Closed rudymatela closed 10 years ago
Interesting. It fails the Travis CI tests with an error I haven't seen before: https://travis-ci.org/UoYCS-plasma/LazySmallCheck2012/builds/29268682
That's the actual error when cabal cannot satisfy dependencies:
...
rejecting: template-haskell-2.9.0.0 (conflict: base==4.6.0.1/installed-8aa...,
template-haskell => base==4.7.*)
rejecting: template-haskell-2.8.0.0, 2.7.0.0, 2.6.0.0, 2.5.0.0, 2.4.0.1,
2.4.0.0, 2.3.0.1, 2.3.0.0, 2.2.0.0 (conflict: lazysmallcheck2012 =>
template-haskell>=2.9 && <3)
Dependency tree exhaustively searched.
...
The error happens because, apparently, the default GHC for Travis is currently 7.6, and:
So actually, using GHC 7.6, cabal is unable to satisfy the dependency. I've added a line in .travis.yml
indicating which GHC to use: ghc: 7.8
. The Travis build now passes!
This is probably a good indication that GHC 7.6 is still widely used and maybe maintaining both versions via CPP could be worth it. Travis even has support of compiling against two different versions of GHC at one go.
Thanks for the contribution!
As a low-priority item in my TODO list is installing two different versions of GHC on my machine. When I do that, I'll try to make a way for it to compile on both versions.
You're welcome! Thanks for making LazySmallCheck2012 in the first place! :-D
This pull request introduces 3 unrelated changes:
*.swp
on.gitignore
.cabal install
.Stats
module was not exposed, that made any package using LSC crash with strange link errors.TySynInstD
constructors.if impl(template-haskell < v2.9) ... cpp-flags: LEGACYTH
to switch between the old and new APIs (I was lazy :grin: and don't have the older version to test here). Also, I'm not sure if it's worth the cost of maintaining that.Feel free to exclude any (or all) of the changes.