Re: hfs files

[Randy Shackelford <shack@onyx.southwind.net>]

David Empson <dempson@actrix.gen.nz> wrote:

: As Jeff said, the IIgs implementation of HFS is _horribly_ slow for
: writing.  Try copying lots of files to an HFS disk, and you can sit
: there waiting while the computer seems to do nothing for several
: seconds, then finally writes a chunk of data.  This is particularly
: noticeable with large files.

: This is especially noticeable (and amusing) if you have a noisy hard
: drive, e.g. a SyQuest.  When copying a large file to a SyQuest, I can
: tell by listening whether I'm writing to a ProDOS disk (reasonably slow
: but steady track stepping) or HFS (long pause, then a sudden burst of
: furious activity).

Like I said elsewhere, that has to be due to variable block sizes and B
tree style directories. That has to take some processing. HFS was designed
for a box with a bit more horsepower than a IIgs.

: This is with an Apple SCSI card and a 7 MHz ZIP GS.  I'd hate to think
: what it is like on an unaccelerated machine.

I have unaccelerated IIgses and it really doesn't seem so bad. Maybe I'm
more patient than most.

: It is pretty good for reading, on the other hand (not much worse than
: ProDOS).

I'll also mention that I've never had file system problems, even without the
FST patch. When I replaced my 105 hard disk with a 230, I copied the contents
of the 70 mb HFS volume to a Mac via localtalk then copied it back to the 200
mb volume on the new disk. That was about 700 files' worth. Then I put the
disk on a Mac and ran Disk Doctor. No problems there. That seems to fly in
the face of all of Nathan's bitching and carrying on about HFS and file
system problems you get with it. Have I mentioned lately how little I miss
him around here?
-- 
Offsite mail to this host gets nuked.