>>>> "sb" == SL Baur <steve(a)xemacs.org>
writes:
sb> Hrvoje Niksic <hniksic(a)srce.hr> writes in
sb> xemacs-beta(a)xemacs.org:
> Totally off-topic, but if someone mentions the "30%"
figure
> once again, I'll SCREAM! It's a myth. Subjectively, Mule is
> at least 3 times slower than non-Mule --
sb> Not to me, but I won't push the matter.
I can't perceive the difference between Mule and no-Mule for most of
the fundamental editing and redisplay operations even on an ancient
Sparc ELC or a 80486/50MHz (both are XEmacs 20.4, though), including
stuff like sort-lines. Once a Japanese font is loaded, changing to
that font is no slower in XEmacs than in a kterm, perceptually. If
Hrvoje can perceive those differences as a half-order of magnitude,
he's quick. I doubt he'd notice any difference using a Mule XEmacs
for `xemacs -batch -l dunnet' ;-)
Where I do perceive differences is in large buffers that need to be
randomly accessed (eg, a threaded VM folder). There the slowdown is
huge, although I'm not sure how much is due to Mule and how much is
due to VM's lack of optimization for non-English. Eg, my "Student
Advising" folder has several hundred MIME BASE64-encoded-words for
Japanese names and subjects, and every time its Summary gets resorted
or something a "Decoding base64 ..." message gets echoed a few hundred
times, which does slow things down I bet.
I wonder how many other lisp packages are written so that in their
default configuration they write scads of informational messages or
otherwise respond to "unusual" occurrances that aren't so unusual in
environments where Mule is essential. Like BASE64 decoding, which is
used for every message header that contains Japanese. I'm sure VM
warns about this because in an ASCII world base64 decoding is used
mostly on objects like images where the operation itself will take a
long time. Decoding 4 JIS characters doesn't deserve mention IMHO but
YMMV :-)
I suspect it's hard to write a good benchmark for this.
--
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences Tel/fax: +81 (298) 53-5091
__________________________________________________________________________
__________________________________________________________________________
What are those two straight lines for? "Free software rules."