Re: [Gtk2hs-devel] patch for ghc-6.12

2010-01-21 Thread Andy Stewart
2009/12/11 Duncan Coutts : > Patch to make gtk2hs (darcs HEAD version) build with ghc-6.12. I can't found Duncan's new patch in gtk2hs repository. This patch just works with GHC-6.12? Axel, we should merge Duncan's patch to avoid user failed to compile with GHC-6.12. Cheers, -- Andy --

Re: [Gtk2hs-devel] patch for ghc-6.12

2010-01-18 Thread Bertram Felgenhauer
Axel Simon wrote: > So is this a ghc problem or something that needs to be fixed in Gtk2Hs? Sorry, my previous mail didn't make it to the list (my mistake). The problem is that ghc-6.12 puts split objs into different subdirectories based on the -osuf option: ghc -c T.hs --> T_o_split/ ghc -

Re: [Gtk2hs-devel] patch for ghc-6.12

2010-01-18 Thread Axel Simon
On Jan 18, 2010, at 2:20, Jens Petersen wrote: >> For now, you can try building gtk2hs with --disable-split-objs - >> that appears to work. > > Thanks again Bertram! Looks good now. :-) > > http://koji.fedoraproject.org/koji/taskinfo?taskID=1928857 So is this a ghc problem or something that n

Re: [Gtk2hs-devel] patch for ghc-6.12

2010-01-17 Thread Jens Petersen
> For now, you can try building gtk2hs with  --disable-split-objs - > that appears to work. Thanks again Bertram! Looks good now. :-) http://koji.fedoraproject.org/koji/taskinfo?taskID=1928857 Jens -- Throughout its 18

Re: [Gtk2hs-devel] patch for ghc-6.12

2010-01-16 Thread Jens Petersen
> but something must be wrong since when I try to make > any of the demos I got a host of linking errors: I meant to add: running in ghci-6.12.1 the demos generally startup ok at least but segfault on exit, etc. Jens --

Re: [Gtk2hs-devel] patch for ghc-6.12

2010-01-16 Thread Jens Petersen
2009/12/11 Duncan Coutts : > Patch to make gtk2hs (darcs HEAD version) build with ghc-6.12. > > It should build and install with 6.10 and 6.12, though it's worth other > people checking that before pushing to the main repo. Built just fine for me on fedora with 6.12.1 :) but something must be wron

Re: [Gtk2hs-devel] patch for ghc-6.12

2010-01-14 Thread Jens Petersen
2010/1/14 Bertram Felgenhauer: > No, what happens is that haddock (and only haddock is affected) > enables the 'rec' keyword, because -fglasgow-exts is used, which > now implies -XDoRec. The code in question is Ah thank you - that helped a lot. :-) http://koji.fedoraproject.org/koji/buildinfo?bui

Re: [Gtk2hs-devel] patch for ghc-6.12

2010-01-13 Thread Bertram Felgenhauer
> Did 6.12 change the way layout is handled? The error message seems > quite strange. It's probably easy to fix but it's slightly worrying. No, what happens is that haddock (and only haddock is affected) enables the 'rec' keyword, because -fglasgow-exts is used, which now implies -XDoRec. The co

Re: [Gtk2hs-devel] patch for ghc-6.12

2010-01-13 Thread Axel Simon
On Jan 13, 2010, at 5:30, Jens Petersen wrote: > 2009/12/11 Duncan Coutts : >> Patch to make gtk2hs (darcs HEAD version) build with ghc-6.12. > > Thank you, Duncan! > > I tried it with ghc-6.12.1 on Fedora and when building the docs got to > the attached build error. Did 6.12 change the way layo

Re: [Gtk2hs-devel] patch for ghc-6.12

2010-01-12 Thread Jens Petersen
2009/12/11 Duncan Coutts : > Patch to make gtk2hs (darcs HEAD version) build with ghc-6.12. Thank you, Duncan! I tried it with ghc-6.12.1 on Fedora and when building the docs got to the attached build error. Jens make all-am make[1]: Entering directory `/home/petersen/fedora/haskell/ghc-gtk2hs/

Re: [Gtk2hs-devel] patch for ghc-6.12

2009-12-19 Thread Jürgen Nicklisch-Franken
Hello Duncan, I've waited for this and started to port Leksah. gtk2hs compiles and installs fine, but after linking I got hundreds of undefined references like: (.text+0x6a02): undefined reference to `gtkzm0zi10zi1_GraphicsziUIziGtkziGeneralziEnums_zdfEqPackingzuzeze_info' ... As I've seen this b

[Gtk2hs-devel] patch for ghc-6.12

2009-12-10 Thread Duncan Coutts
Patch to make gtk2hs (darcs HEAD version) build with ghc-6.12. It should build and install with 6.10 and 6.12, though it's worth other people checking that before pushing to the main repo. The patch bundle also includes another random patch (which I can't remember if I sent in before or not). You