You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Right now, we can't test on GHC HEAD because the latest version of the stm package that hackage has breaks on GHC HEAD. I haven't checked this yet, but I suspect that the HEAD of the stm repo would build correctly since the current hackage version has been broken on GHC HEAD for at least a month.
I don't know enough about travis and cabal to suggest a way to fix this, so I thought I'd bring this to everyone's attention since it is important to continue to check in on compatibility with GHC HEAD.
This seems like the kind of thing that stack could help with, since you can tell it to use a package on a git repo at a specific commit, instead of only being able to rely on package versions that are on hackage. I'm not sure if there's a way to make that conditional so that it would only use that commit if it were being tested against GHC HEAD though.
The text was updated successfully, but these errors were encountered:
We can modify the travis build for HEAD to fetch git repos and build them locally, but yeah, maybe stack is a better answer long term. I know almost nothing about stack, but travis works by defining a 'build matrix' (basically, all the possible ways to build/test), so it shouldn't be a problem to have special behavior/sources for the HEAD build.
Right now, we can't test on GHC HEAD because the latest version of the
stm
package that hackage has breaks on GHC HEAD. I haven't checked this yet, but I suspect that the HEAD of thestm
repo would build correctly since the current hackage version has been broken on GHC HEAD for at least a month.I don't know enough about travis and cabal to suggest a way to fix this, so I thought I'd bring this to everyone's attention since it is important to continue to check in on compatibility with GHC HEAD.
This seems like the kind of thing that
stack
could help with, since you can tell it to use a package on a git repo at a specific commit, instead of only being able to rely on package versions that are on hackage. I'm not sure if there's a way to make that conditional so that it would only use that commit if it were being tested against GHC HEAD though.The text was updated successfully, but these errors were encountered: