Poll : Which is better - FSD Beta 10.9 vs 10.8

[EVNow]

Note : Public poll.

Consider how the two releases performed on your routes. Which release did you like better ?

Result from the last poll below.

 

[momo3605]

Note : Public poll.

Consider how the two releases performed on your routes. Which release did you like better ?

Result from the last poll below.

View attachment 757513

Oh darn I didn’t see the old thread. But agree with the majority there. 10.8 much better than 10.6.1. I’m pretty happy with the overall improvement of FSD between 10.2 and 10.8. Haven’t played with 10.9 yet, but I did get it in the second batch this time. Got 10.8.1 in the very last batch.

 

[EVNow]

Oh darn I didn’t see the old thread. But agree with the majority there. 10.8 much better than 10.6.1. I’m pretty happy with the overall improvement of FSD between 10.2 and 10.8. Haven’t played with 10.9 yet, but I did get it in the second batch this time. Got 10.8.1 in the very last batch.

I think 10.8 is the only release for me that was worse ! Rest have been "slightly better".

 

[d21mike]

I voted sightly better. I wish it was more. The issue of lane changes for absolutely no reason still happens. Even on "Chill". Really wish they would at least indicate why they are doing it. Like with NOA on freeways where it says "Changing lanes to follow GPS" or "Changing lanes for slow traffic" (or something like that). With FSD Beta it will just change lanes for no reason and not notice of why. You might ask why is that a problem? The main problem is due to traffic it might not be able to get back to the original lane so it can follow GPS. I have said since I started FSD Beta why not have a option to say only change lanes when following GPS? Then they added Chill and 2 other options but it made no difference.

 

[S4WRXTTCS]

I went for slightly better.

There is a lot of noise when testing.

Things will happen on one drive, but don't happen on another. Where we can't really know for sure if the FW version played any role.

 

[momo3605]

I went for slightly better.

There is a lot of noise when testing.

Things will happen on one drive, but don't happen on another. Where we can't really know for sure if the FW version played any role.

i usually try to pay attention for areas where the car misbehaves, like there's a right turn where the car tries to go inside the bike lane, instead of recognizing and going wide, and in 10.8, it still does it, but it figures it out quickly and jerks out of the lane. Still not ideal, but noticeable change. Another location where my car just wants to swerve into the other lane and swerve back even if there are no other cars around. Still happens on 10.8, so i'm going to keep an eye on that.

 

[JulienW]

Had to (unfortunately) vote 10.8 is MUCH better for me (hell 10.2-10.8 were all MUCH better to me). Have documented some of my problems in Actual Beta experiences thread.

So far 6 short drives with 5 times it has driven into the (clearly marked with double yellow lines) opposing traffic lanes. Plus all the other disengagement. Can't understand why 10.9 is sooooooo much worse than 10.2-10.8 for me. PLEASE let drive #7 be an improvement.

 

[momo3605]

I'm going to go with 10.8 being slightly better than 10.9. I finally got a chance to take it on my test loops, and while there are some improvements in 10.9, the biggest regression i've seen is consistently stopping 10+ feet before the intersection (not every intersection, but 2 compared to 0 in previous version)....i'm guessing the new native intersection assignment code might have bugs in it. In addition, i've noticed jerkier left turns.