As a Mazda owner, I’m deeply disappointed in Mazda itself about this DCMA takedown.
I used this frequently to ensure there was enough gas in the car (early morning commute and a shared car), alerts to see the windows are up before rain is supposed to arrive, remote unlocking to avoid a long hike back to the car when I left the keys inside, and remote starting as the weather starts to get colder (again, early morning commutes).
Their app is a poor implementation at best.
I am too a Mazda owner, and I am very disappointed in their move. I did read through the DMCA takedown, and I fail to see how their Apple and Android apps are infringing on the Python-based library. It’s not the same code base at all. The code that “provides functionality same as what is currently in Apple App Store and Google Play App Store” is NOT a copyright infringement! I can’t believe GitHub didn’t even read the takedown itself.
Another disappointed Mazda owner. I and my family have purchased 3 in recent years. I sent a message to Mazda in their app, telling how much they have disappointed this customer and encouraged them to reach out to the Home Assistant team as there is a mutual benefit to both.
I encourage others to send a message telling Mazda North America how you feel. It’s simple to do in their app. Maybe we all can help them see their mistake.
It would be nice if Wired or Arstechinca would pick up this for bigger publicity.
Yeah unfortunately if the person who posted the code doesn’t “counter-notice” and if the host of the code doesn’t push back and say “your request doesn’t meet the requirements for a dmca takedown”, then companies will continue to get away with abusing these takedowns. Unfortunately the costs and anxiety of “no I’m actually in the right here” are born by the individual dev not the massive corporation.
And yeah replicating functionality is not copyright infringement. Kind of disappointed that the GitHub process is so lax that there wasn’t more pushback. There was obviously some, since there was the mention that this was a revised request. But like, they didn’t identify any actual content copied…
But I know the host has pretty limited push back abilities under the law. If there is actually no copied code, I would have hoped for a counter notice (which puts it back up and it’s then on the complainer to prove more thoroughly) but I totally understand it would be intimidating.
(Not a lawyer obviously.)
Reach out and complain, car buyers and Mazda owners.
Bizarre decision on their part, and a good reason for me to avoid buying a car from them.
Looks like you can still download pymazda from pypi using the direct links:
https://files.pythonhosted.org/packages/6f/39/148a82a2f38cf770cf943482cf051af60d2aae607b5194260ec027d7ff0e/pymazda-0.3.11-py3-none-any.whl
https://files.pythonhosted.org/packages/c8/98/6da75ea926e4b3566823fcf164114b932d9c4de915e1412f1e0fa016ab58/pymazda-0.3.11.tar.gz
I doubt this will get added. Integrations are supposed to have an external library, which you have done away with.
At what address can we write to them to complain? If there are thousands of us this could change!
I also find this an absurdity and have reported it to a well-known German electronics blog. Perhaps the topic gets thereby again somewhat more range and the pressure on the responsible persons with Mazda increases.
I noticed a potential licensing issue when it comes to relying only on an external package source host like pypi.
If home-assistant receives a GPL source code request(for example due to many of the libraries in the home-assistant docker images being GPL licensed) wouldn’t the source code to all library dependencies(including the source for historical dependencies like pymazda) need to be provided for 3 years(due to the combined home-assistant application effectively falling under GPL requirements) from the date of distribution of the docker image?
I wouldn’t be surprised if this company would not be the reason of this whole issue:
It must be hurtful if somebody uses a service in a way that it can be efficient and useful and mostly for FREE as it is open-source.
That 300 API calls per month gives you total 10 calls per day in average. Just think how often do you check your fuel or battery level per day.
300 per DAY would give you an update every 5 minutes in average.
If you are intending to use the API as a device tracker, your update rate would be around 1 call per minute to have an accurate track of the device.
I use a custom component for the Hyundai I have and it only calls a few times a day (30 mins to the server and 240 mins to the car I think) and only from 06:00 to 21:00. Can’t really use it as a device_tracker but great for keeping an eye on fuel and battery level and mileage. If you call it too much, the battery in car goes flat and I’m sure Hyundai would get fed up of all the API calls. Not sure if that’s the same with Mazdas ?
I think someone requested an integration for that a while back
I think it must have started with complains from Mazda users their service was not available many times.
They did probably an investigation and found that not only their apps are using it, but others as well. Instead of fixing the performance issue they blocked other access.
I guess its typical for companies that do not care for their consumers. I wanted a Mazda 6 at some point, but I stick to Toyota for now 
