On 1 Mar 2003 22:20, Richard Levitte - VMS Whacker wrote:
> In message <[EMAIL PROTECTED]> on Sat, 1 Mar 2003 13:59:55 -0500, Bjorn Knutsson 
> <[EMAIL PROTECTED]> said:
> 
> bjorn> The current version seems to have at least one glitch. Any key I map
> bjorn> in the workspace context gets mapped out in every other context. Since
> bjorn> I have the cursor keys mapped in workspace, that got pretty annoying
> bjorn> fast, so I'm again back to running my version.
> 
> I'm not sure I catch what you're saying.  Care to share your .ctwmrc?

Well, the simple answer is that if you bind a key in the workspace
context, it becomes dead in all other contexts:

"a" = : workspace : f.beep

would make the "a" key generate a beep in the workspace context, as
expected. But it would mean "a" could not be used in Emacs, your XTerm
or any other program. Obviously, this is not what you'd expect or want.

Notice also that this is *not* the behavior you get if you apply my
patch to 3.6, so something else has changed.

> bjorn> I haven't (yet) looked into why it does this, but simple diff of my
> bjorn> patch and the code in 3.7 reveals that the code in 3.7 does not match
> bjorn> the code in my patch in at least one file (events.c).
> 
> Yup, it seems Claude added the flag wspmgr, so you don't have to do
> the compare proposed in your patch (we *are* talking about the
> workspace patch, right?)...

Well, the problem does not seem to be in the patch, or even the
changes made to the patch. Binding keys works great. Except that
whatever key you bind in the workspace context gets mapped out in all
other contexts.
 
> If I had a clearer picture of what you expect, I might be able to fix
> things.  An exampel .ctwmrc with a description of what's expected
> would be great.

Well, I expect that binding a key in the workspace context should not
render it useless everywhere else. Something has changed since 3.6 and
I can't figure out exactly what from the changelog. The changes you
say Claude has made does not show up in the changelog, and looking at
some of the files, event.c for one, there are significant changes.

For example, the test I made for:

Tmp_win == Scr->workSpaceMgr.workspaceWindow.twm_win

is made in other places in the original code. Looking at 3.7alpha4,
some have been changed to a test for:

Tmp_win->wspmgr

others to:

Tmp_win == Scr->currentvs->wsw->twm_win

I don't understand exactly what has changed, so I can't comment on
why, but it *could* be related to the Xinerama-patches.

/Bj�rn

Reply via email to