ADSM-L

Re: Mac ADSM and Mac Word Perfect?

1994-07-13 16:59:18
Subject: Re: Mac ADSM and Mac Word Perfect?
From: Nick Laflamme <NLAFLAMM AT VMA.CC.ND DOT EDU>
Date: Wed, 13 Jul 1994 15:59:18 EST
On Wed, 13 Jul 1994 15:51:01 -0500 David E Boyes said:
>> The symptom is WP crashing on initialization in some instances.  ADSM
>> doesn't have to be doing anything with the server, merely running (or
>> having run earlier, perhaps).  Use of an AppleShare file server seems to be
>> relevant as well, although we're not using ADSM to backup the file server.
>
>Could it possibly be related to the memory leak problem that Brett posted
>here a few days ago with MacTCP and other stuff?  Is it a true AppleTalk
>connection, or is it Ethertalk?

Considering that we're seeing this right after rebooting the Mac for test
purposes, I'd be mildly surprised if it's a memory leak problem, but it
might be.  (Then again, your question reminds me that turning off 32 bit
addressing in the Mac is alleged to help.) It's certainly easy enough to
grab the new Mac client and see if that helps.  (Thanks, IBM, for the FTP
server with the clients on them!)

It's Ethertalk on at least one machine involved.  (Yes, the user is running
ADSM and WP on multiple Macs; the problem shows up on more than one of the
Macs of different vintages. System 7.1 is common, though.)

>> Right now, the evidence is all circumstantial, but there's a disturbingly
>> strong correlation between ADSM and WP not initializing.  I'm trying to get
>> a feel whether others are in a position to confirm or refute fears that
>> there's a bad interaction between ADSM and WP for the Mac.
>
>WP for the Mac is an serious memory hog, so if you are tripping the memory
>leak bug, WP would show the effects most graphically, esp.  if it's having
>to run with a fragmented memory heap and do tons of garbage collection.

I've heard WP for the Mac called lots of things, few of them nice. However,
I can't just blame WP since there are other factors involved, and ADSM
might be one of them.  Best to fix the byte-leak and see what improves.

Thanks,
Nick

* Intelligence demands reasons, not rules.
<Prev in Thread] Current Thread [Next in Thread>