| 7 Nov 2024 |
 emily | VirtualBox maintainer signed off on the idea and I think it's best to get it out of the critical path rather than keeping around a demo-only appliance that adds to the gigabytes of image-related channel bloat and has few advantages over the ISO most users will use | 07:08:37 |
| emily | especially since VirtualBox itself is so crusty (cf. this was in reaction to it breaking the channel because a QEMU change broke the installation procedure for our binary package of the non-Free 90s-vintage Open Watcom compiler that VirtualBox uses to compile its 16-bit BIOS code) | 07:09:26 |
| emily | (thankfully that chain of despair has been broken at multiple points already, but…) | 07:09:44 |
| emily | channel scripts will need deploying if it's approved; I will also make a PR to the homepage to remove that section for 24.11 in that case | 07:11:23 |
| emily | * VirtualBox maintainer (actually two of them) signed off on the idea and I think it's best to get it out of the critical path rather than keeping around a demo-only appliance that adds to the gigabytes of image-related channel bloat and has few advantages over the ISO most users will use | 07:17:54 |
 Vladimír Čunát | I'd say we just merge it. (first the channel-scripts part) | 07:27:01 |
| Vladimír Čunát | It's been hanging there for over a week, not trying to hide, and I haven't noticed any argument against really. | 07:27:36 |
| emily | SGTM | 07:28:15 |
| emily | if you want to hit the button and deploy the scripts stuff | 07:28:29 |
| Vladimír Čunát | Today I feel overwhelmed, but I should do it this week if noone else steps in. | 07:31:09 |
| emily | no worries :) | 07:32:49 |
| emily | (I guess it's the infra team that has access to merging/deploying channel scripts PRs?) | 07:33:09 |
| Vladimír Čunát | Yes. This runs on pluto.nixos.org. | 07:33:46 |
 Paul Meyer (katexochen) | 2nd staging-next is open, are breaking changes already unrestricted on staging? | 13:43:43 |
 Mic92 | In reply to @vcunat:matrix.org
Today I feel overwhelmed, but I should do it this week if noone else steps in. what needs to be done? | 14:15:17 |
 Tristan Ross | In reply to @katexochen:matrix.org
2nd staging-next is open, are breaking changes already unrestricted on staging? For staging, that will be the 8th | 16:07:42 |
 hexa | In reply to @emilazy:matrix.org
if you want to hit the button and deploy the scripts stuff deployed | 16:19:15 |
| Tristan Ross | In reply to @emilazy:matrix.org
https://github.com/NixOS/nixpkgs/pull/352603 is pending a decision for 24.11 Oh, I didn't see this one on my feed lol. I'll take a look. | 16:35:14 |
| 8 Nov 2024 |
| hexa | pkgs.formats.yaml was migrated to yaml 1.2 output in the last remarshal update | 18:01:28 |
| hexa | the output now unquotes certain outputs, that previously needed to be quoted in yaml 1.1 | 18:02:08 |
| hexa | which otoh means that when parsed by a yaml 1.1 parser will lead to unexpected results | 18:02:23 |
| hexa | $ cat test.json
{
"time": "22:30:00"
}
$ json2yaml test.json test.yaml
$ cat test.yaml
time: 22:30:00
$ python3
Python 3.12.7 (main, Oct 1 2024, 02:05:46) [GCC 13.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import yaml
>>> yaml.safe_load(open("test.yaml"))
{'time': 81000}
>>>
| 18:02:34 |
| hexa | $ cat test.json
{
"enable": "on",
"disable": "off"
}
$ json2yaml test.json test.yaml
$ cat test.yaml
enable: on
disable: off
$ python3
Python 3.12.7 (main, Oct 1 2024, 02:05:46) [GCC 13.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import yaml
>>> yaml.safe_load(open("test.yaml"))
{'enable': True, 'disable': False}
| 18:02:50 |
| emily | IIRC YAML 1.2 isn't that widely-adopted (though maybe that's changed?). is there a remarshal flag to use 1.1? | 18:03:31 |
| hexa | I'm contemplating whether to provide the remarshal package in an old version for yaml 1.1 support | 18:03:44 |
| hexa | and have a toggle for the formatter that switches the package around | 18:03:56 |
| emily | I think we shouldn't change the output of those kinds of things in general except for clear bugfixes. even making escapeShellArg omit quotes when unnecessary broke tons of stuff across the entire ecosystem in and out of tree | 18:04:12 |
| hexa | no, they switched from pyyaml to ruamel-yaml, which is a clean 1.1 to 1.2 swap | 18:04:13 |
| emily | ah | 18:04:21 |
| emily | https://github.com/remarshal-project/remarshal/commit/ee9a5b5d5ed82d90223a1d5fc74a4d546a034830 | 18:04:40 |
