bad redrawing performance

[David Kastrup]

Neal Becker <ndbecker2 at gmail.com> writes:

> I'm playing with dxpc (differential x compression).  I find that
> xemacs (21.5.27) behaves terribly.  For example, M-x shell, then run
> say, 'ps aux' redraws over and over.
>
> The same experiment with emacs (21.4-17) behaves very well.

Actually, comparisons between XEmacs 21.5.27 (a current developer
version) and Emacs 21.4 (a rather old stable version) don't make much
sense.  For a head-to-head comparison, you would rather need to use a
pretest of Emacs 22 (the strange version number 21.4-17 suggests some
GNU/Linux packaging, in which case something like emacs-snapshot-gtk
or similar might offer a more realistic comparison).

Sensible comparisons would be with
a) XEmacs 21.4.nottooold (this tells about regressions within XEmacs)
b) Emacs 22.0.somepretest (this tells about the impeding competition)
c) Emacs 20.7 or XEmacs 19.something in case all of the above options
suck terribly and one wants to remind people that good programming is
going down the pits nowadays.

In fact, while you might not be too terribly interested in helping the
competition, option b) might be rather welcome to Emacs developers, I
guess, since they are about to release Emacs 22 any year now, and a
regression check in an area that has not seen much visibility, even if
it is just for kicks, could be nice to have.

Apropos: what's the release idea for XEmacs?  What is the next big
event, and which versions will be declared what ("From now on, ... is
the ... branch, and all ... are expected to upgrade to ... until
... since we are going to phase out support for ... by the date of
...")?

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum