Re: Abysmal state of GTK build

[Yilkal Argaw]

> Po Lu <luangruo@yahoo.com> Unsubscribe
>
> Aug 21, 2022, 2:06 PM (2 days ago)
>
>
> to emacs-devel
> Users of the GTK 3 build experience many, many problems.  The most
> recent such problem is bug#56869, which is definitely a bug in GTK.
>
> Taking into account the very low quality of the GDK X11 backend, which
> is not seeing active maintenance, shouldn't the build default to some
> other toolkit such as Motif, or even better, no toolkit at all?
>
> Especially considering that a GTK developer (the tail) wants to remove
> support for X11 (the dog) in future releases of GTK:
>
>  https://gitlab.gnome.org/GNOME/gtk/-/issues/5004


I recently tried to investigate this issue so I configured emacs with
the "with-x-toolkit=no" option. The thing that bothered me more than
dated look was the fact that the mouse button behaviours (clicks
doubleclicks and whatnot) do not work like many modern users have
gotten used to. I am of that generation who learned computing from in
windows 2000 era so holding the click button to navigate through menu
entries was very confusing. I know this behaviour dates back in the
history of X and that programs like xterm use it but I would think
making the default might turn out to be a lot more confusing to new
users. This behaviour is also a bit hard to use with touch pads (since
many users just tap for clicks). I might be missing something so
correct me if I am wrong.

with regards
Yilkal A.

On Tue, Aug 23, 2022 at 3:44 AM Po Lu <luangruo@yahoo.com> wrote:
>
> Eli Zaretskii <eliz@gnu.org> writes:
>
> >> From: Po Lu <luangruo@yahoo.com>
> >> Cc: djcb@djcbsoftware.nl,  emacs-devel@gnu.org
> >> Date: Mon, 22 Aug 2022 20:59:38 +0800
> >>
> >> Eli Zaretskii <eliz@gnu.org> writes:
> >>
> >> > What about PGTK build on Wayland, not on X?
> >>
> >> It has the same input-related problems to a much lesser degree, and
> >> selections work fine.  And either way, it's the only choice on Wayland.
> >
> > Does it cause crashes?
>
> No, but it causes things like bug#53200, where C-S-u is reported as C-u.
>