Page 2 of 2

Re: Getting back into all this...

Posted: Sat Oct 03, 2009 1:51 am
by sry_not4sale
walkercam wrote:Progress report:

Finally got the serial connection to open up. The problem was being caused by this section in Serial.py:

Code: Select all

        try:
            self._connection = conn = serial.Serial(
                s.port,
                baudrate    = s.baudrate,
                bytesize    = s.bytesize,
                parity      = s.parity,
                stopbits    = s.stopbits,
                timeout     = s.timeout,
                xonxoff     = s.xonxoff,
                rtscts      = s.rtscts,
            )

        except serialutil.SerialException, msg:
            raise comms.interface.CannotconnectException, msg
What I did was take it back to the most simple form of the serial.Serial command and changed it to serial.Serial(0). This worked. I slowly added back in the other operators until it shat itself. Turns out it doesn't like 's.port' for some reason. I hard coded it to be:

Code: Select all

        try:
            self._connection = conn = serial.Serial(
                0,
                baudrate    = s.baudrate,
                bytesize    = s.bytesize,
                parity      = s.parity,
                stopbits    = s.stopbits,
                timeout     = s.timeout,
                xonxoff     = s.xonxoff,
                rtscts      = s.rtscts,
            )

        except serialutil.SerialException, msg:
            raise comms.interface.CannotconnectException, msg
And now it works. I turned on the TA board and packets of data started popping up in the tuner window :P

Aaron - Any ideas as to why the tuner might be unhappy with the s.port operator? I assume it works fine in linux?

Cheers guys.
Hey mate,

I'm guessing it doesn't like the port number being cast as a string, which I imagine happens when it is loaded from the config file. I will look into the code :)

Re: Getting back into all this...

Posted: Sat Oct 03, 2009 3:13 am
by sry_not4sale
Hey mate,

This should be fixed in the devtrunk branch now ;)

Cheers,
Aaron

Re: Getting back into all this...

Posted: Sun Oct 04, 2009 8:51 am
by Fred
Good hacking Cam!! Good turn around time on the fix also :-)

Fred.

Re: Getting back into all this...

Posted: Sun Oct 04, 2009 4:48 pm
by Fred
Cam, I believe the tuner is looking for an id of 261 not 301 for the datalog, grab the code from github (using git!) and get the compiler etc installed and build a pre-release of 0.0.20 to run against. You asked for me to send you mine, but that makes no sense in the medium term so you may as well get setup properly now. If not, I'll be emailing you binaries every few minutes at some point :-)

Fred.

Re: Getting back into all this...

Posted: Sun Oct 04, 2009 8:55 pm
by slacker.cam
Thanks guys. Your replies have been very helpful.

I'll get the compiler set up tonight and have another go at it :)

Re: Getting back into all this...

Posted: Mon Oct 05, 2009 12:04 am
by Fred
Sweet man, good to see you in on the action :-) Big things coming from this project, hard to remembr that good things take time though :-)

Fred.