| 27 Nov 2022 |
 Pol | I use VSCode, and rely on PHPCSFixer. | 18:03:12 |
| Pol | I can make a demo at anytime if you want. | 18:03:20 |
 hexa | I feared as much 😄 | 18:06:47 |
| hexa | maybe I should embrace vscodium at least | 18:06:55 |
| Pol | Use any editor you want, then just run phpcsfixer to reformat the code | 18:16:06 |
| Pol | I can show you everything when you want. | 18:16:27 |
| hexa | thanks, I'll get back to that when I think I'm uncomfortable with what I have | 18:24:12 |
| Pol | Okay | 18:36:02 |
 @etu:failar.nu | In reply to @drupol:matrix.org
etu: I created the backport: https://github.com/NixOS/nixpkgs/pull/203285 against release-22.11 is it ok ? I'm assuming the bot will do that. | 19:24:27 |
| Pol | Oki | 19:24:37 |
| Pol | Thx for merging ! | 19:24:41 |
| @etu:failar.nu | In reply to @hexa:lossy.network
my primary editors these days are neovim and helix I use intelephense (it's nonfree but it's in nixpkgs) with an LSP integration in Emacs, it's great. | 19:25:35 |
| hexa | same with helix | 19:29:06 |
| @etu:failar.nu | However I have some weird issue with intelephense... on my private laptop it works good... on my worklaptop it has started to crash a few weeks back... I even tried to clone the same open source project to the same filesystem path... | 19:33:16 |
| @etu:failar.nu | I don't have an error message other than "it crashed quickly" | 19:33:30 |
| @etu:failar.nu | If I run the same program by full path with same command line flag, it starts up fine. | 19:33:56 |
| @etu:failar.nu | Since both... systems are built from the same nix config locked to the same versions... the only real thing that differs is the... hostname :D | 19:35:06 |
| Pol | Oh | 19:48:42 |
| Pol | Intelephense on vscode ? | 19:48:48 |
| Pol | Never had issues with it :S | 19:48:53 |
| Pol | I wish it could be more active and provide more updates | 19:49:03 |
| Pol | I even paied the license to support the dev | 19:49:14 |
| @etu:failar.nu | In reply to @drupol:matrix.org
Intelephense on vscode ? No, emacs | 20:32:13 |
| hexa | works on helix 😄 | 20:42:42 |
| hexa | can't say I like php tho | 20:42:50 |
| Pol | Just ignore the coding style... and use phpcsfixer... | 21:10:05 |
| hexa | that's the plan 😄 | 22:46:32 |
| 29 Nov 2022 |
 tgerbet | Hey,
After encountering a segfault while running Psalm I saw that JIT SEAlloc in PCRE was re-enabled in https://github.com/NixOS/nixpkgs/pull/184634 (I disabled it in https://github.com/NixOS/nixpkgs/pull/184634)
The issue in the PHP bug tracker is indeed closed because it was considered there was nothing left to do on their side but it is still problematic.
Basically it breaks when the app fork(). This is mentioned in the PCRE README "It does not support fork() operation" https://www.pcre.org/readme.txt
As I said in my initial PR this option is likely to cause more troubles than the advantages it may bring. Especially since PHP 8 has a JIT on its own, running it in restricted env is likely to require more than enabling the PCRE JIT SEAlloc option.
How would you like to proceed on this etu Pol ?
| 16:41:13 |
| tgerbet | I drafted a PR extending the existing PCRE test to demonstrate the issue https://github.com/NixOS/nixpkgs/pull/203634 | 17:54:51 |
| 30 Nov 2022 |
| @etu:failar.nu | I guess we should include that test and a fix for it and backport it to 22.11 as well then :) | 05:58:00 |
