Paul "TBBle" Hampson
Paul "TBBle" Hampson
I think the core need here is for https://github.com/adamrehn/ue4-docker/blob/d832c11d8017fcca2f2894b44a1421152aab51db/ue4docker/infrastructure/WindowsUtils.py#L179-L184 to recognise any actual base tag supported by https://hub.docker.com/_/microsoft-windows-servercore, since #144 means we now have a base image where MS are...
We might need to change how we match OS versions once ltsc2022 is on the table anyway; right now we're using Windows 10 client versions, and mapping those to container...
Yeah, pretty much. I expect that there won't be a Windows 10 (client) release of 10.0.20345.x (or whatever the final build turns out to be, 10.0.20344.x is the latest preview)...
Unfortunately, that is just UBT telling us that UHT crashed. I assume the logs from UHT are in `/home/ue4/UnrealEngine/UHT-UnrealEditor-Linux-DebugGame.txt`, and hopefully will tell us the actual failure.
Well, that forum post makes sense, if it's a bug in Clang or UE4 is passing parameters to clang, that cause it to generate instructions that the older CPU doesn't...
Maybe worth a bug-report to Epic, since I expect the same problem happens if you do the build on the host outside Docker too. They could at least update their...
Sadly, there's nothing more-interesting in [the original commit message](https://github.com/EpicGames/UnrealEngine/commit/e140910c0c9f5c1cf8723d5217c7cf6c62a2e38d) that would explain this. It's also possible that the commit author's confusion was that because MSVC has no flags for extensions...
The only thing we could do here is patch-out the change in the source since that [reported worked](https://github.com/adamrehn/ue4-docker/issues/250#issuecomment-1119595247), but I'd like to see what Epic says first, particularly if it...
I don't think we would ever want to patch _every_ bug, but there is history of patching build-system bugs in ue4-docker, and "delete a line" is well within the scope...
I've never seen this idiom before, using `LABEL` to describe individual layers. My initial reaction is that the `LABEL` should be _after_ the layer it describes, because "layer parent" is...