🚀 go-pugleaf

In <4e2shb$jjt@lantana.singnet.com.sg> englf@singnet.com.sg (felix eng) writes:

>yo all!  can someone help out on this ? pls post to this chap at
>negcjfk@ibm.net. 

>thanks a million.
>=== felix eng ===



>From: negcjfk@ibm.net
>To: englf@singnet.com.sg

>Subject: Floppy drive problems

>Machine: (Pentium, 90 MHz, 24 RAM, two hard drives, one floppy 3.5", no
>compression, 
>not networked).  Specifically, we are suddenly  having trouble accessing
>floppy 
>diskettes. 

>1. The system has been in operation since June of 1995 with no apparent
> problems with the floppy drive.  The other day I was performing a search of 
>some older floppy disks for a particular document I had written a couple years
> ago.  Because disk reads are always been slow under presentation manager,
> I usually perform such searches in a DOS or OS/2 window, using the A:\dir 
>commmand. While doing this, I noticed that the contents of one disk apparently
>matched the contents of the current disk in the drive.  I put in another disk
>and
> typed Dir.  That disks contents was again the same to an earlier disk that
>had 
>045 error, - cannot find sector requested.  At this point, I couldn't make the
>error dialog go away no matter what option I choose.  After repeated efforts 
>and hitting the ctrl-esc combination a few times, I was able to bring up the 
>end program dialog which said A: drive is not responding to the system, etc, 
>etc. do you want to end it?  I did and then my entire desktop went away and 
>was then recreated after about 20 seconds.  I recreated the problem using the 
>drive object for the floppy.  Whatever the contents is of the first disk you 
>access, susbsequent reads will reflect that instead of the true contents of
>the 
>disk currently in the drive.

Sounds to me like there's a bug in the WPS; I know others had some problems
with floopy drives as well, but I don't know if there was an APAR set for it
or not.  You may want to try one of the FixPaks...

What it sounds like to ME, though, is this: When you eject a floppy disk, the
drive is supposed to send a signal stating that the disk has changed.  Some
floppy drives don't do this, or they go bad and the line change signal is
never sent.  It sounds like this is what's happening, so OS/2 doesn't bother
to re-read the disk - it just displays what is currently in memory.  You can
get around this in DOS by hitting <Break> (not <Ctrl>-C) and then trying it
again.  You may want to try this first in a DOS window.  If it works, chances
are that you are going to want to replace your floppy drive.  If it doesn't,
then chances are OS/2's DOS doesn't act like old DOS and try a FixPak.

>2. When the pentium is first booted, it seems to read a good a floppy, but
>only
> once.  For example, iF you ask it to read a fresh pre-formatted floppy then 
>every floppy after that will read empty whether they are empty or not.  It is
>the
>n sometimes possible to get out of this by opening other directories on the
>hardrives 
>and then trying the A: drive again.  Perhaps by doing this we are flushing out
>some 
>sort of 'cache' that remembers the contents of the disk in the floppy drive? 
>Further 
>testing has confirmed that after the machine first boots up it will always
>read a good 
>floppy correctly.  Putting a new floppy in after and then accessing it either
>by the 
>drive object or the command line displays the contents of the first diskette,
>and then 
>you may be caught in the loop witrh the sys0045 error dialog popping up.  You
>sometimes 
>can shut the 0045 error dialog down if you put the first floppy back in the in
>the 
>drive. 

Sounds like a line change problem like I described above.  I don't understand
the 'sys' error you keep getting though, unless it is just a bug in WPS.

>3. ometimes we get sys0279 - extended attribute errors.

???  Are you running under FAT or HPFS?  FAT and EA don't always cooperate.

>4. Another weird thing is that if a file is copied to a fresh floppy disk on
>the
> Pentium, that disk is sometimes rendered unusuable in the 486.  The opposite 
>is sometimes true as well.  Sometimes if a file is erased on a disk in the 486
> and then brought to the pentium the file still shows up on the disk, however
> the reverse is not always true.  However, sometimes if a file is copied to a
> floppy on the Pentium and subsequently erased, and then the disk is placed in
> the 486 and a file copied to it, the file does not get written to the floppy,
>although DOS says that it was (no sound comes from the drive indicating that 
>something has been written.  Putting a new floppy in the 486 and copying the 
>same file will work however. We do not feel the problem lies with the 486 as 
>it works fine with any disk as long it hasn't had OS/2 copy files to it.  
>And that does'nt always cause problems. Floppies have always freely switched
> between the two machines the past. 

Again, sounds like the floppy drive is bad.  If the head isn't calibrated to
the same spec as the 486's, then they won't be able to read each other's disks.
Has someone dropped this drive in the mud recently?

>5. The most irritating symptom of this problem is when the floopy disk is
> continully accessed every five or ten seconds.  In this case, either the 
>sys0045 message keeps popping up which for some reason disables the mouse
> until you press return, or the drive light just comes on every few seconds as
> the drive attempts to read the floppy.

>The above has not always been the case, OS/2 has worked for eight months or
> so without floppy disk problems, now suddenly???  Can anybody help?  I hope 
>we don't have to reinstall of the operating system from the CD.  We do have
>the
> operating system backed up to tape from an earlier time when the system 
>appeared to be functioning normally.  Would a tape reinstall of certain
>directory
> of the operating system do the trick or should we just recopy the tape file
>over
> the operating system overall. Please help!!!!!!

I don't think a reinstall would help at all.  Replace the floppy drive and
see if it fixes it.  Call IBM tech support about the dialog boxes that never
go away and if there's an appropriate FixPak.  If not that, sorry, I can't
tell.  At least, not without more details.

>Jon Kaminsky

--
Christopher Curtis, Sun SysAdmin - http://www.ee.fit.edu/users/ccurtis
Florida Institute of Technology  - telnet bofh.engr.wisc.edu 666
Melbourne, Florida  USA          - Member, Team OS/2