[Oberon] Re. ATAPI Zip drive;

[easlab at absamail.co.za]

> At Thu Aug 11 09:22:14 2005 Chris said,
> > Will anyone want to invest time developing software for
> > redundant [Zip drive] hardware ?
> 
Peter E. wrote:-
> Since that discussion, an ATAPI 100 MB Zip disk 
> has served as a portable workspace with zero errors.
> 
Yes Iomega-Zip 100 MB is amazingly reliable, for a 
unprotected 'floppy'.

> It elliminates the need to move files over the net
> and reconcile various copies.  Saves me at least a 
> few hours each week.  Having all working files 
> immediately available is a great benefit.
> 
> The only problem evident is the minor one of needing 
> two Mount commands in Oberon.Text.  Seems to be a  
> minor idiosyncrasy of the Zip drive with no serious 
> consequence.
> 
Are you saying that NO [not Aos] can read/write Zip ?
IMO only the 'parallel port' version is relevant for
moving data between boxes.

> The SCSI Zip drive tested earlier works well in a 
> Macintosh.  Therefore the problem must have been with 
> Adaptec card or more likely with the Adaptec7 driver.
> 
> As far as I know, Adaptec7 is still considered viable.
> I do not understand the driver well enough to trouble
> -shoot but can cooperate with anyone interested.

Apparently since the beginning of floppies, it was common
to do multiple [try again] reads.  This can be a software 
solution for systems with poor hardware reliability.
A system which doesn't 'crash' when the fd0 detects an 
error is what is needed.  This is easily acheived, and 
probably of more value than investigating Adaptec7.    
But I find the NO 'funny assembler' off-putting.

Re. "Occasional data transfer between boxes",
since most hardware has still got a RS232 port[s], if
the boxes are near to each other NO V24.Panel will
communicate with DOS, Linux ...etc via a null modem.

But a problem of "can't open the port" usually arises.
This can be 'superficialy hacked-around' by:  
    V24.Stop(1) .

Interestingly the same hack [action which is not well
understood by the implimentor] clears the 'lock' which
I've been complaining about, which happens when 
the dialer/ppp has an abort.

It now seems that the subtle problem [solved by all M$
users by re-booting] which I've been chasing for ages, is in
 the V24 module[s] and not the ppp-suite ?
Or it's the task-swithching between them ?
-----------------
So, IMO Iomega Zip was a star at the time, since
all boxes had a centronics port.  But now flash-memory
is more appropriate to occasionally transport data.
Except that USB seems to still not have stabilised ?

I don't think any body wants to invest time with Adaptec7.

== Chris Glur.