Il 17/09/2015 15:04, Dennis Lee Bieber ha scritto:
On Thu, 17 Sep 2015 12:00:08 +0000 (UTC), alister
<alister.nospam.w...@ntlworld.com> declaimed the following:
I can see the data being transmitted snowballing & running away in a +ve
feedback loop very easily.
Especially if a few of the remote devices are configured to ECHO
data... <G>
No ECHO.
Main thing I'd probably change is... Since the COM port identification
is already being provided during initialization of the handler object, why
maintain a list of (com, handler) pairs, and the subsequent subscripting --
just save the com port as an attribute of the object.
One could also make a copy of the object list in the start method, and
at that point, scan the list and remove that one's own identity. That would
remove the need for always testing "is the object I'm about to send to
really me?"
Ok, they are optimizations, but I don't think they solve my issue.
--
https://mail.python.org/mailman/listinfo/python-list
