2020.4.1 incoming

[MarcG]

Release notes on 2017 MX 75D MCU1 HW 2.5 coming from 2019.40.2.3 list
Additional vehicle information
Voice commands
Phone improvements
Camp mode
Backgammon
Voice keyboard
Save dash cam on honk
Driver profiles
Danish support.

Thanks. This list looks like what was included in 2019.40.50.x, which the vast majority already has. So I wonder what is really new for folks going from 2019.40.50.x to 2020.4.1?

 

[azred]

I wish this was true..unfortunately, I've had my Model 3 for a year, and I've never had any problem with laggy response to curves or drifting in the lane until late December this year. I drive the car exactly the same way I always have, but something has definitely changed. It's almost as if the processor is lagging and can't make micro adjustments as frequently as it used to. I have 2.5 (hopefully upgrading soon), but I understand folks on 3 are having similar problems.

My Model 3 is not ponging but S is annoying enough that my wife has asked me to not use autopilot when she is in the S. It is neither my imagination nor incompetence, just a recent release bug. Cars are on different releases.

 

[Nuclear Fusion]

just my opinion, but the driver of a vehicle in most states is responsible for passengers buckling in... adult or not. And while it’s easy to think that an adult can make their own choice for safety, an unbuckled rear seat passenger can kill you up front in an accident.

I disagree with this logic - why shouldn’t ALL adults in a car be fined for any unbuckled adults? The fine should only be given to the unbuckled adult, not the driver.

 

[Edgewood Dome]

I wish this was true..unfortunately, I've had my Model 3 for a year, and I've never had any problem with laggy response to curves or drifting in the lane until late December this year. I drive the car exactly the same way I always have, but something has definitely changed. It's almost as if the processor is lagging and can't make micro adjustments as frequently as it used to. I have 2.5 (hopefully upgrading soon), but I understand folks on 3 are having similar problems.

I noticed the exact same thing and constant, but slight, hunting (ping-pong) even when on a straight section of freeway. Then they upgraded my HW2.5 M3 (purchased with FSD) to the new FSD CPU last week and no more hunting or that feeling that the processor is lagging. Sure hope they upgrade my S90D w/FSD HW2.0 soon, the hunting kinda makes me motion sick.

 

[Nuclear Fusion]

I noticed the exact same thing and constant, but slight, hunting (ping-pong) even when on a straight section of freeway. Then they upgraded my HW2.5 M3 (purchased with FSD) to the new FSD CPU last week and no more hunting or that feeling that the processor is lagging. Sure hope they upgrade my S90D w/FSD HW2.0 soon, the hunting kinda makes me motion sick.

How long did the upgrade take?

 

[Pale_Rider]

New map update?

It’s just a full firmware update instead of an incremental update...

 

[theBurtReynold]

It’s just a full firmware update instead of an incremental update...

2020.4.1 clocks in at 3.0 GB

This has made me wonder ...

The final “feature complete” FSD update will trigger a significant financial event (recognition of all remaining deferred revenue, which is probably a lot of money at this point), meaning the stock will bump as a result.

As such, do you think Tesla will handle the release of “feature complete” no different than any other rollout — incremental, kinda whenever-they-want, as far as timing? (This would be equivalent to announcing a significant financial event during trading, so is it even allowed...?)

Or might Tesla put the feature set behind a feature toggle and more strictly control the timing? Perhaps even during an earnings report ...

 

[powaking]

After seeing hundreds of posts bout this, I'm 100% convinced the ping pong effect is user-induced, or at best, car-specific and not software related. Every release, people claim that all of a sudden, their car is overcome with the poltergeist of ping ponging. I've never seen it on any software version, but what I also know is that I've never been bothered by AP nags, because I know what the car is expecting. I apply slight torque to the wheel by resting one hand on one side of it, and it's flawless - the car pulls against the weight of my hand, maintains a center position in the lane, and as of recent releases, only occasionally goes wide when a lane divider disappears at merges. Meanwhile, people constantly complain about having to wiggle the wheel all the time and their car bouncing back and forth, and I suspect a connection.

I beg to differ. This has been an issue since 2019.40.50.1 and has not been corrected in 2019.40.50.7 (which is what I'm currently on). The car will go into ping pong effect for a few bounces then it will eventually straighten out. Sometimes when approaching a truck it will start the ping ponging. At times it gets annoying to the point where I would just turn it off. Hard to say its driver induced when it hasn't done this in the last 24,000 miles and immediately started since a deployment of a specific firmware version.

 

[kkillebrew]

After seeing hundreds of posts bout this, I'm 100% convinced the ping pong effect is user-induced, or at best, car-specific and not software related. Every release, people claim that all of a sudden, their car is overcome with the poltergeist of ping ponging. I've never seen it on any software version, but what I also know is that I've never been bothered by AP nags, because I know what the car is expecting. I apply slight torque to the wheel by resting one hand on one side of it, and it's flawless - the car pulls against the weight of my hand, maintains a center position in the lane, and as of recent releases, only occasionally goes wide when a lane divider disappears at merges. Meanwhile, people constantly complain about having to wiggle the wheel all the time and their car bouncing back and forth, and I suspect a connection.

Mine does it certain situations. I disagree that it’s user induced, it’s not, it’s triggered by road/lane width and dividing line spacing.

 

[Nglewis92]

I suspect that the 2.5 platform with all the ongoing changes and feature adds is simply nearing processing capacity or the system bus is overloaded. I have no real evidence to back up that assertion other than ongoing feature adds (relative speed merge, one pedal driving, more robust visualization of surrounding traffic including aspect, etc.) are the only changes I'm aware of that could be impacting my autopilot experience. I have tried a couple of things that MAY help (turned down acceleration to chill, reduced lane change aggressiveness, and turned off other services I thought might be processor intensive. It feels like things improve slightly but still degrade over time, and I have to do a hard reset to see that slight improvement in behavior again. That said, it could just be placebo effect. The problem is intermittent, frankly, but when it's there, it's THERE for real...understeer followed by oversteer through curves, late and more aggressive braking, and drifting in the lane. The fact that it varies from car to car would suggest it is very processor load / data bus load dependent. In the same way that everyone doesn't have to reboot their PC due to sluggish performance every day, not everyone is noticing laggy behavior in their MCU every day. My $.02 worth.

 

[BR666T]

Looking at TeslaFi, I'm guessing they halted the rollout of 2020.4.1 as well as 2020.4. Here's to 2020.4.2!

 

[OPRCE]

umm how do i delete the pic?

Hit "report", give reason and request Mod to do it?

 

[ScubaMal]

My 2018 Model X is now installing 2020.4.1. I have had the HW3 installed in late November and I'm still getting a very bad ping pong effect while on autopilot. I took it into the service center for camera calibration and it's still weaving between the lane lines. It was rock solid before Version 10.1.

 

[Puddles]

might want to blur location on that pic

Why would you need to do that?

 

[Mardak]

The final “feature complete” FSD update will trigger a significant financial event (recognition of all remaining deferred revenue

It won't be all at once similar to Smart Summon. It seems that revenue is recognized based on owners who have received the feature, so revenue was still deferred for those in regions that did not get Smart Summon initially. So similarly, if "feature complete" is US-only initially, Tesla would not be able to recognize "all remaining deferred revenue."

And if people who purchased FSD haven't gotten a retrofit yet but "feature complete" FSD software otherwise would have been available to the owner, the act of service replacing the computer would effectively become a revenue generation event going forwards.

 

[CertLive]

I have stopped using Autopilot since the christmas updates the breaking is dangerous. I was hoping 2020.4.1 was going to fix it.

 

[Leeboy22001]

I have this update pending, but nervous about updating.. already have #batterygate #chargegate and #regengate !.. what next !??

 

[CertLive]

You see here is my take on this after 3+ months of ownership. Why would you release updates that make features that worked just fine worse? I mean I still don't have working voice commands or an autopilot that is able to pick up speed without breaking or so some rather odd things with turnings off dual carriageway. These were introduced over the top of things that were working just fine. It starts to make me think the same people who test the software updates are the ones that check the cars before they leave the factory. I have never seen this amount of disregard for peoples safety my car is now worse off keeping in lanes. Why is my car different from any other HW3 SR+ model 3 they should be testing the features before going public on? Are the sensors suites all made so different that the software just made my car a lot worse? Can anyone explain what is Tesla,s thinking on this because its to the point of being unsafe.

 

[CertLive]

I now cant tell if an update is going to give me real improvements or is going to fiddle with the autopilot to the point they never explain and I have to find out when someone rear ends me.

 

[sixela]

I disagree with this logic - why shouldn’t ALL adults in a car be fined for any unbuckled adults? The fine should only be given to the unbuckled adult, not the driver.

You may disagree with the logic, but in my country both the driver and passenger refusing to buckle up are responsible, both in criminal and civil suits (and insurance companies will refuse to cover injuries incurred or caused by the passenger, in case of accidents). And they both get fined. One for not buckling up, the other for driving after failing to force the other for buckling up.

The reason is exactly to discourage the driver from saying it's "not his problem". In that way, I find the nagging helpful: you can force the passenger to buckle up just to stop the nagging, even if he disagrees with the law and is adamant he has the right to put everyone in danger.