Hi Axel,
No, I patched the hsshellscript library distribution to build using
-package-name hsshellscript
as you mentioned. I still got the error. (The documentation
asserts that -package-name is a no-op on unices anyway.
I wonder if that is really still true?)
That's why I said that the error message was confusing: I had
already rebuilt using -package-name and still got an error.
I think there was some module in my test application that
was built by 6.2.1 and deleting its object and interface files
fixed the problem. A slight API breakage from 6.2.1 to 6.2.2
isn't all that surprising.
Best Wishes,
greg
On Dec 1, 2004, at 12:52 PM, Axel Simon wrote:
Hi,
On Wed, 2004-12-01 at 17:36, Gregory Wright wrote:
I cleaned out all of the *.o and *.hi files and rebuilt. Everything
was OK. No doubt there was something stale in one of the *.hi
files, although the error message is perplexing to those of us
who don't channel Simonness ;--)
I came across the same problem today.
crossroads-able> ghc --make -package hsshellscript -o asgraph
asgraph.hs
Chasing modules from: asgraph.hs
Skipping Reformat ( ./Reformat.hs, ./Reformat.o )
Skipping Main ( asgraph.hs, asgraph.o )
Linking ...
ghc-6.2.2: unknown package name: Main
However, building the package (hsshellscript) with the additional
option
-package-name hsshellscript
Files that *use* the package foo keep track of this dependency in the
.hi files. However they track the name that you give with the
-package-name option when compiling the package, not the name given
with
-package when you compile the program. I guess in your case you forgot
to compile the package with -package-name=foo such that the package was
implicitly named Main. Solution: Rebuild the library with the proper
-package-name flag and rebuild everything that uses that package.
Axel.
_______________________________________________
Glasgow-haskell-bugs mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
