From: Jim Brain (brain_at_jbrain.com)
Date: 2004-05-27 09:31:26
>I see, you would prefer to have clean layers in the protocol. Instead of
>escaping of 00 payload bytes, you would like to escape 00 bytes also in the
>parameters of commands. I don't know what you would achieve with that.
>Of course, there must be a resynchronization mechanism, in case of
>intermittent communications failure. I'm using the BREAK condition as a
>signal to reset the microcontroller to idle state.
>
>
If the system gets too far out of whack, command data could get
misinterpreted as a command code (in your example, 00 LISTEN 00 00 00 00
FF (32 bits) coudl get mangled into 00 LISTEN 00 00 00 00 CMD_FF
(whatever this is). At least with escaped command data, sync issues are
less of an issue. Minor nit.
>I didn't consider that, because I have the feeling that the serial bus is
>slower than the 38400 bps line. Most executable programs for the
>Commodore 64 have been compressed, and there are no long sequences of zero
>bytes. I picked zero as the escape byte, because testing against zero is
>easy on the AVR. Testing against some other constant requires that the
>value is loaded to one of the registers 16..31.
>
>
True, but I tried FF, and I got into issues where the system might get
off a byte (bad code), and then it gets worse. FF FE or F2 02 or
something worked much better. I know 00 is easy, but 00 just shows up
so much, In some files, it could double the size of the file.
>I would like to make it possible to upgrade existing C2N232 units, and to
>save trouble in circuit board design. You're right, there's no need to
>stick to a smaller controller. But I like challenges, and I'm not yet
>hitting the limits of the 2313 too hard. (Are you familiar of Levente
>Hársfalvi's RS-232 to 1351 mouse adapter done with a PIC? That's a real
>challenge.)
>
>
Well, I think some standardization on protocol might be useful.
>10 years ago, I was dreaming of something similar: a programmable device
>for the serial bus. Currently available microcontrollers are making the
>dream possible at an affordable cost.
>
>
As was I, but I dind;t have the tools or the expertise to do it. Now, I
have both.
>>I could try. I was waiting a bit until I put the JDos routines in
>>before I laid out the diagram in a package, to make sure I actually
>>capture all of the relevant states.
>>
>>
>
>It'd be good to draw one diagram without JDos and another including it.
>
>
That was the plan, but have them as overlays.
>What does ?ST say after such a failed LOAD attempt?
>
>
66 (with my code, not sure on base 64)
>I don't think that there is EOI handshake for bytes sent under ATN.
>I concluded this after reading the VIC-20 KERNAL routines. Come to
>think of it, EOI handshaking for ATN would make things very complicated
>when you're talking to multiple devices. And there is no EOI handshaking
>for the first ATN byte, is there?
>
>
I am getting the impression that EOI is never used when ATN low.
As stated in another email, I have server code running. save
"/home/brain/incoming/stuff",13 is working
Load needs a bit more thought. It works now (I have a hardcoded app
stored on the AVR for testing), but I need some commands to send to the
AVR to tell it to wait for data, or not to bother, because file cannot
be found.
JIm
--
Jim Brain, Brain Innovations
brain@jbrain.com http://www.jbrain.com
Dabbling in WWW, Embedded Systems, Old CBM computers, and Good Times!
Message was sent through the cbm-hackers mailing list
Archive generated by hypermail pre-2.1.8.