127 lines
6.5 KiB
Plaintext
127 lines
6.5 KiB
Plaintext
|
MY EXPERIENCES WITH DISK OPTIMIZER (tm)
|
|||
|
|
|||
|
P. L. Olympia, Sysop Darwin I and II, 301-251-9206
|
|||
|
11/29/85
|
|||
|
|
|||
|
Disk Optimizer is a commercial product from Softlogic Solutions that
|
|||
|
is intended to solve the problem of badly fragmented files on the hard
|
|||
|
disk. It reorganizes both disk directories and file clusters. Badly
|
|||
|
fragmented files cause delays in program loading or execution, and
|
|||
|
worse (in my opinion) unnecessary punishment on the disk drive.
|
|||
|
|
|||
|
Those of us who run very active bulletin boards know the problem quite
|
|||
|
well. It doesn't take long for an upload subdirectory or a disk
|
|||
|
subdirectory containing the BBS callers and messages files to be so
|
|||
|
badly fragmented that it takes quite a wait for the BBS to locate the
|
|||
|
file you are trying to download. Next time you're on a large and
|
|||
|
active BBS, pull out your stopwatch and do some timing tests.
|
|||
|
|
|||
|
My first experience with Disk Optimizer was nothing to write home
|
|||
|
about. I first tried it on one of my BBS's small hard disk (10 Meg)
|
|||
|
and had to abort after 1.5 hours as there was absolutely no indication
|
|||
|
that it was ever going to finish. The lack of a program report as to
|
|||
|
what it is really doing at a given time didn't help things any. After
|
|||
|
the abort I called up Mr. Norton and Mr. Chkdsk and found that most of
|
|||
|
the disk were indeed reorganized and "defragmentized". But heck, 1.5
|
|||
|
hours for a 10M is no joke - I could just as well backup with one of
|
|||
|
the fast backup programs like DSBACKUP, delete and restore within that
|
|||
|
time. I decided to try again. This time I tried it on the other 10M
|
|||
|
disk. The result was worse. Not only did it take more than an hour (I
|
|||
|
had to abort again), I ended up with a lot of cross-linked clusters
|
|||
|
and two of my subdirectories became damaged. It turned out that the
|
|||
|
cross-linked clusters and the resulting damage might have been my own
|
|||
|
doing as I had several hidden subdirectories on that disk. No matter.
|
|||
|
I decided to return the package. Even if the program works, I wait for
|
|||
|
no one.
|
|||
|
|
|||
|
In the meantime, my main BBS with a 33M hard disk was taking an awful
|
|||
|
lot of punishment - callers who could not read and who tried to
|
|||
|
download files that were not even listed would inevitably throw my
|
|||
|
poor Tulin into a tizzy. I sat there for two weeks listening to the
|
|||
|
agonizing disk head as it hopped from sector to sector. And all
|
|||
|
because I did not have the patience to go thru a backup/delete/restore
|
|||
|
of a 33M disk.
|
|||
|
|
|||
|
Then my friendly dealer called to say that Softlogic has issued a new
|
|||
|
version (1.2). I decided to give it another shot. I tried it on the
|
|||
|
damaged 10M (wiser now, I unhid all the subdirectories first) and
|
|||
|
couldn't believe that I was done in 19 minutes. Mr. Norton and Mr.
|
|||
|
Chkdsk bore out the results of the ANALYZE program that came with the
|
|||
|
package - all files were 100% contiguous. DBASE III took no time at
|
|||
|
all to load and PCBoard (already a speed demon of a BBS software)
|
|||
|
started acting like a bat out of hell. Now, this was on a 10M disk
|
|||
|
with only 550K free space. Still unconvinced, I tried it on an AT
|
|||
|
with a 32M and 5M free that has never been reorganized for 8 months.
|
|||
|
The program did its job in 32 minutes. So far so good.
|
|||
|
|
|||
|
I decided to push my luck. I tried the program on the poor 33M Tulin
|
|||
|
with 2.3M left - an XT clone whose hard disk is as fragmented as one
|
|||
|
can ever get with 2500+ files. Before I did that, I decided to play
|
|||
|
it safe and backup the disk with FASTBACK. Mr. FASTBACK produced a
|
|||
|
"fatal error" after 23 minutes on the 37th diskette. No amount of
|
|||
|
coercing will make it continue. To heck with it - another software
|
|||
|
will get returned tomorrow. I crossed my fingers and ran DISK
|
|||
|
OPTIMIZER anyway - I had a few hours to kill while I work on the
|
|||
|
Thanksgiving turkey. You'd think I was an expectant father while the
|
|||
|
optimizer was running. For a long time, the machine seemed dead - the
|
|||
|
disk light might flash every 5 minutes or so - and this was just in
|
|||
|
Part One - disk directory reorganization. Many times, I came close to
|
|||
|
reaching for the Big Red switch but somehow resisted. I was determined
|
|||
|
to wait till Christmas for this one as I had absolutely no disk backup
|
|||
|
to speak of. After 93 minutes, the program said it was done. Indeed
|
|||
|
all the usual checks said I had a perfectly organized disk - no files
|
|||
|
lost and no fragmented files.
|
|||
|
|
|||
|
And, that, is how I spent Thanksgiving '85.
|
|||
|
|
|||
|
Next time you're on my BBS (before it gets fragmented again) and you
|
|||
|
notice that it is fast - remember these: (1) I am running two jobs at
|
|||
|
the same time via DoubleDos (I have thrown away MultiLink after two
|
|||
|
years and 4 versions, each one slower than the previous one), (2) I am
|
|||
|
running those on a "lowly" XT clone with absolutely nothing produced
|
|||
|
by Big Blue. If the board seems fast it is because of (1) PCBoard
|
|||
|
which is the greatest BBS software in the whole wide world today
|
|||
|
(thank you Fred Clark, Dave Staehlin, et. al) and (2) Disk Optimizer.
|
|||
|
|
|||
|
|
|||
|
TIPS ON USING DISK OPTIMIZER
|
|||
|
----------------------------
|
|||
|
|
|||
|
1. Unhide all files (other than IBMDOS and IBMBIO) and
|
|||
|
subdirectories before you begin
|
|||
|
|
|||
|
2. Do NOT run any memory-resident programs or multitasking or disk
|
|||
|
cache programs (Lightning is good but not now). Never, never have
|
|||
|
DPATH or similar data and overlay finder active.
|
|||
|
|
|||
|
3. In short, run Disk Optimizer on a "Virgin" system with only PC-
|
|||
|
DOS working.
|
|||
|
|
|||
|
4. I ran the program on the AT and XT clones under DOS 2.0, 2.1 and
|
|||
|
3.1 successfully but obviously cannot guarantee that it will work
|
|||
|
for all cheapie disks like mine. That's Softlogic's work after all.
|
|||
|
|
|||
|
|
|||
|
MY BEEF (or Turkey?)
|
|||
|
--------------------
|
|||
|
|
|||
|
You all know that I seldom endorse a product 100% (only Borland
|
|||
|
products are near perfection in my opinion). Alas, I have a few
|
|||
|
suggestions to Softlogic. Forget about your copy-protection. A program
|
|||
|
should never be copy-protected, least of all a disk utility program.
|
|||
|
Also, why don't you tell the Turkey holding his breath while watching
|
|||
|
a machine that seems dead what's going on? A few dots on the screen
|
|||
|
once in a while should be mighty reassuring. Let me also say that I
|
|||
|
love the price. Thank you for that.
|
|||
|
|
|||
|
|
|||
|
CONCLUSION
|
|||
|
----------
|
|||
|
|
|||
|
Should you buy it? I did, didn't I? You could do worse for
|
|||
|
$49.95. BUY IT. This program works. It ain't turkey.
|
|||
|
|
|||
|
Happy Thanksgiving, everybody!
|
|||
|
|
|||
|
|
|||
|
|