rtc@masscomp.UUCP (Richard Carling) (04/18/85)
-----
I posted a previous message about IBM AT disk problems a few months ago
and I thought it appropiate to give an update.
First I should say that I have not taken the system in lately since
I can't spare to loose if for a week or two. It is usable, I suppose.
The repair place will only do one thing at a time (makes sense unless your
the customer), so I presume I would need to bring it in 3 or 4 times
before they would replace the drive. I already had the drive replaced
once so I am not to excited about that.
I have tried the unconditioned and conditioned format a few times off the
advanced AT disk diagnostics. This always gives about a week of use before
severe disk read failures occur.
I tried the DFixer program. Sure enough it found a lot of bad sectors
around the area where the standard format found them. And I was able to
run for three weeks before I started getting disk errors again (a new
record!). Why IBM can't provide a DFixer like program or generally better
software support is not clear to me. Must be an awful lot of dead wood
at IBM. DFixer has its own problems though, they say you can run it on
a disk which has files. You CAN do that but since DFixer doesn't tell you
the filename in which the bad sector occured, only the block number,
that isn't of much use. So you end up only using DFixer on a clean
formatted disk even though their advertising would have you believe
differently (you can run the Norton utilities to find the file name
but it becomes a real pain. Why the DFIxer folks can't finish the
product is beyond me).
Anyway, back to the story, after three weeks of no disk read errors I
was starting to get pretty excited, gee, maybe the CMI drives are OK,
maybe I don't have to keep going to the repair shop.
No such luck, the odd thing was that one night I used the system just
fine, everything was OK. The next day, almost every file I had written
the night before produced a read error that day. Other files
were OK. Anybody have any ideas??? Disk head slip??
Since then I have been getting roughly one read error for every two hours
of use (pretty obnoxious, comes to roughly 1 file in 10 that is created/saved).
I always try to rename the files to badblock.nnn so that they don't get
recycled. I have a feeling, the seek is off, creating bad spots on neighboring
tracks as well as the current one, so even though I isolate the particular
file, other files are also getting damaged.
Now, I am always amazed when I hear CMI and IBM saying users don't use the
machine correctly. You've got to be kidding. I had an XT for 2 years without
a problem. I also use an AT purchased back in August which I have never had one
disk error on. All I want to do is edit, compile and run my programs.
Why does no other manufacturer have this problem (probably cause nobody else
uses CMI drives).
I also hear rumors that it might be DOS 3.0, but then why is this only a
problem on some machines, not all. I'll have to wait for DOS 3.1 and then
try my luck.
And of course CDC/CORE claims that there is a bug in the ROM on the disk
controller. They won't tell IBM the specific problem because this gives
them a market edge. Well that's just swell, thanks guys.
I am starting to wonder about the chances of a lawsuit against IBM.
Gee, I remember those crazy days when I actually thought IBM manufactured
well built products, silly me!
So all you folks who don't have AT's, just think how luckly you are.
As for me, I suppose I will get very very mean when the warranty gets near
its expiration and hope IBM stops playing games and finds a solution
to this problem in the meantime.
Naturally these are my own opinions.
Richard Carling
p.s. If any of you folks from IBM want to send me a new drive
and controller, I would be happy to accept it. Just call
617/ 275-4282 for an appointment.