| 17 Mar 2024 |
 hexa | kinda because php82 is
let
base = callPackage ./generic.nix (_args // {
version = "8.2.17";
hash = "sha256-GRMWwgMmfZYWC0fSL5VdTcEXk96KXzJ+DCp2J1polOo=";
});
in
base.withExtensions ({ all, ... }: with all; ([
| 12:53:01 |
| hexa | so it does include all extensions already? | 12:53:12 |
| hexa | * so it does include alot extensions already? | 12:53:21 |
| hexa | * so it does include alot of extensions already? | 12:53:25 |
| hexa | trimming it down was an afterthought really | 12:53:38 |
| hexa | did I understand that wrong? | 12:54:01 |
| hexa | ok, I am stupid … my local staged changes do actually work. | 12:55:09 |
 tgerbet | They are not all enabled but yes | 12:55:25 |
| tgerbet | Hum in this case it should work out of box | 12:56:10 |
| hexa | and instead of constructing it twice, wouldn't it be better to expose and reuse it as engelsystem.php | 12:56:22 |
 @patka_123:matrix.org | In reply to @hexa:lossy.network
ok, I am stupid … my local staged changes do actually work. No, not stupid, smart. Because now I had the opportunity to learn a few things I hadn't seen before ^^
It was for the greater good
| 12:56:41 |
| hexa | you're welcome | 12:56:59 |
| @patka_123:matrix.org | (Feel free to change my PR or comment on it, I don't mind. But I will get to it tomorrow because I'm not close to a pc) | 12:57:42 |
| tgerbet | Yes we probably should expose something like this for things that are built with buildComposerProject
Looking the phpfpm services declared in nixos/ they would all benefit from it | 12:58:37 |
| hexa | I think making it work and making it pretty can be a two step process | 12:58:43 |
| hexa | yeah, it is the wild, wild west | 12:59:15 |
| hexa | also, I used to set session.gc_maxlifetime to a week or so for engelsystem | 13:03:29 |
| hexa | the default seems to be 24 minutes (1440s) | 13:03:37 |
| hexa | what is up with that? | 13:03:39 |
| hexa | the effect was that people needed to login again on nearly every visit to the engelsystem | 13:04:16 |
| hexa | they got so annoyed that they complained about it 🙂 | 13:04:23 |
| hexa | (I'm fairly certain I could also be holding it wrong) | 13:05:25 |
| tgerbet | It's the PHP default value https://www.php.net/manual/en/session.configuration.php#ini.session.gc-maxlifetime
It can be annoying if the app maintains authentication via the PHP session mechanism (also session.cookie_lifetime)
| 13:26:04 |
| hexa | so I'd better go with shorter sessions, but longer cookies? | 13:38:40 |
| @patka_123:matrix.org | Yes | 13:51:43 |
| tgerbet | Looking at the app implementation they are relying on the Symfony Session component
https://symfony.com/doc/current/session.html#configuring-the-session-ttl
Looking at https://github.com/engelsystem/engelsystem/blob/1217de096a61cd4e0863f587a01a38d141202f21/src/Http/SessionServiceProvider.php#L55-L68 and https://github.com/engelsystem/engelsystem/blob/1217de096a61cd4e0863f587a01a38d141202f21/config/config.default.php#L423-L429 the cookie lifetime seems to be adjusted but not the ttl attribute used by Symfony | 13:55:51 |
| tgerbet | So it likely fallback to the session.gc_maxlifetime setting instead of the $sessionConfig['lifetime'] defined by the ap | 13:56:51 |
| tgerbet | * So it likely fallback to the session.gc_maxlifetime setting instead of the $sessionConfig['lifetime'] defined by the app | 13:56:54 |
| hexa | thanks for looking that up! | 15:54:55 |
| 19 Mar 2024 |
|  NixOS Moderation Botchanged room power levels. | 00:30:24 |
