I have a suspicion that this commit might have made development more painful by bringing longer build times and such. While I don't have concrete proof, I have tried reverting to QLoom 1.0 with that commit, and OOMs prevented that from working
I suspect that the reason this happens is due to the usage of modTestmodImplementation on QSL's modules and therefore having a need to remap all of them compared to Fabric simply pulling in it's module, which doesn't need remapping;
A solution would be reverting the migration while preventing the module from being published, which could work but wouldn't be ideal since we'll want to use Quilt's solution for game tests of our own;
Something must be investigated that keeps the usage of the Quilt Testing API while not causing the poor build system to suffer
I have a suspicion that this commit might have made development more painful by bringing longer build times and such. While I don't have concrete proof, I have tried reverting to QLoom 1.0 with that commit, and OOMs prevented that from working
I suspect that the reason this happens is due to the usage of
modTestmodImplementation
on QSL's modules and therefore having a need to remap all of them compared to Fabric simply pulling in it's module, which doesn't need remapping;A solution would be reverting the migration while preventing the module from being published, which could work but wouldn't be ideal since we'll want to use Quilt's solution for game tests of our own;
Something must be investigated that keeps the usage of the Quilt Testing API while not causing the poor build system to suffer