4 Apr
1999
4 Apr
'99
1 p.m.
16megs is the minimum to run W95 in my experience and it runs much better
with 32m. It's a pig.
<My P200 machine, when with only 32 meg, swapped itself to pieces it seemed
<after a few days since a reboot even though I'd shut down all unused apps
<to run one. System resources still were not released. After kicking up to
<64 meg I now can go for several weeks until needing a restart to recover.
It's called memory leaks. Seems some applications do not return resources
to the pool as do some win95 drivers. often it can be tracked to one bad
driver.
And to think my uVAXII is still running with the 9meg that I used to
service a half dozen users plus network activity.
Allison
4 Apr
4 Apr
8:42 p.m.
Upon the date 02:00 PM 4/4/99 -0400, Allison J Parent said something like:
16megs is the minimum to run W95 in my experience and
it runs much better
with 32m. It's a pig.
<My P200 machine, when with only 32 meg, swapped itself to pieces it seemed
<after a few days since a reboot even though I'd shut down all unused apps
<to run one. System resources still were not released. After kicking up to
<64 meg I now can go for several weeks until needing a restart to recover.
It's called memory leaks. Seems some applications do not return resources
to the pool as do some win95 drivers. often it can be tracked to one bad
driver.
And I have not yet found a memory leak 'cleanup' utility which could either
be run when needed or as a background app which would recover some of the
resources.
While spending a day learning more about this problem and hunting for
solutions I did find a program on ZDnet somewhere which flushes the disk
caches and thus helps slow down using up all the resources. Improved
performance a bit too.
And to think my uVAXII is still running with the 9meg that I used to
service a half dozen users plus network activity.
Well, that helps illustrate to the hard-core windoze fans what a Real
Operating System actually is :) And, to add another log onto the fire of
reason, your mention of drivers being a culprit illustrates how the Wintel
platform's driver development applications and/or sloppy writing of drivers
can contribute to the problem.
Regards, Chris
-- --
Christian Fandt, Electronic/Electrical Historian
Jamestown, NY USA cfandt(a)netsync.net
Member of Antique Wireless Association
URL: http://www.ggw.org/awa
9407
days inactive
9407
days old
1 comments
2 participants
participants (2)
-
allisonp＠world.std.com -
cfandt＠netsync.net