From: Marko Mäkelä (marko.makela_at_hut.fi)
Date: 2003-03-13 14:26:26
On Thu, Mar 13, 2003 at 01:42:54PM +0100, Ullrich von Bassewitz wrote:
> Are you thinking of a SLIP implementation running completely on the C2N232,
> or are you talking about using the C2N232 as a RS232 device?
I'm thinking of using the C2N232 as an RS-232 interface. That would
require the C2N232 to be capable of reading commands from the cassette
interface. The biggest problem is that on some Commodore models (the
264 series or the B series or both; I don't remember), the CASS SENSE
line is input only.
So, only CASS WRITE can be used for signalling the C2N232 from the
Commodore side. When the Commodore requests service by changing CASS
WRITE to active state, it must see whether the C2N232 has data available,
before assuming that the C2N232 is prepared to receive any data. The
C2N232 could use the CASS SENSE line for signalling its status, and
CASS READ for handshaking. But it'd be even nicer to do without the
CASS SENSE line, since in that way, the protocol would work also on the
Tangerine Oric-1 and other computers designed to use an audio tape recorder.
One solution to eliminating the CASS SENSE line is that the C2N232 would
delay its CASS READ acknowledgement to the CASS WRITE pulse by a few
microseconds if no data is available. Alas, this would impose a real-time
requirement to both ends of the tape interface.
The receiver on the Commodore side could be interrupt-driven (on CASS READ).
While the sender is active, interrupts and DMA must be disabled, if my
pulse-width-based protocol
(http://www.funet.fi/pub/cbm/crossplatform/transfer/C2N232/firmware/timing.png)
is to be used.
> The next major version of cc65 will replace the current static RS232 module
> (which supports the swiftlink only) by a loadable driver, so at least the
> latter solution (using the C2N232 as a serial device) should be much easier
> to implement and use than it is now.
Excellent. Will it then also support the built-in 6551 of the plus/4
and the SuperPET?
Marko
Message was sent through the cbm-hackers mailing list
Archive generated by hypermail 2.1.6.