Hi,
I tried to $subject, and ran into small problems.
First, geany_private.rc now includes geany.exe.manifest, but the 1.23.1 bzip lacks this file. Maybe the gzip archive has it, I havent tested. Copying a manifest from the current git works fine. It has the geany version set to 1.24.0.0 though.
"C:\MINGW\BIN\windres.exe: can't open file `geany.exe.manifest': No such file or directory"
--
Second, both the mail announcement for 1.23.1 and the archive files are dated after May 05, but wscript lacks the code which alters the C compiler detection under Win~1. So waf properly detects msvc on my system, and Geany fails to build. Speaking of which, it'll be a good idea to change the C++ compiler detection as well - I don't think a mix of gcc-compiled Geany sources and msvc-compiled Scintilla will work.
If we don't want to release 1.23.1.1 (or 1.23.2) ASAP, I'll change and test the compiler detection under Win~1 tomorrow.
On 05/06/13 19:05, Dimitar Zhekov wrote:
Hi,
I tried to $subject, and ran into small problems.
First, geany_private.rc now includes geany.exe.manifest, but the 1.23.1 bzip lacks this file. Maybe the gzip archive has it, I havent tested.
Sorry, I forgot to add it to Makefile.am. Will fix this later in GIT.
Copying a manifest from the current git works fine. It has the geany version set to 1.24.0.0 though.
This is normal, as GIT master (I guess you pulled it from master) gets a versiom bump after a release.
Second, both the mail announcement for 1.23.1 and the archive files are dated after May 05, but wscript lacks the code which alters the C compiler detection under Win~1. So waf properly detects msvc on my
As expected. This change was not baked into the 1.23.1 as this release was only to address important bugs and build-system-only related issues are not that critical.
system, and Geany fails to build. Speaking of which, it'll be a good idea to change the C++ compiler detection as well - I don't think a mix of gcc-compiled Geany sources and msvc-compiled Scintilla will work.
True. We should do this. Do you volunteer as you seem to have a system with both compilers installed and so you can test this best? :)
If we don't want to release 1.23.1.1 (or 1.23.2) ASAP, I'll change and test the compiler detection under Win~1 tomorrow.
I don't think we want, at least not because of the mentioned issues. Except for maybe three or four people, I guess nobody builds Geany on Windows.
Regards, Enrico
Le 05/06/2013 19:56, Enrico Tröger a écrit :
On 05/06/13 19:05, Dimitar Zhekov wrote:
Hi,
I tried to $subject, and ran into small problems.
First, geany_private.rc now includes geany.exe.manifest, but the 1.23.1 bzip lacks this file. Maybe the gzip archive has it, I havent tested.
Sorry, I forgot to add it to Makefile.am. Will fix this later in GIT.
Indeed, done. I also added an other file missing from the tarball since the last Scintilla update.
Cheers, Colomban
On Wed, 05 Jun 2013 19:56:39 +0200 Enrico Tröger enrico.troeger@uvena.de wrote:
system, and Geany fails to build. Speaking of which, it'll be a good idea to change the C++ compiler detection as well - I don't think a mix of gcc-compiled Geany sources and msvc-compiled Scintilla will work.
True. We should do this. Do you volunteer as you seem to have a system with both compilers installed and so you can test this best? :)
Here. Tested with 1.23.1 (not git), but that's a 1-liner and can't be™ any different.
just to save you some trouble: before compiling, look at https://sourceforge.net/p/scintilla/code/ci/e806db1e88ea1762815760af704cc414... which fixes compilation with older gcc versions and might be necessary,
The wonders of 21-st century C++. Thanks, my gcc seems reasonably new.
On 06/06/13 19:29, Dimitar Zhekov wrote:
On Wed, 05 Jun 2013 19:56:39 +0200 Enrico Tröger enrico.troeger@uvena.de wrote:
system, and Geany fails to build. Speaking of which, it'll be a good idea to change the C++ compiler detection as well - I don't think a mix of gcc-compiled Geany sources and msvc-compiled Scintilla will work.
True. We should do this. Do you volunteer as you seem to have a system with both compilers installed and so you can test this best? :)
Here. Tested with 1.23.1 (not git), but that's a 1-liner and can't be™ any different.
Thanks, committed.
Regards, Enrico
On Fri, 07 Jun 2013 06:59:13 +0200 Enrico Tröger enrico.troeger@uvena.de wrote:
On 06/06/13 19:29, Dimitar Zhekov wrote:
[...] Speaking of which, it'll be a good idea to change the C++ compiler detection as well [...]
Tested with 1.23.1 (not git), but that's a 1-liner and can't be™ any different.
Thanks, committed.
Ah, forgot the import of cxx_compiler. Should have relied more on diff+patch and less on my memory...
On 07/06/13 20:28, Dimitar Zhekov wrote:
On Fri, 07 Jun 2013 06:59:13 +0200 Enrico Tröger enrico.troeger@uvena.de wrote:
On 06/06/13 19:29, Dimitar Zhekov wrote:
[...] Speaking of which, it'll be a good idea to change the C++ compiler detection as well [...]
Tested with 1.23.1 (not git), but that's a 1-liner and can't be™ any different.
Thanks, committed.
Ah, forgot the import of cxx_compiler. Should have relied more on diff+patch and less on my memory...
Same here: I should have looked at the change close enough to see the missing import. Well, fix committed. Thanks.
Regards, Enrico
Hey,
If we don't want to release 1.23.1.1 (or 1.23.2) ASAP, I'll change and test the compiler detection under Win~1 tomorrow.
just to save you some trouble: before compiling, look at https://sourceforge.net/p/scintilla/code/ci/e806db1e88ea1762815760af704cc414... which fixes compilation with older gcc versions and might be necessary, depending on your build environment.
@Colomban: are you ok with backporting this change from Scintilla or just wait for the next Scintilla release? Right now the nightly builds are broken because of this. If you agree, I could do it on the weekend.
Regards, Enrico
Le 05/06/2013 23:27, Enrico Tröger a écrit :
Hey,
If we don't want to release 1.23.1.1 (or 1.23.2) ASAP, I'll change and test the compiler detection under Win~1 tomorrow.
just to save you some trouble: before compiling, look at https://sourceforge.net/p/scintilla/code/ci/e806db1e88ea1762815760af704cc414... which fixes compilation with older gcc versions and might be necessary, depending on your build environment.
@Colomban: are you ok with backporting this change from Scintilla or just wait for the next Scintilla release? Right now the nightly builds are broken because of this. If you agree, I could do it on the weekend.
It's in 3.3.3 so there's no need to wait :) I updated it, so it should be fixed.
Regards, Colomban
On 06/06/13 17:11, Colomban Wendling wrote:
Le 05/06/2013 23:27, Enrico Tröger a écrit :
Hey,
If we don't want to release 1.23.1.1 (or 1.23.2) ASAP, I'll change and test the compiler detection under Win~1 tomorrow.
just to save you some trouble: before compiling, look at https://sourceforge.net/p/scintilla/code/ci/e806db1e88ea1762815760af704cc414... which fixes compilation with older gcc versions and might be necessary, depending on your build environment.
@Colomban: are you ok with backporting this change from Scintilla or just wait for the next Scintilla release? Right now the nightly builds are broken because of this. If you agree, I could do it on the weekend.
It's in 3.3.3 so there's no need to wait :) I updated it, so it should be fixed.
Cool. I didn't even notice 3.3.3 is already out. Thanks for the update.
Regards, Enrico