Joachim Breitner wrote: > ah, ok, so that is an implicit re-assign of the bug to xmonad? I read it > in the ghc6 context.
Oops, as you maintain them both, I had not realized this bug was currently assigned to ghc6 rather than xmonad. Yes, I think this bug should go back to xmonad, maybe as a clone. xmonad currently has this annoying usability problem across ghc6 upgrades, and it could be fixed to avoid it. > I guess I need to add some calls "removeFile", ignoring all exceptions, > to the "recompile" function. Or, to keep the changes against upstream > minimal, at least suggest the work-around in the message that informs > the user about a failed build. I guess it could even try deleting the files and building again if a normal build fails for best of all worlds. But maybe too hackish for the xmonad developers. -- see shy jo
signature.asc
Description: Digital signature
