Page 7 of 9
Re: FreeEMS-Tuner Development Diary - Don't Post!
Posted: Wed Feb 18, 2009 9:58 pm
by sry_not4sale
The nasty test "framework":
http://github.com/srynot4sale/freeems-t ... b2/test.py
Contains lots of dirty hacks to get around threading
Re: FreeEMS-Tuner Development Diary - Don't Post!
Posted: Sun Feb 22, 2009 10:03 pm
by sry_not4sale
Pushed a new File comms adaptor. This takes a binary dump file and process it on connection.
Re: FreeEMS-Tuner Development Diary - Don't Post!
Posted: Mon Feb 23, 2009 4:49 am
by sry_not4sale
Push logging to file and config options
Re: FreeEMS-Tuner Development Diary - Don't Post!
Posted: Wed Feb 25, 2009 8:57 pm
by sry_not4sale
So...
Uploaded my port of Fred's C serial parsing code to python for shits and giggles.
Also a pre-commit git hook for updating version.py. Has a small bug so not quite ready for prime time.
And finally some additions to the debug frame, including a tidy up of comms/protocol plugin version names and including the user's config details!
Re: FreeEMS-Tuner Development Diary - Don't Post!
Posted: Sun Apr 26, 2009 9:21 pm
by sry_not4sale
So, Monday is here and I am back into it - as promised
Mainly fixes to support Python 2.6 (as I upgraded to Ubuntu 9.04) and bugs when using a fresh install of the tuner:
Code: Select all
822d707dc6eaf5520988c265f9a0ddffb877b5a2
Fixed un-escaped ampersand
d8fa0a3780a3f1195c44ab9ae660c8635af4a897
Removed references to the old logger (except in debugTab - todo)
fcf25463dec9ed1b24e9e6b29a01fb216a98576e
Cast payload to a string before setting
81f642e936a3feaa12c1eea0d340df1179d8704c
Tidying up docstring format
446425152b2d7f2743f44e70c79f08f66e9d0910
Fixed misspelling of isLogging method name
d6e42ffa33d9a0fc2b0e8a2d758f04a4e7b69bce
Updated default comms plugin to be Test
4415ff69237fb0bfd8b85cd6e714b158b601f89c
Fixed bad config file path and bug in new installs where no config file exists and debugFrame crashes
845d716b203eebb422af2330c09cecacc9de36f5
Fixes to make the tuner compatible with python 1.6
I have also started on threading code to revive crashed worker threads
Re: FreeEMS-Tuner Development Diary - Don't Post!
Posted: Tue Apr 28, 2009 11:09 pm
by sry_not4sale
Some more features I completed yesterday:
* Added ablity to switch on and off gui diagnotics
* Added check for datalog status
And a bug fix:
* Hack around receive thread returning empty payloads as a list
Re: FreeEMS-Tuner Development Diary - Don't Post!
Posted: Fri Aug 28, 2009 11:04 am
by sry_not4sale
So... after a few months off...
Started looking around for recommendations on python test frameworks, and documentation generators / code visualisers.
I want to get some decent test coverage happening, and document alot more to get the new dev barrier of entry down.
I would love to start the MegaSquirt comms plugin also! Anyone have the MS hardware and are willing to coordinate some testing with me?
Re: FreeEMS-Tuner Development Diary - Don't Post!
Posted: Mon Aug 31, 2009 3:32 am
by sry_not4sale
Had a chat with Fred about stuff, next thing on the list is fixing some incorrect payload id's and getting the form for writing data to memory locations finished.
Also read up about a much nicer (and looks to be better performing) method of processing frames in binary streams in Python! Will start a thread about it soon, although I think some comprehensive test coverage of the existing serial processing code is in order first.
Re: FreeEMS-Tuner Development Diary - Don't Post!
Posted: Tue Sep 01, 2009 5:13 am
by sry_not4sale
Finished reading up about co-routines in Python. Will be interesting to see how much performance we can gain from them.
Re: FreeEMS-Tuner Development Diary - Don't Post!
Posted: Wed Sep 02, 2009 10:26 pm
by sry_not4sale
Sooo finally got back into some code last night...
Fixed a few miscellaneous bugs and removed some unused code.
Updated the app title and status bars to show more useful information, including the comms plugin you are using and what Serial port / Input file you are connected to.
Also renamed Comms Logging to Comms Recording to avoid confusion with application logging. Found a bug with the dialog bog for selecting a file to record too - the Cancel button not canceling the Start Recording operation... hehe
I got 90% of the way through having the Real-time Data tab updating directly from incoming data log packets - just need to finish adding a random data log packet to the Test adaptor, and have the code check to see if the tab is visible before attempting to update all the gui (otherwise it could be a bit of a performance hog) - and to update from saved values when it comes into focus.
If anyone has any idea on improving the current real-time data interface - please comment here:
http://issues.freeems.org/view.php?id=103
Also created an
issue regarding the Firmware needing some request/response packets for returning the current status of data logging (ON or OFF) and assigned it to Fred.