[darcs-users] darcs 2.3.0: compile fails with GHC 6.8.3 on FreeBSD 6.3
Mark Stosberg
mark at summersault.com
Mon Jul 27 20:19:37 UTC 2009
On Mon, 27 Jul 2009 16:11:32 -0400
Mark Stosberg <mark at summersault.com> wrote:
> > The cabal-install program has an annoying feature that if a dependency
> > fails to build, it sometimes still tries to build the original
> > package. So I'm willing to bet that a diagnostic message was output,
> > but was buried in the middle of the darcs build output.
> >
> > Running "cabal install haskeline" should provide a better error
> > message. To get it working, you may need to provide extra arguments
> > to point to iconv's headers/library files, e.g.
> >
> > cabal install haskeline --extra-include-dirs=/usr/include
> > --extra-lib-dirs=/usr/lib
>
> A variation of that worked:
>
> cabal install haskeline --extra-include-dirs=/usr/include --extra-lib-dirs=/usr/local/lib
>
> Perhaps the inclusion of "/usr/local/lib" can be automated on FreeBSD in the future?
>
> Now that I'm past that I'm getting a notice about "curl" being missing,
> which should be straightforward to address. I'll post further notes to
> darcs-users once this compile/install is done.
The compile completed by using this:
cabal install darcs --extra-lib-dirs=/usr/local/lib
It would be nice if "/usr/local/lib" could be automatically used
for the darcs build on FreeBSD, since it's a standard library path
there.
A little more feedback on the "cabal" system:
- The compile should a status of "1 of 136", which I took was like a
progress bar, but then when it got to 136, it started the counting
over again. It actually completed after two round of counting.
- I would like to know if there's a way to run the test suite via cabal,
as that gives me extra confidence that darcs is working as expected on
my platform.
Mark
--
. . . . . . . . . . . . . . . . . . . . . . . . . . .
Mark Stosberg Principal Developer
mark at summersault.com Summersault, LLC
765-939-9301 ext 202 database driven websites
. . . . . http://www.summersault.com/ . . . . . . . .
More information about the darcs-users
mailing list