On Thu, Aug 18, 2011 at 11:21:28AM +0200, Pascal Stumpf wrote: > On Thu, Aug 18, 2011 at 10:17:31AM +0200, Mark Kettenis wrote: > > Makes no sense. The keepSelection check is already done in > > ScrnDisownSelection(). The only place where DisownSelection() is > > called directly is SelectSet(), and that should only happen if you > > explicitly shrink your selection to "nothing". > > > > As far as I know, xterm keeps the selection just fine. If I select > > some text in one xterm, then click at some random place in that same > > xterm such that the text is no longer highlighted, I can paste it just > > fine into another xterm using the middle mouse button. > > Having done some simple tests, the situation seems to be like this: > > If you highlight some text in an xterm, it's copied to PRIMARY just fine > and kept there if keepSelection is true. However, PRIMARY cannot be > accessed by gtk apps, which use the CLIPBOARD. So most people do > something like > > xterm*VT100.translations: #override <Btn1Up>: select-end(PRIMARY, CLIPBOARD, > CUT > _BUFFER0) > > in their .Xdefaults. And here's the problem: If you select text, it gets > copied to both PRIMARY and CLIPBOARD. If you then click somewhere else > in the same xterm, it is kept in PRIMARY, but *not* in CLIPBOARD (so > it's lost for gtk apps).
This is the behavior that should happen with "xterm*keepSelection: false" > Hope this helps. It doesn't because this isn't what the man page states for keep selection. It only works with Scrn* functions and does not get tested when DisownSelection is called, which is called from some other spots directly as well instead of going through Scrn*.
