On Dec 3, 2004, at 9:24 AM, [EMAIL PROTECTED] wrote:

Your results are interesting, if not confusing. Do you think this is a bug
(& therefore should be entered in bugzilla)?

IIRC, it looks like 3 bugs:

Char read commands come up short on XP for reading 256 chars.
Multi-char read commands throw away data after LF on OS X.
Codes 10, 11 and 12 are garbled in reading uint1 on OS X.
Multi-uint1 read commands sometimes come up short on OS X based on garbage at end (?).
uint1 read commands input garbage if the data is not there.



Fortunately, since the phone system only uses a certain set of chars, I
can filter out all of the junk, & make this work for my needs. BTW, I'm
using both the keyspan usa-28 & usa-28x on OSX, & com1 on the PC. The
results are the same with either keyspan adapter.

It looks like you might get garbage if you try to read a uint1 and none is there. The time might come when one of the garbage bytes looks like a valid letter.


It might be that my short data for 256 char reads does not apply to single char reads.

There are several reported bugs for serial on OS X; you will have to work around those.

Windows serial I/O is blocking (up to 2 seconds) and OS X serial I/O is not (except for a weird OS X feature in opening the modem where a 3 second delay is forced).

I think someone (Sarah?) has reported better luck with usa-28 than with usa-19.

I suspect that Rev is not setting some terminal parameters on ports, leaving them as they are, on OS X.

Dar

****************************************
    Dar Scott Consulting
    http://www.swcp.com/dsc/
    Programming Services
****************************************

_______________________________________________
use-revolution mailing list
[EMAIL PROTECTED]
http://lists.runrev.com/mailman/listinfo/use-revolution

Reply via email to