Re: Xy3 memory

["..." yesss@xxxxxxxx]

Hi, Myron. I'm unfamilar with views.pm but--hoping that
I understand the prob--I write a XyW3.57 macro that opens
the same 144+K file in two windows and in the second
instance jumps to a phrase that first occurs two-thirds
of the way through. When run, #testmac lands the cursor
where it should be in window #2 and in window #1 the cursor
is still at TOF. If #testmac works, the problem isn't xyW3
and using xyW4 is hardly the only solution.

If you suspect that you have a memory prob, get a func ME
report and act accordingly. How big are the printer and help
files you keep loaded while editing, e.g.? With the 144+K
file open in both windows, my func ME reports available
memory is 186 [whatever ME units are]. With only a dir
listed in each window the total available is 464. My PR3
and HL3 occupy 1 each. The files are 1695 and 7468 bytes
respectively. The Help file is the only one I ever load,
but this PR3 is strictly a screen driver (like a xyWin
XWSET.DFL that doesn't load another PR3--I load one specific
to my printer only while actually printing). The other
killer is your user dictionaries. I have no hyphenation
dictionary and MY.SP3 is only 1539 bytes but, unloaded,
it and DICT.SPL knock off 4. ... Ciao. 		--a

=======================================
adpFisher  nyc
xyWrite 3 supplements !xyWise and !xyWiz +
Wolfgang Bechstein's seafaring adventures:
http://www.escape.com/~yesss/ 
=======================================

≪ Call a 100K file.
Jump to the beginning.
Run Views.pm and choose 2 views.
Open view #2.
Search for a piece of text 2/3 of the way through.
Now jump to view #1. You should be at the beginning of the
	file, but you won't be. You can jump to View #2 correctly,
	but not to #1 (when it it near the beginning of the file).
Jump to the beginning of the file.
Search for something a few lines down.
Now you can jump from a View #2 (with a cursor location
	at, say, 70K to a View #1 (with the cursor at or near
	the beginning.)

It seems to have something to do with searching over
a distance of approximately 64K or more and then trying
to use Views.pm. Shorter searches don't affect Views.pm,
and running a shorter search seems to correct the problem
temporarily. Any suggestions? ≫

--Myron