search

rpendleton / mazda-trip-tracker

CMU tweak for the Mazda Connect system that allows you to automatically track various trip statistics
4 stars 2 forks source link

Installation instructions missing #1

Open sergesyrota opened 8 years ago

sergesyrota commented 8 years ago

This looks awesome! But I'm not sure how to build it, and there are no mentioning of that in the readme. Is it supposed to be compiled on Mazda system? Or is there a way to compile on linux machine, and then just copy to the car?

Thanks!

rpendleton commented 8 years ago

It's easiest if you compile it on a Linux machine and then copy it to the car. You'll need to use a cross-compiling toolchain, such as the one I've created at rpendleton/mazda-toolchain.

I have to warn you that none of this is an easy task though. Additionally, this isn't a finished project. The program has to be started manually and causes some bugs with the navigation system. (I think the navigation bugs might be related to my unreleased autostart feature though, so I'm not positive.) I'm planning on finishing it eventually, since I want to use it to make taxes easy for me, but I won't be able to do that until I'm on break from school during the summer.

sergesyrota commented 8 years ago

I understand that I'm wandering in uncharted territory :) Thanks, I'll take a look at that. If I'm successful, I might even submit a PR for this issue to write up a manual.

and causes some bugs with the navigation system

What bugs are those? Just want to make sure I'm on the lookout when I get to running the program. And I'm assuming autostart thing is not committed to this repo? Not seeing anything that might look like that, but wanted to double check.

rpendleton commented 8 years ago

The autostart isn't committed to this repo since it involved making changes to the CMU (and no changes to this project). Essentially, I noticed that if I started my trip tracker too early, the navigation wouldn't work the rest of the drive. I'm not sure if I'm just connecting to the GPS device too early or if something more complicated is going on, but it's pretty easy to reproduce if you leave it enabled for a few days of driving.

If you do run into that problem and can't get around it, it's not actually that hard to fix. I just haven't had time to do it myself. Rather than using the GPS device to get location data, there's a method you can call on the d-bus to get the current location. You'd have to poll it every second rather than read from it as data comes in, but it's worth looking into. The data is slightly less accurate if I remember right, but not by much. And by the time it gets to d-bus, it's already been parsed, so there wouldn't be a need to mess with NMEA data.