When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
I jumped around in the video and only listened to a total of 2-3 minutes, but I do believe Tesla will be licensing all kinds of their tech to other companies in the coming years.
I jumped around in the video and only listened to a total of 2-3 minutes, but I do believe Tesla will be licensing all kinds of their tech to other companies in the coming years.
yea. He’s hard to listen to because he’s an obnoxious self-indulgent pompous little ***** even though he’s right on a lot of things. Looking at him is like watching a thumb wiggle and pretend to talk.
This is extremely interesting. But Elon has been talking about this AI self driving system for a while, and how all his vehicles will instantly increase in worth simply after he flips the switch and turns it on. So interesting to me that they spent probably billions on the current FSD Beta writing code but it won't be used in the end.
So interesting to me that they spent probably billions on the current FSD Beta writing code but it won't be used in the end.
As a programmer I can tell you almost everything I write ends up being tossed. I don't know of another way, you fail and fail and fail again until you get it right. Sometimes you chuck the entire thing and start over.
Alternative is to patch, work-around, cobble, and otherwise Frankenstein to try and make the whole mess work.
As a programmer I can tell you almost everything I write ends up being tossed. I don't know of another way, you fail and fail and fail again until you get it right. Sometimes you chuck the entire thing and start over.
+1 It's the only way.
Sometimes refactoring can get us by but eventually like you said, time to chuck it and start over.
I'm doing that right now when migrating to faster ARM chips. The newer chips allow me to chunck in serial stream data using DMA techniques in the Mhz's.
Sometimes the code is good but have to start over for hardware reasons too.
Originally Posted by LeX2K
Alternative is to patch, work-around, cobble, and otherwise Frankenstein to try and make the whole mess work.
And nobody wants to work on that mess or be the very unlucky guy who spends his life debugging others people's code. Haha 😂
Last edited by Margate330; Aug 28, 2023 at 12:37 PM.
And nobody wants to work on that mess or be the very unlucky guy who spends his life debugging others people's code. Haha 😂
OR, debugging your own code you wrote many years ago and totally forgot about! I've been there it was a strange experience I didn't remember writing one line.
OR, debugging your own code you wrote many years ago and totally forgot about! I've been there it was a strange experience I didn't remember writing one line.
BTW established OEMs have severe code bloat.
Got that right! Haha
We get in a zone and then the brain reboots. Lol
The other day I was looking at my Devicenet program I wrote a couple years ago and didn't remember one line of it.
Good thing I heavily use descriptive variable names because I rarely comment my code.
On the FSD, I predict they will make a case for necessary sensors buried in the road and connected to 5g to advance the FSD in a pseudo closed loop system.
I could never survive as a programmer...I have a mild form of ADD. My mind would be wondering off while looking at oceans of code. Dear God, just thinking about that is making my head spin
Good thing I heavily use descriptive variable names because I rarely comment my code.
Good friend of mine that taught me a massive amount drilled this into my head, document everything. He's an incredible coder I'm jealous. Some descriptors tend to be ridiculously long but damn does it help later.
So back to FSD if you want to nerd out for an hour
As a programmer I can tell you almost everything I write ends up being tossed. I don't know of another way, you fail and fail and fail again until you get it right. Sometimes you chuck the entire thing and start over.
Alternative is to patch, work-around, cobble, and otherwise Frankenstein to try and make the whole mess work.
tossing out code that isn’t worth debugging is one thing. But they consciously kept adding to the beta code while simultaneously creating AI version. Beta was just a place holder.
tossing out code that isn’t worth debugging is one thing. But they consciously kept adding to the beta code while simultaneously creating AI version. Beta was just a place holder.
"in the beginning nobody knows anything"
20/20 hindsight is not real if it was they would have never attempted FSD using heuristics. I can think of numerous analogies to explain how version 12 happened the simplest is you have to walk the wrong path to know it's wrong. If you don't walk it at all you'll never learn.
One thing is for sure. Dan O’Dowd is TICKED OFF about Teslas success and I’m loving every minute of it. 🤣🤣🤣🤣🤣. For those who don’t know he owns a company that supplies LiDAR which Tesla for good reason refuses to use.