bug#35058: [PATCH] Use display-graphic-p in more cases

[Eli Zaretskii]

> From: Alex <agrambot@gmail.com>
> Cc: 35058@debbugs.gnu.org
> Date: Tue, 02 Apr 2019 23:14:06 -0600
> 
> >> > If it is very inconvenient to use hyper on text terminals, then I
> >> > think we shouldn't require users to do that.
> >> 
> >> It wouldn't be a matter of requiring, but instead forcing the modifier
> >> back to 'meta if both in terminal Emacs and cua-rectangle-modifier-key
> >> is 'hyper.
> >
> > Sorry, I don't understand.  Can you show how would you like to remove
> > the condition?
> 
> I don't agree with my position in my last email since I found out that
> my terminal has a special workaround for 'super.
> 
> Removing the condition would just mean always using
> cua-rectangle-modifier-key in cua--init-keymaps even if in a terminal:
> 
> (setq cua--rectangle-modifier-key cua-rectangle-modifier-key)
> 
> This would mean that terminal users that can't use super/hyper modifiers
> easily would might find it confusing, but the current behaviour is a bit
> confusing as is.

I think the use case behind that condition is of a user who uses both
GUI and TTY frames, and have cua-rectangle-modifier-key customized to
something that doesn't easily work on TTY.  Perhaps the right solution
would be to have 2 separate defcustom's, one each for every frame type.

> Why do you only say that TTY frames _generally_ don't have function
> keys as symbols?

Because there are exceptions, like the MS-Windows console.

> How about display-symbol-keys-p (if TTY frames can't support symbols in
> key sequences)? 

Fine with me, thanks.  Although having 2 defcustom's might eliminate
the need for the test and for the predicate.