[net.micro.mac] Write protect tabs

gus@Shasta.ARPA (01/31/86)

> > 
> > Then how come my Sony disks work? The write-protect mechanism on those
> > diskettes doesn't include a hole going through the diskette - there is a 
> > small tab that can be moved back and forth, but you can't see through the
> > diskette in either mode.
> > 
> > 	--MKR
>  No, you don't seem to understand.  The tab that you slide back & forth IS
>  the hole that I'm referring to. (I never said the hole was in the actual 
>  media).  When the tab is moved so that you can see through the CASE of the
>  microdisk, then the disk is locked.
> 
> 						Robert J. Hammen

No! YOU don't undestand! Some early Sony disks don't have a writeprotect hole
at all, just as the previous poster described. Unless an LED can
differentiate between thick blue plastic and thin blue plastic, the switch
MUST be mechanical

gus@Shasta.ARPA (01/31/86)

> > > 
> > > Then how come my Sony disks work? The write-protect mechanism on those
> > > diskettes doesn't include a hole going through the diskette - there is a 
> > > small tab that can be moved back and forth, but you can't see through the
> > > diskette in either mode.
> > > 
> > > 	--MKR
> >  No, you don't seem to understand.  The tab that you slide back & forth IS
> >  the hole that I'm referring to. (I never said the hole was in the actual 
> >  media).  When the tab is moved so that you can see through the CASE of the
> >  microdisk, then the disk is locked.
> > 
> > 						Robert J. Hammen
> 
> No! YOU don't undestand! Some early Sony disks don't have a writeprotect hole
> at all, just as the previous poster described. Unless an LED can
> differentiate between thick blue plastic and thin blue plastic, the switch
> MUST be mechanical
> 

No, I didn't understand. A recent message describes what someone found when
closely examining his 400K and 800K drives. I took apart my 400K external
drive and found thing to be just as described. Two mechanical sensing levers
controlling two photointerrupters with very visible red LED's.

I also brought out Larry Kenyon's old disk test program. (The one with all
the buttons, switches, scroll bars, etc...) and the hardware chapter of
Inside Mac. I started playing around with the "Interface test" window and
found things to be just as described in Inside Mac. I was able to eject
the disk by manipulating the right control lines which were set up as check
boxes. Nowhere in the documentation does it say anything about turning these
LEDs on or off or disabling the write protect. Undauted, I noticed, that
four control lines provide access to 11 registers documented in Inside Mac.
The disk test program seemed to indicate the existence of three more - Eject,
a second Sides and a /Exists. None of these were really interesting since
they were read-only. What WAS interesting, however, was that there were only
FOUR writable registers - DIRTN (Set head step direction), STEP (Step the
head), EJECT (Eject the disk) and MOTORON (Turn on/off the disk motor)

I also noticed that one of the lines was always 0 (SEL). If there were any
extra writable controls, they would probably be accessable by setting SEL
to 1, with CA0 and CA1 providing up to four more registers, and CA2 being
the data bit to store there.

I observed no reaction when experimenting with these four combinations which
would indicates that it is NOT possible to defeat the Write Protect in 
software. 

Note that my tests were not exhaustive, and I could have missed something.
Also, since I didn't write the disk test program, I don't know what internal
protections (if any) there might be to avoid a possibly damaging combination
of events.

I still remain convinced, however, that it is impossible for a fully
functioning drive to write to a write protected disk. The proof is in the
pudding. The scratch disk that I tested all this on still booted perfectly
when I rebooted the machine, even though I was experimenting on track 0, 
which contains directory and boot information.

Now I really feel silly, because we just had Larry Kenyon talk at the SMUG
Developers meeting yesterday. If he couldn't resolve this issue one and for
all, no one can! Unfortunately, I forgot to ask him!

						Gustavo Fernandez