bug#6677: 24.0.50; NS variant cannot copy to pasteboard correctly

[David De La Harpe Golden]

On 20/07/10 11:13, Peter Dyballa wrote:
> Hello!
>
> Although it obviously works to copy some text in an X client or in a Mac
> OS X application and paste it into the *scratch* buffer it does not work
> to mark some text in any buffer of the NS variant and either use the
> Copy entry from the Edit menu or just press ⌘-C (Cmd-C) and insert
> (paste) it into the X client or a Mac OS X application (the name of the
> application in the  bar only flashes, because some bad content, not
> fitting into the text pane, is held in pasteboard?).

My MacOSX/GNUStep/ObjC, while not quite non-existent, is not amazing.

But it looks to me like src/nsselect.m  needs some adjustment so it can 
handle recent changes to defaults (and N.B. in fact to handle settings 
that have been possible for a long time)

It looks like it has been mapping NSGeneralPBoard (the macosx "normal" 
cut/copy/paste clipboard) to an emulated X11 PRIMARY, when in fact 
NSGeneralPBoard is much more closely analagous to X11 CLIPBOARD.

*** Since x11 emacs is now actually using X11 CLIPBOARD, it is likely 
that the attached  _untested_ but minor patch or something very like it 
is correct on nextstep.

Can you try building an emacs with the patch on macosx?

Note that even with the patch applied, it may not act quite the same as 
before, but that is not necessarily a bug but rather a change.

(N.B. it should be possible to get historical behaviour back (even 
though looking at nsselect.m it was probably a bit odd) by twiddling 
customizations).

> Additionally double-click to mark a word is interpreted as
> <down-mouse-1> <mouse-1>, bound to mouse-drag-region. I don't think that
> this is desirable...

N.B. Mouse-drag-region calls mouse-drag-track which calls 
mouse-start-end to _implement_ double-click to mark a word, perhaps not 
intuitively given the name. So does double-clicking nonetheless actually 
mark the word on NS?

ns-clipboard-pboard_r1.diff
Description: Text Data