racekarl
Well-Known Member
First, there is no cost for CarPlay (pretty sure the same is true for AA). Second, there are various levels of integration possible, and the OEM implementing has a choice about how far they take it. The idea that it could "exponentially" increase testing is untrue and unhelpful hyperbole. CarPlay (and presumably AA, but I am less familiar with that) can be pretty well "sand-boxed" and not present a substantial regression risk to other areas of the car.Rivian's software isn't mature enough (probably never will) but they are improving much faster than Tesla back in the day.
Every time you add another major UI like AA or CP, it exponentially increases the testing cycle, it's not just that CP works but rather running function A in CP doesn't break or alter something else somewhere unexpected. What happens if you add NewFeature-Y? Got to make sure it didn't break or alter anything else, including AA or CP.
You have to make sure everything works as expected because it's a car, lives are at risk. Software bugs come up in mysterious ways. Problems are compounded when you mix and match software you did not write or may not have source code access to. For example: I really do not want my car interface to go blank if my streaming video breaks.
That is enough reason not to expand your testing and integration footprint, otherwise you will not get updates every 1-3 months, instead it will be once a year, with minor incremental changes.
Even if the above is addressed, why would Rivian want to share the car data and telemetry with others? What are the costs involved to add CP or AA? I have a pretty good idea of what Rivian is doing to add Cast support and it's not cheap (dev + $$$) - we definitely will be paying for the privilege to stream video, at least with a data plan.
At the end of the day it is also essentially screen-casting, so the costs and risk are not likely that far off from this Google announcement.
Sponsored