🚀 go-pugleaf

In message <na.3777bc470f.a500d0steve.pampling@argonet.co.uk> S Pampling wrote:

> Now strangely enough, on PC's the problem was the bios NOT the hardware. Old
> PC's which you wouldn't expect to run bi-directionally (and don't for
> Windoze etc) will run Laplink.

*Any* printer port is capable of bidirectional communication. Even
unidirectional ones (well, OK the Beeb can't but it wasn't a true
Centronics port). To facilitate a reverse channel with
unidirectional ports, you need to use 'nibble-mode' where the 5
'status' inputs turn into 4 data inputs and <summat else>. I would
assume that Laplink will fall back to this mode if it can't get a
true 8-bit reverse channel. 

> Anyone feel like having a look at RO3.1 to look at the printer port
> routines?

No. :-)

> or did we have some unidirectional buffer hardware between the
> port chip and the physical parallel port in the old machines?

Yes. We did. Somewhere. Maybe even in the port chip itself.  IIRC.

> Anyone got a circuit diagram handy?

Nope. But IIRC it's available at Acorn's ftp server.

Cheers,

Pat.

-- 
.=============================================================================.
| Patrick Herborn,      |  pat@nanobyte.demon.co.uk  |  Primary Mail Address  |
| Electronic Engineer.  |  pah@sn2.ee.umist.ac.uk    |  Alternative  Address  |
'============================================================================='
Immensely inspiring tagline follows....
... Drop your carrier ... we have you surrounded!