| 1 Dec 2024 |
|  @mmkaram:matrix.org joined the room. | 21:23:13 |
| @mmkaram:matrix.org changed their display name from Mahdy M. Karam to mmkaram. | 21:26:55 |
| 2 Dec 2024 |
 Mindavi | Has anyone ever benchmarked nix with huge pages? Does it help anything with evaluation? | 17:24:47 |
| 3 Dec 2024 |
 fricklerhandwerk | In reply to @emilazy:matrix.org
that page has "Shipped with the previous stable release" but I thought the support policy was that only the latest version + the version in stable NixOS were supported? so isn't that going to be pointing to the manual for an unsupported version soon? Uhm yes, but that’s not a contradiction. Just because the previous Nixpkgs release ships Nix 2.18 it doesn’t mean we have to actively support it | 14:43:50 |
| 4 Dec 2024 |
 Mic92 | emily: It would be a bit of a catch-22 if we could never remove a nix version that is still available in the previous nixpkgs version. It took unfortunally a bit to migrate all packages that were using nix 2.18 in nixpkgs. We will support nix 2.24 for sure. | 06:59:08 |
| Mic92 | I will try to also port all patches back for the important stuff that was still using 2.18 in the release. | 06:59:47 |
| Mic92 | * emily: It would be a bit of a catch-22 if we could never remove a nix version that is still available in the previous nixpkgs version. It took unfortunally a bit to migrate all packages that were using nix 2.18 in nixpkgs. We will support nix 2.24 for sure as this is the version that is used by default. | 07:00:07 |
| Mic92 | * emily: It would be a bit of a catch-22 if we could never remove a nix version that is still available in the previous nixpkgs version. It took unfortunately a bit to migrate all packages that were using nix 2.18 in nixpkgs. We will support nix 2.24 for sure as this is the version that is used by default. | 07:00:18 |
| Mic92 | Ah that was not the actual thing you asked actually | 07:02:11 |
| Mic92 | Well now it seems to point to nix 2.24. | 07:04:13 |
| Mic92 | * Well now the manual seems to point to nix 2.24. | 07:04:30 |
|  @tonton:envs.net left the room. | 22:49:41 |
| 5 Dec 2024 |
|  @rasmus:fricloud.dk left the room. | 17:36:54 |
 Matej Urbas | Hi Arian! You probably got to the bottom of this problem by now, right?
For the future though, I just added this S3 logging change to Nix: https://github.com/NixOS/nix/pull/12007
This will print an ID which you can use to find more information in CloudWatch. | 19:09:19 |
|  @nollie:matrix.org removed their display name nollie. | 20:38:27 |
| @nollie:matrix.org left the room. | 20:38:47 |
| 6 Dec 2024 |
|  @cafkafk:fem.gg changed their profile picture. | 03:43:51 |
 Robert Hensing (roberth) | Perhaps we should adopt the "oldstable" name? I've seen that name pop up during the support window overlap before | 13:59:55 |
 p14 | I just hit what looks like this old issue: https://github.com/NixOS/nixpkgs/issues/138157
It appears to repro on nixpkgs staging with nix 2.24.10 doing nix develop -f . hello. I get:
| 14:40:03 |
| p14 | * I just hit what looks like this old issue: https://github.com/NixOS/nixpkgs/issues/138157
It appears to repro on nixpkgs staging with nix 2.24.10 doing nix develop -f . hello. I get:
error: [json.exception.parse_error.101] parse error at line 38, column 445: syntax error while parsing value - invalid string: control character U+001E (RS) must be escaped to \u001E; last read: '" \n local sep=\"$1\";\n local name=\"$2\";\n local type oldifs;\n if type=$(declare -p \"$name\" 2> /dev/null); then\n local -n nameref=\"$name\";\n case \"${type#* }\" in \n -A*)\n echo \"concatStringsSep(): ERROR: trying to use concatStringsSep on an associative array.\" 1>&2;\n return 1\n ;;\n -a*)\n local IFS='<U+001E>'
| 14:40:52 |
| p14 | * I just hit what looks like this old issue: https://github.com/NixOS/nixpkgs/issues/138157
It appears to repro on nixpkgs staging (b8a6909c649634b9ef2183dd2848dfd52e49d305) with nix 2.24.10 doing nix develop -f . hello. I get:
error: [json.exception.parse_error.101] parse error at line 38, column 445: syntax error while parsing value - invalid string: control character U+001E (RS) must be escaped to \u001E; last read: '" \n local sep=\"$1\";\n local name=\"$2\";\n local type oldifs;\n if type=$(declare -p \"$name\" 2> /dev/null); then\n local -n nameref=\"$name\";\n case \"${type#* }\" in \n -A*)\n echo \"concatStringsSep(): ERROR: trying to use concatStringsSep on an associative array.\" 1>&2;\n return 1\n ;;\n -a*)\n local IFS='<U+001E>'
| 14:41:23 |
| p14 | I see a handful of issues of this style open against nix as well. https://github.com/NixOS/nix/issues?q=is%3Aissue+json+json.exception.parse_error+is%3Aopen
Is this a recently introduced stdenv issue | 14:44:37 |
| Mic92 | p14: https://github.com/NixOS/nix/pull/11921 | 15:53:27 |
| p14 | This is while running 'nix develop' though, there is no indication of any build taking place | 15:54:37 |
| p14 | * This is while running 'nix develop' on any package in staging though, there is no indication of any build taking place | 15:54:55 |
| Mic92 | p14: I have the feeling if all packages are doing that, we should also fix this in nixpkgs. It should not echo invalid json | 16:11:30 |
| Mic92 | Do you know what change introduces this? | 16:11:46 |
| p14 | In reply to @joerg:thalheim.io
Do you know what change introduces this? I do not, sorry, my time budget blew and I am now afk for a bit. Seems master is ok but not staging. | 16:16:50 |
| Mic92 | Maybe give a heads up in the staging channel at least. | 16:17:23 |
| Mic92 | The pull request is only improving the error message, but not fixing the root cause, which is on the nixpkgs side. | 16:17:55 |
