4 ms·

Rivians software is quite good. The number of times I've wished it had carplay was low and is now zero that they moved to Google maps instead of their own in t

Rivians software is quite good. The number of times I've wished it had carplay was low and is now zero that they moved to Google maps instead of their own in the past 6 months (the maps themselves and the interface were totally fine, but traffic prediction and were bad at times). They are fairly good at changing when change is needed. The latter suggestion is IMHO a really bad idea for companies the size of rivian. it would cost a lot of time and energy to integrate carplay and android auto very very well. Most cars are already a mishmash of software pieces, so replacing 60% of that with a simple carplay integration might buy them a lot. My wife's car is the first model year from that manufacturer with carplay and it's just horribly done. Duplicate voice assistants, and carplay only uses 75% of the screen (this is not fixed by the software changes in ios26 around this. This is the manufacturer not giving a shit) just to start. Despite all that it is still better than the smorgasbord of shit that it is without carplay. That is not true of rivian. It is already good software that is well integrated. if they moved, the same has to remain true. The upfront cost of moving would be large, not provide tons of benefit to existing customers (it's not in the top 5 for most owners), and they aren't large enough to have parallel software teams, so it would slow down development on things owners and near-purchasers care more about. In the end good product development is at least as much about saying no as it is about saying yes. saying yes to everything and making them all options does not actually make for good products (or softwaare) in the end, they seem pretty good about holding themselves accountable to whether they’re succeeding in their various bets or not. and doing something reasonable with that info. As long as they continur that I'm not too worried, because it’ll lead to the CarPlay if it needs to.

labcomputer 1y ago
> The latter suggestion is IMHO a really bad idea for companies the size of rivian. it would cost a lot of time and energy to integrate carplay and android auto very very well. This can't possibly be true. CarPlay and Android Auto are just an MPEG video stream to the car's display, plus a simple API to send taps back to the phone. Hell, you can buy a complete CarPlay/AA head unit for $200 at Best Buy. I mean, if it is true that Rivian can't even grab an OSS video decoder and run it on their system... that makes VW's due-diligence (before throwing billions of dollars at Rivian for software) look very shaky.
- matthewdgreen 1y ago
  The part of CarPlay I’m unhappy with is the (in)ability to plan EV trips with charging stops. Adding charging in a good native nav interface is simple: you enter the route and it picks chargers for you if the distance exceeds current range. In standard CarPlay this doesn’t happen automatically. Presumably it could, but in addition to charging station search and planning, CarPlay would need an interface to take current battery level and some other preferences that would tune the algorithm for specific cars. Not sure if that exists, but the cars I’ve used haven’t exhibited it.
- DannyBee 1y ago
  That is if you use it to replace the audio functionality, sure. If all your center console/stereo did before was a broken spotify interface, and now has lots of functionality, then yes, carplay is much better. As i said, that would constitute really shitty integration for rivian, becuase that isn't what their center console was like. In rivians case you would now have a massive ton of center console functionality and driving display that worked and looked differently from the apps piece (2x differently if they did both carplay and android auto). If you wanted it to look and act right WRT to the rest of the car you would have to do the carplay ultra thing, and write ios apps for your other functionality so that it actually all felt the same. Do the same for android auto, too. Its the same with tesla. Can I just suggest you spend 5 minutes looking at one (you don't have to drive it) and then see if you think replacing 40% of the center console functionality with something that looks and acts very different than the rest of the car would really qualify as "good integration". None of what i referred to as the cost is the literal technical cost of displaying the video stream (Their UI is built in unreal and they have plenty of gpu power, so this part is trivial for sure), but instead, of building out the apps, functionality, and other pieces necessary to make the car software look and act coherently. The reason VW threw money at them, other than wanting a hedge against Tesla, is precisely because VW has shitty software that feels and acts like a mess (my wife's car is a VWAG car, but not a VW), and Rivian doesn't. In VW's case, i'm sure all the pieces are written by different suppliers, so i can forgive them a little bit, but still, it's a mess. It's ironic then to suggest that they didn't do their diligence because you think it's as easy as just displaying the video stream to make it good, when VW is paying rivian for precisely the opposite! It's also interesting to try to discuss this with folks who have strong opinions but don't appear to have ever used one. Overall, the best way i could put it is: Imagine you have an ipad, but you have to use a bunch of apps that are only available in android all the time. So 60% of your ipad screen is taken up by an android emulator that runs all the time. The other 40% of the screen is IOS. The two pieces know nothing about each other. Do you believe this would be a good experience? I don't. Would it suddenly be good if you could make the android emulator full screen? No, that wouldn't change a lot for me, though it would probably be better in the relative sense. It would be good when i didn't have to run the android emulator, and all of my apps felt like they aren't fighting each other or ignoring each other :)

labcomputer 1y ago

> The latter suggestion is IMHO a really bad idea for companies the size of rivian. it would cost a lot of time and energy to integrate carplay and android auto very very well. This can't possibly be true. CarPlay and Android Auto are just an MPEG video stream to the car's display, plus a simple API to send taps back to the phone. Hell, you can buy a complete CarPlay/AA head unit for $200 at Best Buy. I mean, if it is true that Rivian can't even grab an OSS video decoder and run it on their system... that makes VW's due-diligence (before throwing billions of dollars at Rivian for software) look very shaky.

matthewdgreen 1y ago
The part of CarPlay I’m unhappy with is the (in)ability to plan EV trips with charging stops. Adding charging in a good native nav interface is simple: you enter the route and it picks chargers for you if the distance exceeds current range. In standard CarPlay this doesn’t happen automatically. Presumably it could, but in addition to charging station search and planning, CarPlay would need an interface to take current battery level and some other preferences that would tune the algorithm for specific cars. Not sure if that exists, but the cars I’ve used haven’t exhibited it.
DannyBee 1y ago
That is if you use it to replace the audio functionality, sure. If all your center console/stereo did before was a broken spotify interface, and now has lots of functionality, then yes, carplay is much better. As i said, that would constitute really shitty integration for rivian, becuase that isn't what their center console was like. In rivians case you would now have a massive ton of center console functionality and driving display that worked and looked differently from the apps piece (2x differently if they did both carplay and android auto). If you wanted it to look and act right WRT to the rest of the car you would have to do the carplay ultra thing, and write ios apps for your other functionality so that it actually all felt the same. Do the same for android auto, too. Its the same with tesla. Can I just suggest you spend 5 minutes looking at one (you don't have to drive it) and then see if you think replacing 40% of the center console functionality with something that looks and acts very different than the rest of the car would really qualify as "good integration". None of what i referred to as the cost is the literal technical cost of displaying the video stream (Their UI is built in unreal and they have plenty of gpu power, so this part is trivial for sure), but instead, of building out the apps, functionality, and other pieces necessary to make the car software look and act coherently. The reason VW threw money at them, other than wanting a hedge against Tesla, is precisely because VW has shitty software that feels and acts like a mess (my wife's car is a VWAG car, but not a VW), and Rivian doesn't. In VW's case, i'm sure all the pieces are written by different suppliers, so i can forgive them a little bit, but still, it's a mess. It's ironic then to suggest that they didn't do their diligence because you think it's as easy as just displaying the video stream to make it good, when VW is paying rivian for precisely the opposite! It's also interesting to try to discuss this with folks who have strong opinions but don't appear to have ever used one. Overall, the best way i could put it is: Imagine you have an ipad, but you have to use a bunch of apps that are only available in android all the time. So 60% of your ipad screen is taken up by an android emulator that runs all the time. The other 40% of the screen is IOS. The two pieces know nothing about each other. Do you believe this would be a good experience? I don't. Would it suddenly be good if you could make the android emulator full screen? No, that wouldn't change a lot for me, though it would probably be better in the relative sense. It would be good when i didn't have to run the android emulator, and all of my apps felt like they aren't fighting each other or ignoring each other :)