🚀 go-pugleaf

Roderick W. Smith <rodsmith@indiana.edu> wrote:
>In message <322D04BF.75C5@ibm.net> - Steve Behman <sbehman@ibm.net> writes:
>>Roderick W. Smith wrote:
>>>robdel@ozemail.com.au (Robert De Lazzari) writes:
>>>>
>>>> My EA file is in the order of 22 Megs. I am using a FAT formatted
>>>> HD and OS2 Warp.
>>>
>>>Without knowing the size of the partition or how many files have EAs,
>>>it's hard to say if this is excessive, but my hunch is it's normal.
>>
>>I have 4 Fat partitions on my system all about 150Mb in size -- the
>>largest ea data. sf is 1.6 Mb and the sum for the 4 partitions is less
>>than 3Mb.  I would, therefore, suggest that yours is EXCESSIVE!
>
>This comparison is too simple.  For instance, your four FAT
>partitions are rather small, and use 4kB allocation block sizes.
>With larger allocation block sizes, EAs will be stored in larger
>chunks -- perhaps as high as a multiple of 32kB per file rather than
>the 4kB you've got.  This difference alone could account for the
>difference.  There's also the question of how many files have EAs
>attached.

Furthermore, it will depend on usage.

If you, for instance, run a REXX script to set a WPS attribute (say,
"maintain sorted order") on every directory on the drive, then every
directory will get an EA - meaning at least one disk block in the EA
data file.

If you, perhaps, attach a custom icon to every text file, then the
size of the ICO file will be written to each file's EAs, taking up
more space in the file.

The fact that an EA data file is big doesn't mean anything's wrong
with the system.  It may simply mean that your files have a lot of EAs
attached to them.

---------------------+--------------------------------------------------------+
David Charlap        | The contents of this message are not the opinions of   |
david@visix.com      | Visix Software, nor of anyone besides myself.          |
Visix Software, Inc. +------------------------------------------------+-------+
Member of Team-OS/2  | Vote Dogbert for supreme ruler of the universe |
---------------------+------------------------------------------------+