🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.sys.mac.apps
Article #11602

Re: Is Auto-Doubler worth it?

#11602
From: John T. Chapman
Date: Mon, 11 Jan 1993 15:35
61 lines
3914 bytes
In article <C0o9xt.4pt@gpu.utcs.utoronto.ca> Doug Moen,
dmoen@gpu.utcs.utoronto.ca writes:
>I own both Auto-Doubler 2.0 (which I use) and Norton Utilities 2.0.
>I'm running system 7.1.  Apple's Disk First Aid 7.1 tells me my disk is
>okay.  But when I run Norton Disk Doctor, it reports numerous problems
>(with the catalogue B-tree header, volume bitmap, etc), and offers to
>fix them.  After NDD "fixes" these problems, my disk is corrupt.  This
>corruption manifests itself in two ways:
> (a) the finder tells me my disk is damaged at boot time, and offers to
>     reformat it
> (b) even when (a) does not occur, Disk First Aid will still report disk
>     problems after my disk has been "Nortonized".  After Disk First Aid
>     fixes the disk, my machine runs okay, but now Norton thinks my disk
>     is corrupt again.

I also had this problem and talked to Symantec's tech-support for a 
while, trying to solve the problem.

In a nutshell, the problem occurs if you are working with a fairly full 
disk and running a transparent compression program such as Autodoubler
that 
creates/deletes a lot of temp files. What happens is that the desktop's 
catalog extents and such get _really_ fragmented due to the file creation 
and deletion. The fix was to grab several large files, move them off the 
disk, and run NDD Speed Disk to defragment the volume. (The actual
process 
was a bit trickier, but that's the basic idea.) 

Symantec, of course, recommended that such compression programs not be 
used. A more practical (if possibly less safe) recommendation is to keep
a 
fair bit of space free and run Speed Disk fairly often to try and keep
the 
desktop and directory from becoming extensively fragmented. Another 
suggestion is to lock files and applications that open 'write' even if 
they're really only being used 'read-only' and run the AIC for
applications 
that do modify themselves. This will cut down on the number of temp files 
that AutoDoubler generates (most of the decompresses will be to memory 
only). 

From what I could tell, this is not exactly a problem with Autodoubler as 
such, but rather with the file system in general when working with 
limited disk space. Autodoubler does a good job at making the problem
worse, 
however, due to the creation of lots of temp files of varying sizes. On 
the other hand, I imagine Autodoubler should be the best of transparent, 
non-driver level compression programs because it does allow decompression 
to memory only. Again, this is only speculation.

Well, good luck. I hope this helps a bit.

-------------------------------------------------------------------------
  John T. Chapman                                     jtc1@cornell.edu
  Office of Computing and Statistical Consulting
  Cornell University - College of Human Ecology
-------------------------------------------------------------------------
"Truth is after all a moving target/Hairs to split, and pieces that don't 
fit/How can anybody be enlightened?/Truth is after all so poorly lit"
                                             - N. Peart, Turn The Page
-------------------------------------------------------------------------
Technical Details
Message-ID: <1is448INNphf@newsstand.cit.cornell.edu>
Path: arc1-ca.pugleaf.net!sparky!uunet!zaphod.mps.ohio-state.edu!sdd.hp.com!caen!batcomputer!cornell!newsstand.cit.cornell.edu!newsstand.cit.cornell.edu!usenet
References: <20779.2B4E801D@zeus.ieee.org> <C0o9xt.4pt@gpu.utcs.utoronto.ca> <43310@sdcc12.ucsd.edu>

Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!sparky!uunet!zaphod.mps.ohio-state.edu!sdd.hp.com!caen!batcomputer!cornell!newsstand.cit.cornell.edu!newsstand.cit.cornell.edu!usenet
From: John T. Chapman <jtc1@cornell.edu>
Newsgroups: comp.sys.mac.apps
Subject: Re: Is Auto-Doubler worth it?
Date: 11 Jan 1993 15:35:35 GMT
Organization: Office of Computing and Statistical Consulting
Lines: 61
Sender: jtc1@cornell.edu (Verified)
Distribution: world
Message-ID: <1is448INNphf@newsstand.cit.cornell.edu>
References: <20779.2B4E801D@zeus.ieee.org> <C0o9xt.4pt@gpu.utcs.utoronto.ca> <43310@sdcc12.ucsd.edu>
NNTP-Posting-Host: 132.236.42.176
X-UserAgent: Nuntius v1.1.1d17
X-XXDate: Mon, 11 Jan 93 15:48:39 GMT