
I have finally moved the new ghc to arch repos. With this cleanup comes a bit of change. The decision has been made that ghc will be updated to the latest stable version and haskell-platform will be moved to the aur. This means you should remove the haskell-platform related packages before updating. I will also keep only essential packages and their dependencies related to haskell in [extra]. For now this means cabal-install and ghc + their depends. I have closed the following bugs with this release: https://bugs.archlinux.org/task/23247 - [ghc 7.0.2] ghci -package ghc causes panic https://bugs.archlinux.org/task/27132 - [ghc] re-enable the strip makepkg option https://bugs.archlinux.org/task/18917 - [ghc] ghc.install :: test if directory exists https://bugs.archlinux.org/task/28758 - [haskell-text] uninstallable due to removal of haskell-deepseq https://bugs.archlinux.org/task/25991 - haskell-glut: undefined references to pretty much all glut functions https://bugs.archlinux.org/task/25072 - [alex][happy][cabal-install] missing makedepnds The following bugs are still open which I will fix as soon as cabal-install and darcs are buildable: https://bugs.archlinux.org/task/27249 - [cabal-install] Add bash completion support https://bugs.archlinux.org/task/26955 - [darcs] missing makedependence haskell-hashed-storage<0.6 There are still some issues left with the current situation of haskell on arch, but these will take time and communication with upstream. *I need to figure out what to do about haddock since ghc only packages the binary and thus provides only half the package with ghc. This is the reason why we packaged haskell-haddock even though the ghc package already provides the haddock binary. *The post-install messages related to haskell packages can get pretty long. There is a proposed solution with a patch, but I need to figure out if there is any way we can work it out with upstream first. *There is no cabal-install or darcs version that is marked as stable and currently builds with ghc 7.4.1 These packages are statically linked so it shouldn't matter for now. I will be keeping an eye on these packages and waiting for upstream to provide new stable versions that builds against the latest ghc. Cheers!

On Sat, Mar 03, 2012 at 11:35:17AM -0600, Thomas Dziedzic wrote:
I have finally moved the new ghc to arch repos.
Excellent news. I'll try to make sure [haskell] catches up ASAP. [...]
There are still some issues left with the current situation of haskell on arch, but these will take time and communication with upstream. *I need to figure out what to do about haddock since ghc only packages the binary and thus provides only half the package with ghc. This is the reason why we packaged haskell-haddock even though the ghc package already provides the haddock binary.
There should be a release on Hackage of 2.10.0 already.
*The post-install messages related to haskell packages can get pretty long. There is a proposed solution with a patch, but I need to figure out if there is any way we can work it out with upstream first.
This relates to a completely superfluous 'echo' in /usr/share/doc/ghc/html/libraries/gen_contents_index (line 50). Until it's addressed you can get rid of the flooding by just deleting the line locally. /M -- Magnus Therning OpenPGP: 0xAB4DFBA4 email: magnus@therning.org jabber: magnus@therning.org twitter: magthe http://therning.org/magnus Perl is another example of filling a tiny, short-term need, and then being a real problem in the longer term. -- Alan Kay

Great news! I jumped in to try, but I'm having trouble with haskell-gtk. Every piece of code using event handlers segfaults. I have removed every haskell package and installed ghc and haskell-gtk from [extra], then also built haskell-gtk from abs, but with same (crashing result). I don't know if I should report the error upstream, would be great if someone could verify. (For example demo/hello/World.hs which segfaults on my machine). cheers /Johan

I've found the problem (actually, the compiler warned about it). It turns out there is a bug report related to this (http://hackage.haskell.org/trac/gtk2hs/ticket/1256), but that the provided patch isn't enough to fix the problem. I'll alert the gtk2hs team about this. cheers /Johan

On Sun, Mar 04, 2012 at 01:14:11PM +0100, Johan Holmquist wrote:
Great news!
I jumped in to try, but I'm having trouble with haskell-gtk. Every piece of code using event handlers segfaults.
I have removed every haskell package and installed ghc and haskell-gtk from [extra], then also built haskell-gtk from abs, but with same (crashing result).
I don't know if I should report the error upstream, would be great if someone could verify. (For example demo/hello/World.hs which segfaults on my machine).
Have you verified it against my [ghc74] repo? It's quickly done and would at least give you one other data point. /M -- Magnus Therning OpenPGP: 0xAB4DFBA4 email: magnus@therning.org jabber: magnus@therning.org twitter: magthe http://therning.org/magnus Most software today is very much like an Egyptian pyramid with millions of bricks piled on top of each other, with no structural integrity, but just done by brute force and thousands of slaves. -- Alan Kay

The same problem with ghc74 (I did not use [testing] -- only [ghc74],
[extra] and [community]).
You can try this with the haskell-gtk demos, e.g. the hello demo. Make
it, run it and push the button and bang -- Segfault.
/Johan
2012/3/4 Magnus Therning
On Sun, Mar 04, 2012 at 01:14:11PM +0100, Johan Holmquist wrote:
Great news!
I jumped in to try, but I'm having trouble with haskell-gtk. Every piece of code using event handlers segfaults.
I have removed every haskell package and installed ghc and haskell-gtk from [extra], then also built haskell-gtk from abs, but with same (crashing result).
I don't know if I should report the error upstream, would be great if someone could verify. (For example demo/hello/World.hs which segfaults on my machine).
Have you verified it against my [ghc74] repo? It's quickly done and would at least give you one other data point.
/M
-- Magnus Therning OpenPGP: 0xAB4DFBA4 email: magnus@therning.org jabber: magnus@therning.org twitter: magthe http://therning.org/magnus
Most software today is very much like an Egyptian pyramid with millions of bricks piled on top of each other, with no structural integrity, but just done by brute force and thousands of slaves. -- Alan Kay
_______________________________________________ arch-haskell mailing list arch-haskell@haskell.org http://www.haskell.org/mailman/listinfo/arch-haskell

Checked the code at http://code.haskell.org/gtk2hs/glib/System/Glib/hsgclosure.c and it is obviously fixed. I don't know why it didn't work from the community or ghc74 repositories (maybe I did a mistake somewhere). I will try to fetch the source for haskell-glib with abs from [ghc74] when I get home to see if it has the correct (?) code or the faulty one. Still it is (was yesterday) faulty in [community]. I'll check that again and provide a bug report if necessary. cheers /Johan

On Mon, Mar 5, 2012 at 12:51, Johan Holmquist
Checked the code at http://code.haskell.org/gtk2hs/glib/System/Glib/hsgclosure.c and it is obviously fixed. I don't know why it didn't work from the community or ghc74 repositories (maybe I did a mistake somewhere).
Since it looks like you're looking at the upstream source repo it could be that they haven't released a fixed version yet. In putting together [ghc74] I had to make some minor changes to some of the gtk2-related packages in order to make them compile. At the time those changes were already in upstream, but no version with those changes had been released yet.
I will try to fetch the source for haskell-glib with abs from [ghc74] when I get home to see if it has the correct (?) code or the faulty one. Still it is (was yesterday) faulty in [community]. I'll check that again and provide a bug report if necessary.
Both [community] and [ghc74] are built from the versions available on Hackage. I believe both are unpatched, so you can just check the version on Hackage. /M -- Magnus Therning OpenPGP: 0xAB4DFBA4 email: magnus@therning.org jabber: magnus@therning.org twitter: magthe http://therning.org/magnus

On Mon, Mar 5, 2012 at 13:44, Johan Holmquist
2012/3/5 Magnus Therning
: Both [community] and [ghc74] are built from the versions available on Hackage. I believe both are unpatched, so you can just check the version on Hackage.
You're right, glib-0.12.2 on Hackage is not patched. Waiting for that to happen then.
In addition to waiting you can always raise a bug and/or send an email to the author/maintainer of the packages to encourage them to release a new version ASAP... or at least to find out what the hold-up is ;) /M -- Magnus Therning OpenPGP: 0xAB4DFBA4 email: magnus@therning.org jabber: magnus@therning.org twitter: magthe http://therning.org/magnus
participants (3)
-
Johan Holmquist
-
Magnus Therning
-
Thomas Dziedzic