Re: Is Unicode character set a problem when compiling GTK sources?

[Arnavion <arnavion gmail com>]

(Re-adding codekiddy and Fan to To:)

On Tue, Feb 17, 2015 at 11:18 AM, Arnavion <arnavion gmail com> wrote:
> Compiling as either MultiByte or Unicode should not make a difference.
> That setting only affects the use of unannotated Windows API, but
> glib/gtk should already be using the W forms of those API (with an
> intermediate utf8<->utf16 step). MultiByte is probably there in the
> project files because it's the default for new projects. Changing it
> should make no difference.
>
> If you find something that breaks because it uses the unannotated form
> of a Windows API on a utf8 string expecting to get the A version, this
> would break when you changed it to Unicode since it would now try to
> run the W version on a utf8 string. If you find such a case, I think
> you should file a bug for it.
>
> (That said, I don't see the point of changing this. In the best case,
> you would find a bug where an A version of some API is being called
> with a utf8 string instead of W version with utf8<->utf16 conversion
> step. In the worst case, you'll achieve nothing.)
>
> -Arnav
>
> On Tue, Feb 17, 2015 at 11:09 AM, John Emmas <john creativepost co uk> wrote:
> > On 17/02/2015 18:36, codekiddy wrote:
> > > (even Fan's projects which I'm using) which were
> > > set to Multibyte character set are now set to *unicode character set*
> > >
> > > Can you tell what consequnces could that have?
> > > I'm starting to think it would be better to re-compile everything with
> > > Multibyte instead because I have bad problems with GTK+ on run-time, there
> > > should be reason why those projects found in official sources are all set to
> > > Multibyte??
> >
> > Hi there,
> >
> > As someone else who builds a lot of GTK related stuff with MSVC, my
> > experience has been that the most commonly used Unicode format (for GTK+
> > based projects) tends to be UTF-8.  UTF-8 can only be represented easily
> > when you select the multibyte character set. Windows Unicode uses precisely
> > 2 bytes to represent each character. But UTF-8 can use anywhere between 1
> > byte and 4 bytes (in fact, I think the early implementations could use up to
> > 6 bytes!)  It probably makes things a lot easier too for Glib::ustring.
> >
> > Hope that explains it.
> >
> > John
> > _______________________________________________
> > gtk-list mailing list
> > gtk-list gnome org
> > https://mail.gnome.org/mailman/listinfo/gtk-list