Re: Viper generates lots of X11 traffic -- how to debug?

[Michael Kifer]

>>>>> "KG" == Kai =?iso-8859-1?q?Gro=DFjohann?= <Kai> writes:

    KG> address@hidden (Michael Kifer) writes:

    >> It seems that somehow when you hit this tab state gets changed numerous
    >> times (which is strange).

    KG> It turns out that it is not necessary to use the strange pacakges to
    KG> see this.  It is sufficient to hit ESC at the end of the Subject
    KG> header (after putting something in it).

    KG> Then I see this:

    KG> * Cursor is narrow bar (as is normal in insert state, in command state
    KG> it's the normal large block).  The cursor is green.  I hit ESC now.

    KG> * Cursor changes color to black -- still narrow bar.

    KG> * Cursor changes color to green -- still narrow bar.

    KG> * Cursor changes color to black -- still narrow bar.

    KG> * Cursor moves one char to the left, changes to block.  (Here, the
    KG> order is not so clear, it happens fast.)

Yes, two color changes are redundant in this case, but viper tries to paint
the cursor in the right color in all cases. There are cases when you switch
to a different buffer in the same or another frame and then back, and all
these cases need to be handled. It might be possible to optimize this
further, but will complicate the code. I didn't try this because it was
fast enough (I gave up trying to use emacs in the X mode over slow
connections long ago).
I think a more reasonable solution is to provide a variable that can turn
this feature off. Ideally, this should be done automatically when
connection is slow. Does somebody know if this is possible to find out from
within Elisp?


        --michael