You can subscribe to this list here.
2008 |
Jan
|
Feb
(53) |
Mar
(145) |
Apr
(22) |
May
(7) |
Jun
(14) |
Jul
(14) |
Aug
(9) |
Sep
(10) |
Oct
(48) |
Nov
(59) |
Dec
(45) |
---|---|---|---|---|---|---|---|---|---|---|---|---|
2009 |
Jan
(36) |
Feb
(5) |
Mar
(33) |
Apr
(28) |
May
(5) |
Jun
(6) |
Jul
(1) |
Aug
(7) |
Sep
(11) |
Oct
(3) |
Nov
(31) |
Dec
|
2010 |
Jan
(8) |
Feb
(3) |
Mar
|
Apr
(2) |
May
(9) |
Jun
(1) |
Jul
(2) |
Aug
|
Sep
(1) |
Oct
(1) |
Nov
|
Dec
|
2011 |
Jan
(1) |
Feb
(3) |
Mar
(4) |
Apr
(1) |
May
(2) |
Jun
(12) |
Jul
(36) |
Aug
(7) |
Sep
(40) |
Oct
(6) |
Nov
(40) |
Dec
(8) |
2012 |
Jan
(54) |
Feb
(8) |
Mar
(1) |
Apr
(16) |
May
(2) |
Jun
(12) |
Jul
(1) |
Aug
(1) |
Sep
(2) |
Oct
(2) |
Nov
|
Dec
(7) |
2013 |
Jan
(8) |
Feb
|
Mar
(13) |
Apr
(2) |
May
(13) |
Jun
(44) |
Jul
|
Aug
(13) |
Sep
(12) |
Oct
(11) |
Nov
(7) |
Dec
(6) |
2014 |
Jan
(3) |
Feb
(4) |
Mar
(9) |
Apr
(1) |
May
|
Jun
(3) |
Jul
(1) |
Aug
(1) |
Sep
(1) |
Oct
|
Nov
|
Dec
|
2015 |
Jan
|
Feb
(2) |
Mar
|
Apr
|
May
(2) |
Jun
(3) |
Jul
|
Aug
(2) |
Sep
(5) |
Oct
(2) |
Nov
(1) |
Dec
(1) |
2017 |
Jan
|
Feb
|
Mar
|
Apr
(1) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(1) |
Dec
|
2020 |
Jan
(1) |
Feb
|
Mar
|
Apr
(1) |
May
(1) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
From: Eric K. <eri...@gm...> - 2008-03-05 19:53:48
|
> The problem is that the module imports Data.Time, which do not exist in > GHC 6.3.3. If I uncommit the patch everything compiles. And the examples > in samples/wx runs - the two I tried anyway. Would it be fair to ask people to install the time package? http://hackage.haskell.org/cgi-bin/hackage-scripts/package/time-1.1.2.0 If not, we could conceivably have a ghc-6.6 branch without the patch. I think 6.6 is pretty common (Ubuntu Gutsy uses it), although I personally have given up on using the GHC that comes with my system, and have just gone with whatever packages the GHC folks themselves provide. -- Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow> PGP Key ID: 08AC04F9 |
From: Mads L. <mad...@ya...> - 2008-03-05 19:43:23
|
Hi all I've been trying to compile WxHaskell with GHC 6.6.3, as this is the stable version in Debian. I know I should have done this regularly, but I have unfortunately been getting into the bad habit of only compiling with GHC 6.8.1 and GHC 6.8.2. First of all, I assume that if people want to compile with old versions of GHC, they will have to live with using the Makefile version. E.g.: chmod 755 configure ../configure --with-opengl --prefix=/usr/local/ make .... Luckily, it all went nicely except for this patch: "Sun Aug 19 22:42:30 CEST 2007 shelarcy <she...@gm...> * Use Data.Time.getCurrentTime instead of (System.)Time.getClockTime to avoid build failure by generated non-UTF8 String" The problem is that the module imports Data.Time, which do not exist in GHC 6.3.3. If I uncommit the patch everything compiles. And the examples in samples/wx runs - the two I tried anyway. Anyway, my question is, how do we remove this dependency on Data.Time for GHC 6.6 only, without forcing GHC 6.8 users to also not use Data.Time? Also, do other people compile with GHC 6.6? Or is it just me who have to, because Debian's is so out of date. Greetings, Mads Lindstrøm |
From: shelarcy <she...@gm...> - 2008-03-05 18:12:09
|
On Thu, 06 Mar 2008 02:20:29 +0900, Eric Kow <eri...@gm...> wrote: > Is there a good > reason for wx to be calling wxcore-clean? I forgot about that. I'll check that morning or afternoon. (I'm sleepy now.) > By the way, I just noticed why you originally suggested waiting for > wxWidgets 2.8 support (because 2.6 can't build on Leopard). Not sure > what to do about it. Some online forums suggest compiling it with the > 10.4 SDK I don't know about that. How about add that information and link to installation guide? I think that would be FAQ. Best Regards, -- shelarcy <shelarcy hotmail.co.jp> http://page.freett.com/shelarcy/ |
From: Eric K. <eri...@gm...> - 2008-03-05 18:01:42
|
Tue Mar 4 22:43:27 GMT 2008 Eric Kow <eri...@gm...> * Bump to 0.10.3. Wed Mar 5 10:32:28 GMT 2008 Eric Kow <eri...@gm...> * Add vim modeline for makefile.lib. It's really irritating otherwise; vim thinks it's COBOL. Wed Mar 5 11:14:15 GMT 2008 Eric Kow <eri...@gm...> * Don't do wildcard expansion in cp-echo. It seems to expand non-wildcard strings like into the empty string. Not sure if this breaks anything else, though. Wed Mar 5 11:23:24 GMT 2008 Eric Kow <eri...@gm...> * Fix dep: wx-bindist requires wx Wed Mar 5 11:24:01 GMT 2008 Eric Kow <eri...@gm...> * Withdraw haddock support for GHC < 6.2. Portability is nice and all, but this is getting hard to maintain. Wed Mar 5 11:26:39 GMT 2008 Eric Kow <eri...@gm...> * Update URL for haddock interface files. Wed Mar 5 11:27:20 GMT 2008 Eric Kow <eri...@gm...> * (configure/haddock) Set normdocroot on non-Cygwin. Wed Mar 5 11:38:37 GMT 2008 Eric Kow <eri...@gm...> * Also take advantage of -split-objs for wx. Wed Mar 5 13:48:26 GMT 2008 Eric Kow <eri...@gm...> * Update hdocbases to work with haddock2 and ghc 6.8.2 I'm not sure which, but this makes it compile on my end. Wed Mar 5 13:52:06 GMT 2008 Eric Kow <eri...@gm...> * Update list of developers (homepage). Wed Mar 5 13:52:27 GMT 2008 Eric Kow <eri...@gm...> * Update repository info (now using darcs, not CVS). (homepage) Wed Mar 5 13:54:36 GMT 2008 Eric Kow <eri...@gm...> * Remove scintilla from wishlist (homepage). Already implemented as far as I know. Wed Mar 5 14:18:06 GMT 2008 Eric Kow <eri...@gm...> * Add bugs to the dist files. We want people to know we're aware of them, and we want them to get the idea of writing bug reports with demonstrators. Wed Mar 5 14:18:41 GMT 2008 Eric Kow <eri...@gm...> * Add XRC as a wishlist task (homepage). Wed Mar 5 14:20:57 GMT 2008 Eric Kow <eri...@gm...> * Last modified (homepage). Wed Mar 5 17:22:00 GMT 2008 Eric Kow <eri...@gm...> * More portable haddocking for Windows. As suggested by shelarcy. Wed Mar 5 17:35:55 GMT 2008 Eric Kow <eri...@gm...> * Don't call wxcore-clean when building wx. Also, build wxcore in its own directory, so we don't get overlapping with wx. (I'm guessing that the overlap is why we called wxcore-clean) Wed Mar 5 17:38:25 GMT 2008 Eric Kow <eri...@gm...> * Fix mispelling of Mads's family name. Sorry! |
From: Eric K. <eri...@gm...> - 2008-03-05 17:20:43
|
Hi, > It seems that your patch break bindist. Argh :-) > ifdef ENABLE-PROF > wx: wx-main-only wx-prof > else > wx: wx-main-only > endif > wx-main-only: wxcore-clean wx-main Huh. So that's why it kept recompiling wxcore. Is there a good reason for wx to be calling wxcore-clean? > $docroot isn't in share directory on Windows platform. > How about change following in this part? > > if test "$cygpathfound"; then > docroot="$ghcroot/doc/libraries" > else > docroot="$ghcroot/share/doc/ghc/libraries" > fi > ;; Fair enough. By the way, I just noticed why you originally suggested waiting for wxWidgets 2.8 support (because 2.6 can't build on Leopard). Not sure what to do about it. Some online forums suggest compiling it with the 10.4 SDK -- Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow> PGP Key ID: 08AC04F9 |
From: shelarcy <she...@gm...> - 2008-03-05 17:10:09
|
Hi Eric, On Wed, 05 Mar 2008 23:22:48 +0900, Eric Kow <eri...@gm...> wrote: > More stuff related to a 0.10.3 release candidate. Basically fixes to 'make > bindist' on my end, and some documentation. Now we need people to make > packages for their platforms, test them out, etc. It seems that your patch break bindist. > [Fix dep: wx-bindist requires wx > Eric Kow <eri...@gm...>**20080305112324] hunk ./makefile 698 > @$(call cp-srcdist, $^) > = > > # bindist > -wx-bindist: = > > +wx-bindist: wx > @$(call cp-bindist,$(WX-OUTDIR),$(BINDIST-LIBDIR),$(WX-BINS)) > ifdef ENABLE-PROF > @$(call cp-bindist,$(WX-OUTDIR),$(BINDIST-LIBDIR),$(WX-PROF-BINS)) bindist includes wxcore-bindist and wx-bindist. And "make wx" removes wxcore file. ifdef ENABLE-PROF wx: wx-main-only wx-prof else wx: wx-main-only endif wx-main-only: wxcore-clean wx-main So, now "make bindist" removes wxcore file before generating bindist. It is unexpected behaviour. And I know one more problem on Windows. > [Update hdocbases to work with haddock2 and ghc 6.8.2 > Eric Kow <eri...@gm...>**20080305134826 > I'm not sure which, but this makes it compile on my end. > ] { > > (snip) > > # Find .haddock files > if test "$hdocfound" =3D "yes"; then > hunk ./configure 683 > 6.0* | 6.1* | 6.2* | 6.3* | 6.4* | 6.5* | 6.6*) > docroot=3D"$ghcroot/doc/html/libraries";; > 6.*) = > > - docroot=3D"$ghcroot/doc/libraries";; > + docroot=3D"$ghcroot/share/doc/ghc/libraries";; > *) > docroot=3D"$ghcroot/doc/libraries";; > esac $docroot isn't in share directory on Windows platform. How about change following in this part? if test "$cygpathfound"; then docroot="$ghcroot/doc/libraries" else docroot="$ghcroot/share/doc/ghc/libraries" fi ;; -- shelarcy <shelarcy hotmail.co.jp> http://page.freett.com/shelarcy/ |
From: Eric K. <eri...@gm...> - 2008-03-05 14:23:02
|
More stuff related to a 0.10.3 release candidate. Basically fixes to 'make bindist' on my end, and some documentation. Now we need people to make packages for their platforms, test them out, etc. Tue Mar 4 22:43:27 GMT 2008 Eric Kow <eri...@gm...> * Bump to 0.10.3. Wed Mar 5 10:32:28 GMT 2008 Eric Kow <eri...@gm...> * Add vim modeline for makefile.lib. It's really irritating otherwise; vim thinks it's COBOL. Wed Mar 5 11:14:15 GMT 2008 Eric Kow <eri...@gm...> * Don't do wildcard expansion in cp-echo. It seems to expand non-wildcard strings like into the empty string. Not sure if this breaks anything else, though. Wed Mar 5 11:23:24 GMT 2008 Eric Kow <eri...@gm...> * Fix dep: wx-bindist requires wx Wed Mar 5 11:24:01 GMT 2008 Eric Kow <eri...@gm...> * Withdraw haddock support for GHC < 6.2. Portability is nice and all, but this is getting hard to maintain. Wed Mar 5 11:26:39 GMT 2008 Eric Kow <eri...@gm...> * Update URL for haddock interface files. Wed Mar 5 11:27:20 GMT 2008 Eric Kow <eri...@gm...> * (configure/haddock) Set normdocroot on non-Cygwin. Wed Mar 5 11:38:37 GMT 2008 Eric Kow <eri...@gm...> * Also take advantage of -split-objs for wx. Wed Mar 5 13:48:26 GMT 2008 Eric Kow <eri...@gm...> * Update hdocbases to work with haddock2 and ghc 6.8.2 I'm not sure which, but this makes it compile on my end. Wed Mar 5 13:52:06 GMT 2008 Eric Kow <eri...@gm...> * Update list of developers (homepage). Wed Mar 5 13:52:27 GMT 2008 Eric Kow <eri...@gm...> * Update repository info (now using darcs, not CVS). (homepage) Wed Mar 5 13:54:36 GMT 2008 Eric Kow <eri...@gm...> * Remove scintilla from wishlist (homepage). Already implemented as far as I know. Wed Mar 5 14:18:06 GMT 2008 Eric Kow <eri...@gm...> * Add bugs to the dist files. We want people to know we're aware of them, and we want them to get the idea of writing bug reports with demonstrators. Wed Mar 5 14:18:41 GMT 2008 Eric Kow <eri...@gm...> * Add XRC as a wishlist task (homepage). Wed Mar 5 14:20:57 GMT 2008 Eric Kow <eri...@gm...> * Last modified (homepage). |
From: shelarcy <she...@gm...> - 2008-03-05 12:26:29
|
I checked that now this patch also works on Windows. Thank you for your work. On Wed, 05 Mar 2008 19:49:30 +0900, Eric Kow <eri...@gm...> wrote: > I hope this works better... > > Sun Mar 2 23:37:10 GMT 2008 Eric Kow <eri...@gm...> > * Create a --enable-split-objs option for configure and makefile. > For wxcore, --enable-split-objs calls GHC with -split-objs and > combines the splitted up .o files in the archive. > > (snip) > > Tue Mar 4 22:35:24 GMT 2008 Eric Kow <eri...@gm...> > * Back off from using split-objs for Graphics.UI.WXCore.WxcTypes > I don't really understand why we have to do this, but I get undefined > symbol errors linking sample apps when using this with --enable-split-objs > > Wed Feb 27 22:31:52 GMT 2008 Eric Kow <eri...@gm...> > * Add an --hcflags option to the configure script. > For --split-objs and friends. > > (snip) > > Wed Mar 5 10:48:56 GMT 2008 Eric Kow <eri...@gm...> > * Use xargs to feed split objs to $(AR) 25 at a time. -- shelarcy <shelarcy hotmail.co.jp> http://page.freett.com/shelarcy/ |
From: Eric K. <eri...@gm...> - 2008-03-05 10:49:40
|
I hope this works better... Sun Mar 2 23:37:10 GMT 2008 Eric Kow <eri...@gm...> * Create a --enable-split-objs option for configure and makefile. For wxcore, --enable-split-objs calls GHC with -split-objs and combines the splitted up .o files in the archive. Tue Mar 4 15:40:47 GMT 2008 Eric Kow <eri...@gm...> * Add a wxcREFUSE_MEDIACTRL flag (for the same reason as opengl) Tue Mar 4 17:07:29 GMT 2008 Eric Kow <eri...@gm...> * Fix configure bug where ENABLE-SPLITOBJS was being defined. It was being set to 'no' (perhaps the makefile should just check if ENABLE-SPLITOBJS='yes') Tue Mar 4 22:35:05 GMT 2008 Eric Kow <eri...@gm...> * Warn about GHC older than 6.4 (to match removed functionality). Tue Mar 4 22:35:24 GMT 2008 Eric Kow <eri...@gm...> * Back off from using split-objs for Graphics.UI.WXCore.WxcTypes I don't really understand why we have to do this, but I get undefined symbol errors linking sample apps when using this with --enable-split-objs Wed Feb 27 22:31:52 GMT 2008 Eric Kow <eri...@gm...> * Add an --hcflags option to the configure script. For --split-objs and friends. Tue Mar 4 22:43:27 GMT 2008 Eric Kow <eri...@gm...> * Bump to 0.10.3. Wed Mar 5 10:32:28 GMT 2008 Eric Kow <eri...@gm...> * Add vim modeline for makefile.lib. It's really irritating otherwise; vim thinks it's COBOL. Wed Mar 5 10:48:56 GMT 2008 Eric Kow <eri...@gm...> * Use xargs to feed split objs to $(AR) 25 at a time. |
From: shelarcy <she...@gm...> - 2008-03-05 03:40:58
|
On Wed, 05 Mar 2008 07:38:26 +0900, Eric Kow <eri...@gm...> wrote: > Using --enable-split-objs, I've been able to reduce the size of the UTFSampler > binary from 3M stripped to 1M stripped. And it still seems to work. > If somebody could explain to me why we need the last patch, I would be even > happier. Your current patch uses $hcflag variable. On Wed, 05 Mar 2008 07:38:26 +0900, Eric Kow <eri...@gm...> wrote: > [Create a --enable-split-objs option for configure and makefile. > Eric Kow <eri...@gm...>**20080302233710 > > (snip) > > +if test "$enablesplitobjs" =3D "yes"; then > + hcflags=3D"$hcflags -split-objs" But you forgot to use this variable in HCFLAGS. > HCBASENAME=$hcbasename > HCFLAGS= So we need last patch's below part. > [Add an --hcflags option to the configure script. > Eric Kow <eri...@gm...>**20080227223152 > For --split-objs and friends. > ] { > (snip) > hunk ./configure 825 > HCVERSION=$hcversion > HCNAME=$hcname > HCBASENAME=$hcbasename > -HCFLAGS= > +HCFLAGS=$hcflags > HCPKG=$hcpkg > ENABLE-PROF=$enableprof > HC-PROF-FLAGS=-hisuf p_hi $hcprof > } And I meet another problem in your patch. $ runhaskell Setup.lhs build (snip) c:\ghc\ghc-6.8.2\bin\ar.exe: creating dist/wx/libwxcore0.a ar -sr dist/wx/libwxcore0.a dist/wx/imports/Graphics/UI/WXCore/WxcObject_split/* .o && ar -sr dist/wx/libwxcore0.a dist/wx/imports/Graphics/UI/WXCore/WxcClassTy pes_split/*.o && : /bin/sh: /cygdrive/c/ghc/ghc-6.8.2/bin/ar: Argument list too long make: *** [dist/wx/libwxcore0.a] Error 126 Build Failed. Windows' ar can't use long argument. And GHC uses xargs to avoid that problem. http://hackage.haskell.org/trac/ghc/wiki/Building/Windows#Aregumentlisttoolong http://darcs.haskell.org/ghc-6.8/ghc/mk/target.mk Could you use xargs to link split-objs? Best Regards, -- shelarcy <shelarcy hotmail.co.jp> http://page.freett.com/shelarcy/ |
From: shelarcy <she...@gm...> - 2008-03-05 03:16:51
|
Hi Eric, On Mon, 03 Mar 2008 08:42:46 +0900, Eric Y. Kow <eri...@gm...> wrote: >> If we want to do it, to use -split-objs for spliting the single object file, >> to use -f* options (I don't know this is good for wxHaskell or not), or to >> use strip for Haskell program. > > Well, it seems like using strip is pretty simple advice, and quite > helpful in that it shrank the UTF8Sampler from 8M to 4M. But I suppose > -split-objs would really make the biggest difference? Windows doesn't have strip in defalt. He must install MinGW or installing strip by hand for to use strip. On the other hand, ghc always has -split-obj option in any platform. Of cource, that can use Windows platfrom too. So I think -split-obj is better than strip. On Mon, 03 Mar 2008 08:42:46 +0900, Eric Y. Kow <eri...@gm...> wrote: >> (OutputDir)/(ModuleName)_split directory has ld.script that describe just INPUT(...). >> So, I think we must pass that per modules. > > (snip) > > The problem is that I don't really know what we do with them. The one > difference I noticed (and that I pounced on) is that it is these files > that get put into the libfoo.a archive files. I'm guessing that this is > enough because when you pass in a whole bunch of object files as > arguments to a linker, the linker is smart enough to ignore any files > that are not used at all? > > Also, I'm guessing that the ld.script stuff is GNU-specific (we don't > get them on MacOS X), and that it isn't even neccessary for GNU (it's > just an alternative to having a very long list of command line > arguments). I see. I don't know why ghc generate ld.script. But I find an problem by linking whole object by ar on Windows platform. I report that in next mail. Best Regards, -- shelarcy <shelarcy hotmail.co.jp> http://page.freett.com/shelarcy/ |
From: Eric K. <eri...@gm...> - 2008-03-04 22:47:36
|
Folks, With the recent --enable-split-objs patch, I think we have met our criteria for releasing 0.10.x (assuming it actually works). Shall we try to prepare some binaries for this weekend? Volunteers for your respective platforms? What would also be interesting is updates to the webpage, documentation, etc, generally tidying up. I propose we call the next set of binaries 0.10.3rc1 [On a side note, I think 1.0 should come after (maybe right after) 0.11, basically, after the dust settles a bit] Tue Mar 4 22:43:27 GMT 2008 Eric Kow <eri...@gm...> * Bump to 0.10.3. |
From: Eric K. <eri...@gm...> - 2008-03-04 22:38:33
|
Using --enable-split-objs, I've been able to reduce the size of the UTFSampler binary from 3M stripped to 1M stripped. And it still seems to work. If somebody could explain to me why we need the last patch, I would be even happier. Sun Mar 2 23:37:10 GMT 2008 Eric Kow <eri...@gm...> * Create a --enable-split-objs option for configure and makefile. For wxcore, --enable-split-objs calls GHC with -split-objs and combines the splitted up .o files in the archive. Tue Mar 4 15:40:47 GMT 2008 Eric Kow <eri...@gm...> * Add a wxcREFUSE_MEDIACTRL flag (for the same reason as opengl) Tue Mar 4 17:07:29 GMT 2008 Eric Kow <eri...@gm...> * Fix configure bug where ENABLE-SPLITOBJS was being defined. It was being set to 'no' (perhaps the makefile should just check if ENABLE-SPLITOBJS='yes') Tue Mar 4 22:35:05 GMT 2008 Eric Kow <eri...@gm...> * Warn about GHC older than 6.4 (to match removed functionality). Tue Mar 4 22:35:24 GMT 2008 Eric Kow <eri...@gm...> * Back off from using split-objs for Graphics.UI.WXCore.WxcTypes I don't really understand why we have to do this, but I get undefined symbol errors linking sample apps when using this with --enable-split-objs |
From: Eric Y. K. <eri...@gm...> - 2008-03-04 12:32:08
|
> Tue Mar 4 07:01:18 CET 2008 Mads Lindstroem <mad...@ya...> > * Added testcase for FontFixed problem (1906476) Thanks! Pushed. -- Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow> PGP Key ID: 08AC04F9 |
From: Mads L. <mad...@ya...> - 2008-03-04 06:05:47
|
Tue Mar 4 07:01:18 CET 2008 Mads Lindstroem <mad...@ya...> * Added testcase for FontFixed problem (1906476) |
From: Mads L. <mad...@ya...> - 2008-03-03 19:48:43
|
Hi It seems that I put a wrong bug-number in the patch-comment. Sorry :( The real number is 1906476. Greetings, Mads Lindstrøm Mads Lindstroem wrote: > Mon Mar 3 20:38:20 CET 2008 Mads Lindstroem <mad...@ya...> > * Added testcase for FontFixed problem (1179394) > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ |
From: Mads L. <mad...@ya...> - 2008-03-03 19:42:57
|
Mon Mar 3 20:38:20 CET 2008 Mads Lindstroem <mad...@ya...> * Added testcase for FontFixed problem (1179394) |
From: SourceForge.net <no...@so...> - 2008-03-03 19:40:13
|
Bugs item #1906476, was opened at 2008-03-03 20:40 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=536845&aid=1906476&group_id=73133 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None Status: Open Resolution: None Priority: 5 Private: No Submitted By: Mads Lindstrupermule) Assigned to: Mads Lindstrupermule) Summary: textCtrl monospace font is ignored for user input Initial Comment: Setting a textCtrl's font to fontFixed works fine for text set with "set textCtrl [ text := "bla bla" ]". But fontFixed is ignored when the GUI-user writes into the textCtrl. See wxhaskell/bugs/FontFixed.hs if you want to reproduce the bug. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=536845&aid=1906476&group_id=73133 |
From: Eric Y. K. <eri...@gm...> - 2008-03-02 23:43:09
|
On Thu, Feb 28, 2008 at 15:49:45 +0900, shelarcy wrote: > We must tell the linker to link splited objects. > -split-objs option generates it in (OutputDir)/(ModuleName)_split directory per moudles. > You can see that your own directory or below cabal code. > > http://www.haskell.org/ghc/docs/latest/html/libraries/Cabal/src/Distribution-Simple-GHC.html#getHaskellObjects > > (OutputDir)/(ModuleName)_split directory has ld.script that describe just INPUT(...). > So, I think we must pass that per modules. I've had a look at this, and I confess that I'm pretty lost. I see that passing -split-objs to GHC causes it to create many tiny .o files per module (one per top-level function, says the manual), in addition to the one .o file for that module. The problem is that I don't really know what we do with them. The one difference I noticed (and that I pounced on) is that it is these files that get put into the libfoo.a archive files. I'm guessing that this is enough because when you pass in a whole bunch of object files as arguments to a linker, the linker is smart enough to ignore any files that are not used at all? Also, I'm guessing that the ld.script stuff is GNU-specific (we don't get them on MacOS X), and that it isn't even neccessary for GNU (it's just an alternative to having a very long list of command line arguments). Anyway, my latest patch shows how I currently understand things. I hope it's progress! > If we want to do it, to use -split-objs for spliting the single object file, > to use -f* options (I don't know this is good for wxHaskell or not), or to > use strip for Haskell program. Well, it seems like using strip is pretty simple advice, and quite helpful in that it shrank the UTF8Sampler from 8M to 4M. But I suppose -split-objs would really make the biggest difference? -- Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow> PGP Key ID: 08AC04F9 |
From: Eric K. <eri...@gm...> - 2008-03-02 23:39:38
|
I'm afraid I don't really understand all this compiler stuff. Does this do what we want? Sun Mar 2 23:37:10 GMT 2008 Eric Kow <eri...@gm...> * Create a --enable-split-objs option for configure and makefile. For wxcore, --enable-split-objs calls GHC with -split-objs and combines the splitted up .o files in the archive. |
From: Eric Y. K. <eri...@gm...> - 2008-02-29 00:28:49
|
Pushing these in. (although I wonder if it would have been useful to turn the ListControl thing into an automated regression test; maybe there is a way to generalise it, like setting a value in a widget and getting it back should return the same thing) -- Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow> PGP Key ID: 08AC04F9 |
From: shelarcy <she...@gm...> - 2008-02-28 08:30:38
|
DarcsURL: C:/home/shelarcy/wxhaskell MIME-Version: 1.0 Content-Type: multipart/mixed; boundary="=_" --=_ Content-Type: text/plain Content-Transfer-Encoding: quoted-printable Thu Feb 28 17:16:14 =93=8C=8B=9E (=95W=8F=80=8E=9E) 2008 shelarcy <shelarc= y...@gm...> * Remove List Control test case (bug 1742979 is already fixed on darcs). Thu Feb 28 17:26:49 =93=8C=8B=9E (=95W=8F=80=8E=9E) 2008 shelarcy <shelarc= y...@gm...> * Fix: bugs/makefile's "make clean" doesn't clean Windows things. --=_ Content-Type: text/x-darcs-patch; name="remove-list-control-test-case-_bug-1742979-is-already-fixed-on-darcs__.dpatch" Content-Transfer-Encoding: quoted-printable Content-Description: A darcs patch for your repository! New patches: [Remove List Control test case (bug 1742979 is already fixed on darcs). shelarcy <she...@gm...>**20080228081614] { hunk ./bugs/ListControl.hs 1 -import Graphics.UI.WX - -bugtext =3D unlines - [ "Bug: the contents of this listCtrl ('Aa' and 'Bb') should appear" - , "in the following label, but they do not:"] - -main :: IO () -main =3D start gui - -gui - =3D do f <- frame [text :=3D "Test"] - pa <- panel f [] - lc <- listCtrl pa [columns :=3D [("Col1", AlignLeft, 100), ("Col2",= AlignLeft, 100)], - items :=3D [["Aa", "Bb"]] - ] - i <- get lc items - let l =3D unlines $ [ bugtext, "----------8<-----------" ] - ++ map unwords i - ++ [ "----------8<-----------" ] - set f [layout :=3D container pa $ column 5 [ fill $ widget lc , lab= el l ] ] rmfile ./bugs/ListControl.hs hunk ./bugs/makefile 10 - ListControl\ } [Fix: bugs/makefile's "make clean" doesn't clean Windows things. shelarcy <she...@gm...>**20080228082649] { hunk ./bugs/makefile 24 - rm -rf $(APPS) *.o *.hi *.app + rm -rf $(APPS) *.o *.hi *.app *exe *.manifest } Context: [Remove parasite configure change. Eric Kow <eri...@gm...>**20080227134757 Move around some variables to make more sense. ] = [Add wx-prof target for Windows binary distribution. shelarcy <she...@gm...>**20080227035051] = [Fix bugs in enable wxcore profiling. shelarcy <she...@gm...>**20080227034845] = [Specify what OS BoxedCombinator bug fails on. Eric Kow <eri...@gm...>**20080227135140] = [Fix bugs makefile under Linux Eric Kow <eri...@gm...>**20080227134848] = [Add a configure option to enable wxcore profiling. Eric Kow <eri...@gm...>**20080226215155] = [Fix some bugs pointed out by shelarcy on wxcore-prof. Eric Kow <eri...@gm...>**20080226211227] = [Add wxcore-prof target for compiling profiling versions of wxcore. Eric Kow <eri...@gm...>**20080225231635] = [Remove support for GHC prior to 6.4. Eric Kow <eri...@gm...>**20080225223554] = [Maintainer is wxhaskell-devel not wxhaskell-users. Eric Kow <eri...@gm...>**20080225003139] = [Change samples/contrib/Camels.hs encoding to UTF-8. shelarcy <she...@gm...>**20080225142551] = [Fix: docdist and bindist's wx documents don't link wxcore documents. shelarcy <she...@gm...>**20080225141105] = [Fix: configure doesn't find GHC's Haddock document directory on GHC 6.8.x = or higher. shelarcy <she...@gm...>**20080225135754] = [Fix bug pointed out by shelarcy in --enable-mediactrl flag. Eric Kow <eri...@gm...>**20080225110340 It was setting the opengl flag instead. ] = [Add a test case for start >> start (1610984) Eric Kow <eri...@gm...>**20080225001328] = [Implement an --enable-mediactrl configure flag. Eric Kow <eri...@gm...>**20080225000109 with mediactrl disabled by default. ] = [Add a diagram of wxhaskell components. Eric Kow <eri...@gm...>**20080224185557] = [(OS X) Skip intermediate step of compiling master.o (revisited!) Eric Kow <eri...@gm...>**20080218225200 = I'm going to quote some mails from Malcolm Wallace on wxhaskel-users in ea= rly 2008-02. = First message: > I'm attempting to install wxHaskell from the darcs repo at > darcs.haskell.org, under ghc-6.6. on MacOS 10.4.11, with > wxMac-2.6.4. > > The compilation of wxHaskell seems to proceed OK until the link > stage: > > g++ -r -keep_private_externs -nostdlib -o out/wxc/master.o .... > > which failed because there is no -lstdc++-static on my machine. I > found the place in the makefile to replace -lstdc++-static with > the dynamic -lstdc++, and compilation proceeds further, but not > much: > > g++ -dynamiclib -install_name\ > /usr/local/wxhaskell/lib/libwxc-mac2.6.4-0.10.1.dylib -undefined suppres= s\ > -flat_namespace -o out/wxc/libwxc-mac2.6.4-0.10.1.dylib out/wxc/master.o= \ > -lwx_macu_gl-2.6 -L/usr/local/lib -framework QuickTime -framework IOKit\ > -framework Carbon -framework Cocoa -framework System -lwx_macu_media-2.6= \ > -lwx_macu-2.6 > ld: out/wxc/master.o undefined symbol 12387 (__ZTI10wxListBase) can'= t be a weak definition > /usr/bin/libtool: internal link edit command failed > > The offending symbol __ZTI10wxListBase looks like it might be a > z-encoded symbol from Haskell-land, but I'm not sure how to proceed. = Second message: > A correct build on MacOS 10.4.11 with ghc-6.6 requires that > wxWidgets is configured with --enable-shared, and that this patch: > > (OS X) Check architecture to use intermediate step of compiling > master.o or not. > > be darcs unpulled from your local repo. > > With these tweaks, wxHaskell builds, compiles, and runs the demo > programs. = ] = [Eliminate mandatory --with-opengl on Linux. Eric Kow <eri...@gm...>**20080219171912 = If wxWidgets is compiled with opengl, its header files #define wxUSE_GLCANVAS 1 = (See, for example, /usr/lib/wx/include/gtk2-unicode-release-2.6/wx/setup.h= on Ubuntu Gutsy Gibbon), which causes us to compile glcanvas as if we were go= ing to link against the wxWidgets opengl library. But since we are not linkin= g against that, the user gets errors compiling their applications. = Here we introduce an extra preprocessor flag to really insist that no, des= pite what wxWidgets says, we don't want to use GLCANVAS. ] = [Fix: wxPrint**, wxPreviveFrame** and other function causes link error by u= ndefined reference on Windows (Visual Studio). shelarcy <she...@gm...>**20080219034144] = [Cleanup passing of std/media/stc/opengl to wx-config. Eric Kow <eri...@gm...>**20080218130345] = [Export TabPage type synonym (bug 1349475) Eric Kow <eri...@gm...>**20080218005112] = [Add ListControl test case (bug 1742979) Eric Kow <eri...@gm...>**20080218002441 This might only be a bug on Windows, if at all. ] = [Add test case textColor attribute (bug 1224727). Eric Kow <eri...@gm...>**20080218000738] = [Fix Landscape printing (bug 1168903). Eric Kow <eri...@gm...>**20080217235724 = Filed on behalf on Lennart Augustson <le...@au...>, who submitted a regular Unix patch. ] = [Add BoxedCombinator test case (bug 1549363) Eric Kow <eri...@gm...>**20080217233719] = [Add Sound type's Media class stop method. shelarcy <she...@gm...>**20080218160511] = [Fix STC* test cases. shelarcy <she...@gm...>**20080218155734] = [Fix: wrapper.h and sound.cpp doesn't care about wxWindows 2.4.2. shelarcy <she...@gm...>**20080218155605] = [Sync Visual Studio Project's source flatten away ewxw directories. shelarcy <she...@gm...>**20080218142201] = [Add the submenu bug. Eric Kow <eri...@gm...>**20080217174326] = [Add NonModalDialog test case (bug #1372529) Eric Kow <eri...@gm...>**20080217174203] = [Add a bugs directory for test cases. Eric Kow <eri...@gm...>**20080217174142] = [Fix UTF8Sampler test case. Eric Kow <eri...@gm...>**20080217130553] = [Fix typo pointed out on bugtracker. Eric Kow <eri...@gm...>**20080217114303] = [dos2unix files that have both CRLF and LF line terminators Eric Kow <eri...@gm...>**20080217095711 = Text editors get confused by these. Get rid of trailing whitespace while we're at it. ] = [Flatten away ewxw directories. Eric Kow <eri...@gm...>**20080217094929 = This is to mirror the wxc project, making it easier to compare our version of wxc with the standalone one. ] = [Remove unused ewxw files. Eric Kow <eri...@gm...>**20080217082029 = There is an overlap between ./wxc/src/ewxw and ./wxc/src. The makefile se= ems to ignore the ones in ewxw. ] = [Synch Visual Studio Project's wxc library version shelarcy <she...@gm...>**20080216161420] = [Synch Visual Studio Project's output dir same as makefile shelarcy <she...@gm...>**20080216155656] = [Ugly hack to make setup haddock work without first building. Eric Kow <eri...@gm...>**20080216135157] = [Bump to 0.10.2. Eric Kow <eri...@gm...>**20080216130854] = [Mimick cabal in placement of haddocks. Eric Kow <eri...@gm...>**20080216130819] = [Move cabal stuff to end of makefile. Eric Kow <eri...@gm...>**20080216125445 so that all variables are defined. ] = [Add Hs-Source-Dirs for wxcore.cabal. Eric Kow <eri...@gm...>**20080216125112 In case we ever switch to the Simple build type. ] = [Re-add time package dependency (needed by wxdirect). Eric Kow <eri...@gm...>**20080216124853] = [Improve description and synopsis of cabal files. Eric Kow <eri...@gm...>**20080216124832] = [Split doc building into wxcore and wx version. Eric Kow <eri...@gm...>**20080216124746] = [Modernise haskell98 imports in wxdirect. Eric Kow <eri...@gm...>**20080216120113] = [Prefer /usr/local/wxhaskell/bin/wx-config to the one on the path. Eric Kow <eri...@gm...>**20080216113844 = This makes it easier to install wxhaskell via cabal-install. Just set up a symbolic link to point to point to your wxWidgets, no need for the --wx-config flag. ] = [Add some --bindir, --datadir, --libexecdir to configure script. Eric Kow <eri...@gm...>**20080216113837 For Cabal. The last two are ignored. ] = [Change output dir to 'dist' to mimick cabal. Eric Kow <eri...@gm...>**20080216104847] = [Modernise haskell98 imports in wxcore. Eric Kow <eri...@gm...>**20080216100020] = [Consolidate build targets which exist purely for the sake of Cabal. Eric Kow <eri...@gm...>**20080216095845] = [Fix some dependencies that Ross Paterson pointed out. Eric Kow <eri...@gm...>**20080215181444] = [TAG 0.10.1 Eric Kow <eri...@gm...>**20080215173503] = Patch bundle hash: b21f13e6aa353e56e3c830480285f577ae47b58c --=_-- . |
From: shelarcy <she...@gm...> - 2008-02-28 06:50:04
|
Hi Eric, On Thu, 28 Feb 2008 07:34:14 +0900, Eric Y. Kow <eri...@gm...> wrote: >> wxHaskell doesn't support GHC's -split-objs option any profiling option, >> but Cabal does. > > I'm sending a quick little patch to let the user pass in flags to GHC. > We call the linker ourselves, though, so I guess my patch isn't actually > going to help matters? No, It isn't. We must tell the linker to link splited objects. -split-objs option generates it in (OutputDir)/(ModuleName)_split directory per moudles. You can see that your own directory or below cabal code. http://www.haskell.org/ghc/docs/latest/html/libraries/Cabal/src/Distribution-Simple-GHC.html#getHaskellObjects (OutputDir)/(ModuleName)_split directory has ld.script that describe just INPUT(...). So, I think we must pass that per modules. >> Current wxHaskell generates big binary. So I think this is important task, too. > > wxhaskell has a --enable-upx flag. Would that be good enough? No, that wouldn't. I want to say about executable file, not about library file. This flag and --enable-strip flag mean that using upx to compress wxc library. But I doesn't affect to generate Haskell program size. If we want to do it, to use -split-objs for spliting the single object file, to use -f* options (I don't know this is good for wxHaskell or not), or to use strip for Haskell program. http://www.haskell.org/ghc/docs/latest/html/users_guide/options-phases.html#options-linker http://www.haskell.org/ghc/docs/latest/html/users_guide/options-optimise.html#options-f http://www.haskell.org/ghc/docs/latest/html/users_guide/smaller.html -- shelarcy <shelarcy hotmail.co.jp> http://page.freett.com/shelarcy/ |
From: Eric Y. K. <eri...@gm...> - 2008-02-27 22:34:59
|
> Wed Feb 27 13:48:48 GMT 2008 Eric Kow <eri...@gm...> > * Fix bugs makefile under Linux > > Wed Feb 27 13:51:40 GMT 2008 Eric Kow <eri...@gm...> > * Specify what OS BoxedCombinator bug fails on. These look harmless enough to go in. -- Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow> PGP Key ID: 08AC04F9 |
From: Eric Y. K. <eri...@gm...> - 2008-02-27 22:34:26
|
> wxHaskell doesn't support GHC's -split-objs option any profiling option, > but Cabal does. I'm sending a quick little patch to let the user pass in flags to GHC. We call the linker ourselves, though, so I guess my patch isn't actually going to help matters? > Current wxHaskell generates big binary. So I think this is important task, too. wxhaskell has a --enable-upx flag. Would that be good enough? -- Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow> PGP Key ID: 08AC04F9 |