Comparative Speed of Xy4-DOS vs. XyWin [Was: "xy windows [again]"]

[Robert Holmgren holmgren@xxxxxxxx]

** Reply to note from Adriano Ortile Sun, 20 Dec 1998

I automated your four tests, and ran them repeatedly against Xy4-DOS and XyWin in
various configurations -- and I'll be damned if I'm not getting roughly the same
result that you found, in favor of XyWin. Let me see if I can dig up those tests
that Carl and I did some years back, which produced results very much in Xy4-DOS's
favor. I'm pretty sure I posted them here (or somewhere, maybe on the BBS), along
with the program that generated them. I don't remember what exactly they did (Carl
may have them). But some contemporary anecdotal evidence is that the programs UVB
and UVR (User Variable Backup and Restore) in XYWWWEB.U2 take about three times as
long to run in XyWin as in Xy4-DOS (intensive searching, replacing, and cursor
movement). So I'm startled by your evidence. The only divergence from your
results that I noticed was that Test 3 would not run in either Xy4-DOS
(Error #18, not enough memory) or XyWin (Protection Violation, crashed session).
There are just too many so-called "page elements", basically a variety of
formatting commands, for EDITOR to process (thousands and thousands of them; I
think the theoretical limit is something like 867 page elements in one file --
all this means is that you can't display such a file in Page-Line mode).


-----------
Robert Holmgren
holmgren@xxxxxxxx
-----------