| 29 Nov 2020 |
 Chris 🌱️ | There are two approvals already, wondering if we should merge. CC: Alexandre Franke, danigm | 20:36:44 |
 David Heidelberg | Mine feelings about !626: https://www.youtube.com/watch?v=1HVv2saWOl0 :D | 21:09:26 |
|  jjardon left the room. | 22:53:07 |
| 30 Nov 2020 |
 Alejandro Domínguez | In reply to @brainblasted:gnome.org
There are two approvals already, wondering if we should merge. CC: Alexandre Franke, danigm No, Alexandre found another regression that went unnoticed and didn't appear in previous iterations of the MR, so I suspect the culprit is somewhere in matrix-sdk and/or ruma (I made several version updates in the course of compensating for the lack of spec-compliancy of Synapse). BTW could you check !663 so I can merge it? I want to rebase !626 on top of it | 05:23:24 |
| Alejandro Domínguez | Kai: What Synapse version are you using as your homeserver? And how can I check which one is in matrix.org? | 08:55:57 |
 Kai | Both are at 1.23.0 The easiest way is probably this:
https://federationtester.matrix.org/#nasnotfound.de
https://federationtester.matrix.org/#matrix.org | 08:58:07 |
| Alejandro Domínguez | I've been looking around in the contents of matrix-rust-sdk and ruma, and I don't think there's anything in ruma that's causing this. matrix-sdk uses reqwest underneath which is too high level to trigger by using it the cancelation in hyper (for more context about this, see my last update on that thread in !626) | 09:01:07 |
| Alejandro Domínguez | So it's either that matrix.org has broken its setup lately or the error is buried inside reqwest or hyper itself | 09:02:10 |
| Alejandro Domínguez | * I've been looking around in the contents of matrix-rust-sdk and ruma, and I don't think there's anything in ruma that's causing this. matrix-sdk uses reqwest underneath which is too high level to trigger by using it the cancelation in hyper (for more context about this, see my last update on that thread in !626) | 09:02:24 |
 Alexandre Franke | Can you try and talk to the Synapse people? | 09:03:06 |
| Alexandre Franke | Could it be that the request is too large and we need some kind of pagination? | 09:03:28 |
| Alejandro Domínguez | The response is already paginated | 09:03:51 |
| Alexandre Franke | Oh. | 09:04:20 |
| Alejandro Domínguez | Because that task is cancelled in hyper, the decoder then finds that the size of the response is 0, so it returns exactly the "UnexpectedEof" error | 09:05:19 |
| Alejandro Domínguez | How it's possible that a month ago the directory stuff got a lot speedier in Fractal with this very same MR and now it's failing while it's waiting even longer than before is beyond my understanding | 09:06:57 |
| Alejandro Domínguez | * Because that task is cancelled in hyper, the decoder then finds that the size of the response is 0, so it returns exactly the "UnexpectedEof" error | 09:08:14 |
| Alejandro Domínguez | * How it's possible that a month ago the directory stuff got a lot speedier in Fractal with this very same MR and now it's failing while it's waiting even longer than before is beyond my understanding | 09:08:49 |
| Alejandro Domínguez | In reply to @afranke:matrix.org
Can you try and talk to the Synapse people? Just tried that, let's see if they can point me to the right direction | 09:13:22 |
| Alejandro Domínguez | I'm quite annoyed that an error as "stupid" as this one is the last blocker in !626 tbh | 09:13:59 |
 Gitlab Notifications | [GNOME/fractal] Alexandre Franke approveded merge request Change statics for consts (!663) | 09:24:27 |
| Gitlab Notifications | [GNOME/fractal#master] 1 new commit by Alejandro Domínguez | 09:24:59 |
| Gitlab Notifications | - Change statics for consts (2864e630)
| 09:24:59 |
| Gitlab Notifications | [GNOME/fractal] Alejandro Domínguez merged merge request Change statics for consts (!663) | 09:25:00 |
| Kai | Yeah, that is this kind of doing everything right and having it still break :/ I hope the Synapse people will have a fix for you. | 09:26:14 |
| Gitlab Notifications | [GNOME/fractal] Alejandro Domínguez opened merge request Replace legacy gtk functions (!665) | 09:59:33 |
| Gitlab Notifications | [GNOME/fractal#master] 1 new commit by Alexandre Franke | 10:31:16 |
| Gitlab Notifications | - Replace legacy gtk functions (5db67ad6)
| 10:31:16 |
| Gitlab Notifications | [GNOME/fractal] Alexandre Franke merged merge request Replace legacy gtk functions (!665) | 10:31:16 |
 Danny Colin | Where can I find an svg version of Fractal's logo? | 17:44:33 |
| Kai | Danny Colin: https://gitlab.gnome.org/GNOME/fractal/-/blob/master/fractal-gtk/res/icons/hicolor/scalable/apps/org.gnome.Fractal.svg | 18:29:29 |
