| 21 Nov 2025 |
K900 | Yes | 09:38:09 |
K900 | @vcunat linux 4601, darwin 178 | 09:45:48 |
vcunat | I'd merge immediately. Any reason not to? | 09:48:35 |
K900 | Not that I know of | 09:48:49 |
vcunat | (but testing that it fixes these issues will still be great even after the merge) | 09:49:03 |
uep | build host in my neighbour's attic is probably making some noise now | 09:49:46 |
K900 | Wait | 10:00:57 |
K900 | Hold on | 10:00:58 |
K900 | Actually | 10:01:00 |
K900 | HOLD THE FUCK ON. | 10:01:02 |
K900 | My work laptop is Intel | 10:01:11 |
K900 | Which I keep forgetting about | 10:01:14 |
K900 | How did I not see any of this | 10:01:22 |
K900 |  Download image.png | 10:06:04 |
K900 | Anyway time to go brr | 10:06:08 |
vcunat | After several hours we might have more binaries on cache.nixos.org, so that this is more accessible for testing by the affected people. | 10:07:30 |
K900 | Honestly most of it is just other unrelated updates | 10:08:02 |
K900 | I barely have any gtk4 stuff | 10:08:07 |
vcunat | So maybe that's why you haven't seen the bug? | 10:10:58 |
K900 | Possibly | 10:12:45 |
uep | yeah, mostly quite apps are fine | 10:16:22 |
uep | fractal is a bit jumpy because of the layout they use and matrix read-markers moving all the time | 10:16:48 |
uep | the system monitor graph thing is wild though | 10:17:01 |
uep | * yeah, mostly quiet apps are fine | 10:17:29 |
K900 | That's probably doing another render pass | 10:17:47 |
K900 | And then compositing the graph onto the window surface | 10:17:58 |
K900 | And when your synchronization primitive is "trust me bro" | 10:18:10 |
uep | yeah | 10:18:32 |
uep | it really does look like that.. chunks of other buffers in all the wrong places at all the wrong scales | 10:18:59 |
uep | flash flash jump jump | 10:19:08 |