
On Tue, Oct 4, 2011 at 12:49 PM, Johan Tibell
On Tue, Oct 4, 2011 at 7:57 AM, Duncan Coutts
wrote: In general, yes I think benchmarking is a sensible follow-on from test suites. We can follow the same approach of specifying a protocol so that we can add more sophisticated interfaces later as we gain experience wit them.
What do you think about the very basic interface I gave? It basically just makes it possible to create executables that are distinguishable as benchmarks. That seems like a very small step but it solves more or less all of my problems*.
Should I go ahead and implement it?
* I prefer not to use the workaround as having spurious executables confuses Hackage (which lists them) and users.
I've had decent luck with adding a benchmark as a 'test' section, which puts the executable in a predictable place.
Cheers, Johan
_______________________________________________ cabal-devel mailing list cabal-devel@haskell.org http://www.haskell.org/mailman/listinfo/cabal-devel