6f89c5fabcf5d9433b45578cad1/pkgs/by-name/vm/vmagent/package.nix#L5 Thus it did still create a PR ( https://github.com/NixOS/nixpkgs/pull/331673 ). I would have assumed that I would just create a PR for VictoriaMetrics now, does somebody know what is wrong? | 10:46:37 |
Mindavi | In reply to @asymmetric:matrix.dapp.org.uk but naming is so hard 🙈 nixpkgs-update-notify-bot? nun (nixpkgs-update-notifier)? 😅 I like nixpkgs-update-notifier | 13:52:12 |
asymmetric | In reply to @hexa:lossy.network better yet, the url to the project repo 🙂 done | 14:06:52 |
hexa | 🙂 | 14:07:07 |
3 Aug 2024 |
zowoq | In reply to @shawn8901:matrix.org
Is there a way how to disable r-ryantm to update a specific package if it's build out of another? I was trying to disable vmagent for being updated (as we build it out of VictoriaMetrics package via override). I was trying to add # nixpkgs-update: no auto update being added as comment. https://github.com/r-ryantm/nixpkgs/blob/a9e94032782cc6f89c5fabcf5d9433b45578cad1/pkgs/by-name/vm/vmagent/package.nix#L5 Thus it did still create a PR ( https://github.com/NixOS/nixpkgs/pull/331673 ). I would have assumed that I would just create a PR for VictoriaMetrics now, does somebody know what is wrong? If you want the comment to work need to fix meta.position:
❯ nix eval -f. vmagent.meta.position
"/Users/zowoq/src/nixpkgs/pkgs/by-name/vi/victoriametrics/package.nix:87"
diff --git a/pkgs/by-name/vm/vmagent/package.nix b/pkgs/by-name/vm/vmagent/package.nix
index f31526026319..612144762d6f 100644
--- a/pkgs/by-name/vm/vmagent/package.nix
+++ b/pkgs/by-name/vm/vmagent/package.nix
@@ -3,7 +3,7 @@
# This package is build out of the victoriametrics package.
# so no separat update prs are needed for vmagent
# nixpkgs-update: no auto update
-lib.addMetaAttrs { mainProgram = "vmagent"; } (
+lib.addMetaAttrs { mainProgram = "vmagent"; position = toString ./package.nix + ":1"; } (
victoriametrics.override {
withServer = false;
withVictoriaLogs = false;
❯ nix eval -f. vmagent.meta.position
"/Users/zowoq/src/nixpkgs/pkgs/by-name/vm/vmagent/package.nix:1"
Could add it to the skiplist instead: https://github.com/nix-community/nixpkgs-update/blob/main/src/Skiplist.hs
| 05:36:45 |
shawn8901 | In reply to @zowoq:matrix.org
If you want the comment to work need to fix meta.position:
❯ nix eval -f. vmagent.meta.position
"/Users/zowoq/src/nixpkgs/pkgs/by-name/vi/victoriametrics/package.nix:87"
diff --git a/pkgs/by-name/vm/vmagent/package.nix b/pkgs/by-name/vm/vmagent/package.nix
index f31526026319..612144762d6f 100644
--- a/pkgs/by-name/vm/vmagent/package.nix
+++ b/pkgs/by-name/vm/vmagent/package.nix
@@ -3,7 +3,7 @@
# This package is build out of the victoriametrics package.
# so no separat update prs are needed for vmagent
# nixpkgs-update: no auto update
-lib.addMetaAttrs { mainProgram = "vmagent"; } (
+lib.addMetaAttrs { mainProgram = "vmagent"; position = toString ./package.nix + ":1"; } (
victoriametrics.override {
withServer = false;
withVictoriaLogs = false;
❯ nix eval -f. vmagent.meta.position
"/Users/zowoq/src/nixpkgs/pkgs/by-name/vm/vmagent/package.nix:1"
Could add it to the skiplist instead: https://github.com/nix-community/nixpkgs-update/blob/main/src/Skiplist.hs
Let me try the position first :) thx! | 07:25:21 |
shawn8901 | In reply to @zowoq:matrix.org
If you want the comment to work need to fix meta.position:
❯ nix eval -f. vmagent.meta.position
"/Users/zowoq/src/nixpkgs/pkgs/by-name/vi/victoriametrics/package.nix:87"
diff --git a/pkgs/by-name/vm/vmagent/package.nix b/pkgs/by-name/vm/vmagent/package.nix
index f31526026319..612144762d6f 100644
--- a/pkgs/by-name/vm/vmagent/package.nix
+++ b/pkgs/by-name/vm/vmagent/package.nix
@@ -3,7 +3,7 @@
# This package is build out of the victoriametrics package.
# so no separat update prs are needed for vmagent
# nixpkgs-update: no auto update
-lib.addMetaAttrs { mainProgram = "vmagent"; } (
+lib.addMetaAttrs { mainProgram = "vmagent"; position = toString ./package.nix + ":1"; } (
victoriametrics.override {
withServer = false;
withVictoriaLogs = false;
❯ nix eval -f. vmagent.meta.position
"/Users/zowoq/src/nixpkgs/pkgs/by-name/vm/vmagent/package.nix:1"
Could add it to the skiplist instead: https://github.com/nix-community/nixpkgs-update/blob/main/src/Skiplist.hs
it seems that setting the position fixes my overall issue. Tho not being on the ignored list but complaining before that the old version is not defined in the master derivation (as those are defined on the overriden ones). I suspect that nixpkgs-update might should check for ignores first, and then for sanity [as it would at least in this case not result in a faulty error log]. Tho it still fine for my result. ill keep the ignore comment still, so that the initial intention is clear why meta.position is added. | 19:07:56 |
shawn8901 | * it seems that setting the position fixes my overall issue. Tho not being on the ignored list but complaining before that the old version is not defined in the master derivation (as those are defined on the overriden ones). I suspect that nixpkgs-update might should check for ignores first, and then for sanity [as it would at least in this case not result in a faulty error log]. Tho it still fine for my result. ill keep the ignore comment still, so that the initial intention is clear why meta.position is added. Or do you have other suggestions for that? 🙂 | 19:22:18 |