We didn't know that at the time. Early June when I ordered.
Actually I exercise my analytical skills every day as a software engineer and having heard every interview with Elon, I developed an appreciation of his belief in adaptive general software systems over specifically tailored ones (i.e his Self driving approach compared to the competition demonstrates that), I was convinced that Track Mode would be a general closed loop adaptive feedback system. I still believe it is designed that way. That's how I would have written it.
Exactly! I see no reason Track Mode would not work with AWD as is. It is kept away from them artificially to justify the premium cost of performance models. I believe Track Mode could even work with RWD with some slight modifications or even none if the software was written to detect capabilities.
This software design pattern is not new in Silicon Valley and has been around for 20+ years. For example, the video card and PC gaming industry works like this. Why do you think modern games work on basically all PCs with a wide range of capabilities? The game rendering engine adapts by providing different levels of rendering quality based on either detection or the caps reported by the video card. Who pioneered capability reporting in video cards? NVIDIA. What chips does the model 3 use on their two main boards? NVIDIA chips.
The only thing I see in the + as being more track capable are the larger rotors which will provide a longer time before brake fade. But the road conditions, temperature, driving characteristics, as well as other non hardware factors by themselves would change the threshold characteristics of the fade. So software would need to detect brake fade anyway even on P+. So even though the P- will fade a little sooner, the software should detect that and act accordingly for either car.
In June when we were ordered, we didn’t even know that track mode was coming to the Model
3 at all!
My internal analysis said that IF it came, it would be guaranteed on the 3P+. Not guaranteed on the 3P-.
I’m not a software engineer but your rationale sounds good. I’m sure you are competent at your work but everyone, myself included has to self evaluate daily to calibrate how smart we really are.
I would say there are two fallacies that can bust your argument.
Having Starcraft work on a wide range of computers is not the same as software on 4000lb cars. What happens when Blizzard fails to account for some adrenaline driver or some cuda driver? Slower frame rates? Graphic artifacts? Worst case is the game crashes on loading. People die when software in vehicles malfunction.
Tesla is leaving NVidia by the way with AP3. Using its own in house chip.
Second fallacy that will bust anyone is assymetric information. You can watch thousands of hours of Musk interviews and get insight into their development process - but unless you are in the decision meetings it doesn’t matter. All the reasonings in the world that make sense to you doesn’t happen because Tesla knows something you don’t. Or Tesla could just be making a bonehead decision. Either way your choice based on your analysis was wrong in the end.
Also, adaptive track mode while possible may not be profitable. Tesla HAS to spend time on adaptive EAP because they are betting on Tesla vision versus waymo geofenced and LIDAR approach.
Where I was wrong was I never would have imagined Tesla would roll PUP overnight into P and drop the whole package by 5K.
Simplify production - check
Drop whole package by 5K? - senseless. Drop it by 2-3K, minimal outrage.
I can’t think of why Tesla did what they did, unless they felt it only a drop of that magnitude can sustain sella.
Assymetric information, I can’t see their price elasticity of demand data.
One needs to avoid overestimating their knowledge no matter what field they work in. Someone was willing to wager I didn’t even have a college degree. Wasn’t hard for me to bet 10000/10000/100000 anything they wanted. I knew something they didn’t - I know if I have a degree or not and how to verify it.