Re: [Pan-users] 64-Bit Windows build

[Petr Kovar]

On Wed, 26 Apr 2017 09:57:18 +0000
Steve Davies <address@hidden> wrote:

> Bugzilla - That would have been the best answer - particularly as I already
> have an account now you mention it! :)
> 
> While I am here - the following is from memory, I'm away from my pan dev
> environment.
> 
> I notice that COMPRESS_GZIP and ENABLE_COMPRESS_GZIP are defined in
> compress.h, which is #included in many places, but both values are used
> only once each in nntp.c - On my version of g++ this causes many "defined
> but not used" compile warnings. Would it be better to move those 2 lines
> into nntp.c? Or replace them with #defines?

I think moving them to nntp.cc would be a nice fix, yes, though I don't
really know that code so it would need some testing, too.

Could you submit a patch? I think people following pan master could be able
to test it.

Thanks,
pk 


> On Tue, 25 Apr 2017 at 23:51 Petr Kovar <address@hidden> wrote:
> 
> > On Tue, 25 Apr 2017 20:31:22 +0100
> > Steve Davies <address@hidden> wrote:
> >
> > > On 25/04/2017 20:23, Petr Kovar wrote:
> > > >
> > > >> Yes, happens to me too (I had not noticed!) on Windows 10.
> > > >>
> > > >> I'll try and update everything I can and re-build. It is not a
> > > >> fundamental date formatting issue because the article pane shows the
> > > >> date just fine...
> > > > Was just wondering, do you build Pan with GTK+ 2 or 3? I wouldn't be
> > > > surprised if this was a problem in older versions of GTK+.
> > > >
> > > > pk
> > > >
> > > GTK+3 is not yet usable on Windows builds as far as I can tell, but I
> > > did find the issue - Not sure how to submit the following, so here is a
> > > "patch" for you to check and merge if you like it.
> >
> > Thanks, Steve, for a quick fix! :)
> >
> > > commit cfbf70be8e083b8a35fc94fad5d5c902c024a223
> > > Author: Steve Davies <address@hidden>
> > > Date:   Tue Apr 25 20:13:33 2017 +0100
> > >
> > >      Fix the strftime test for mingw64
> > >
> > >      mingw64's strftime returns an empty string when the %l and %h
> > > tokens are not understood.
> > >
> > > diff --git a/configure.ac b/configure.ac
> > > index 4caae04..cce6c24 100644
> > > --- a/configure.ac
> > > +++ b/configure.ac
> > > @@ -240,13 +240,14 @@ dnl Check to see if strftime supports the use of
> > > %l and %k
> > >   AC_MSG_CHECKING(for %l and %k support in strftime)
> > >   AC_RUN_IFELSE([AC_LANG_SOURCE([[
> > >   #include <string.h>
> > > +#include <stdlib.h>
> > >   #include <time.h>
> > >   int main(int argc, char **argv) {
> > >     char buf[10];
> > >     time_t rawtime = time(0);
> > >     struct tm *timeinfo = localtime (&rawtime);
> > >     strftime(buf, 10, "%l %k", timeinfo);
> > > -  exit (strstr(buf, "l") || strstr(buf, "k") || !strcmp(buf," "));
> > > +  exit (strstr(buf, "l") || strstr(buf, "k") || !strcmp(buf," ") ||
> > > !strlen(buf));
> > >   }]])],[
> > >   AC_DEFINE(HAVE_LKSTRFTIME, 1, [strftime supports use of l and k])
> > >   ac_cv_lkstrftime=yes
> > >
> > >
> > > The above will probably be destroyed in-transit :(
> > > Steve
> >
> > Yeah, the line breaks didn't survive it, so I recreated it manually. I
> > think it's better to send it as an attachment, since git format-patch
> > creates a patch in a mail format. Or, pan is mirrored on GitHub if you're
> > there. Or, Bugzilla...
> >
> > Cheers,
> > pk
> >