Firmware Update Storm?

[Aeg07046]

I am wondering what is happening at the Tesla Firmware team. Generally Tesla has been issuing firmware updates every 4-8 weeks. But this month (August 2017), according to ev-fw, they have issued 4 firmware updates:

17.32 9xxx
17.32.6
17.32.17, and now
17.34

This seems excessive and not well controlled, meaning even after a later update is released, cars of all flavors continue to receive older firmware. Just yesterday and today, different cars got all four updates. Seems very weird. it is possible that ev-fw is wrong, otherwise, why would you let four different firmware versions be installed on cars during the same 48 hour period. Some (unlucky?) soul got 17.32 9xxx today, a firmware version that is a month and four versions old.

Some of you software development people can weigh in as to why a FirmwareStorm of this type would be proper and why same model cars would get four different versions over a two day period.

Separately, someone mentioned on another thread (don't have time to find the link) that they were told by Tesla Customer Service that they have decided that "non-major updates" will not be rolled out fleet-wide and only major updates will. That, if correct, seems to be a new policy.

If anyone thinks that all this is well-planned, managed, orderly and proper, I would love to hear your thoughts.

My concern is that this is not a PC firmware update where the worst thing happens is that it bricks the system, this is a four thousand pound mass going 90 miles an hour with this firmware controlling its steering. If AS suddenly steers left at 90 mph, regardless of how firm one's hands are on the steering wheel, it is nearly impossible not to hit a nearby median. (Think of a scenario with no Autosteer, driving 90 mph next to a median, the passenger unexpectantly reaches over and pushes the steering wheel left.) I am incredibly impressed in Tesla's QC on firmware where I have heard of no accidents caused by Autosteer. Does the FirmwareStorm endanger this great streak?

 

[Sir Guacamolaf]

AP1 here, I'm still on a very old firmware version. I don't care really since the new updates have nothing new in them.
I'd rather be on an older version for the reasons you mention, unless there is a notable new feature .. but as an AP1 owner, I am not sure what really groundbreaking new stuff to expect at this point anyway.

 

[boonedocks]

I am still waiting on one newer than a month and 1/2 ago. Haven’t received one of those. My AP2 seriously sucks right now.


#StillWaitingOnMySilkySmoothOrSomethingSpecialElon

 

[outie]

Think of a scenario with no Autosteer, driving 90 mph next to a median, the passenger unexpectantly reaches over and pushes the steering wheel left.) I am incredibly impressed in Tesla's QC on firmware where I have heard of no accidents caused by Autosteer. Does the

Actually, if I wasn't holding the steering wheel driving on the carpool lane or fast lane causing overrides I'd have hit the median many times over and it would be a ghost typing this.

 

[boaterva]

That first release is really 17.32, and the other two are additions to the week 32 one. The 'main' one for a week doesn't have a third number, it's the useless hash code (seven digits).

As for not everyone will get a release, I can't see that point unless it's for a specific purpose (like the P90D fix). Testing on a subset seems like a mess.

I'd like to hear someone steeped in development weigh in on this, also. From supporting software and other IT projects, this seems hard to support when people are all over the place, version-wise.

 

[kort677]

I am in the process of updating right now and I'll post which flavor I received on ev-fw.com

 

[RichardL]

My understanding is that Tesla often do this to try out different algorithms in real world situations - not necessarily changing the actual behavior of the car - they once said they sometimes run the new stuff in the background to compare it's suggested behavior to the current software or even the driver's actual behavior, then it can report back to the mothership.

So multiple small releases can mean they are trying out several new approaches.

 

[kort677]

I am in the process of updating right now and I'll post which flavor I received on ev-fw.com

replying to self, this is the version I received today, 17.32.6ca28227, no notes

 

[kort677]

My understanding is that Tesla often do this to try out different algorithms in real world situations - not necessarily changing the actual behavior of the car - they once said they sometimes run the new stuff in the background to compare it's suggested behavior to the current software or even the driver's actual behavior, then it can report back to the mothership.

So multiple small releases can mean they are trying out several new approaches.

isn't it great being a beta tester?

 

[bmah]

isn't it great being a beta tester?

It's more like this:

A/B testing - Wikipedia

My AP1 car is running an update from about 6 weeks ago, and that's fine by me. No updates == no regressions.

Bruce.

 

[mrElbe]

At least one of the newer versions is reportedly fixing the MPH to km/h conversion problem.

 

[brianman]

At least one of the newer versions is reportedly fixing the MPH to km/h conversion problem.

Yay!

 

[kort677]

It's more like this:

A/B testing - Wikipedia

My AP1 car is running an update from about 6 weeks ago, and that's fine by me. No updates == no regressions.

Bruce.

call it what you may, we, the car owners are apparently being used as testers, functionality of what version you were given isn't really relevant unless you've gotten a buggy version

 

[MarcusMaximus]

At least in the past, we've also seen different version numbers going out to different types of cars(I.e one for AP1, one for AP2, one for ludicrous only, etc). It's possible some of these are just the same update for different configurations.

 

[JohnnyG]

Here's a possible theory... (please note the fixes are made-up just for this example)

Version:    Fix:                                 Cars Pushed To:
17.32       Bug with the UHFS                    Only cars with UHFS
17.32.6     Bug with Ludicrous                   Only cars with Ludicrous
17.32.17    Bug with heated wipers               Only cars with heated wipers
17.34       Bug with Pano & vented seats         Only cars with Pano OR vented seats
17.34.x     Bug with Rev2 inverter & Rev1 TPMS   Only cars with Rev2 inverter OR Rev1 TPMS
17.34.xx    Bug with SAS on RWD only             Only cars with RWD that have SAS

You get the pattern here...

 

[boaterva]

Here's a possible theory... (please note the fixes are made-up just for this example)

Version:    Fix:                                 Cars Pushed To:
17.32       Bug with the UHFS                    Only cars with UHFS
17.32.6     Bug with Ludicrous                   Only cars with Ludicrous
17.32.17    Bug with heated wipers               Only cars with heated wipers
17.34       Bug with Pano & vented seats         Only cars with Pano OR vented seats
17.34.x     Bug with Rev2 inverter & Rev1 TPMS   Only cars with Rev2 inverter OR Rev1 TPMS
17.34.xx    Bug with SAS on RWD only             Only cars with RWD that have SAS

You get the pattern here...

Yeah, OMG is all I can say. This would be fine if people actually got more releases more often. But we know how *that* works!

 

[Shock-On-T]

Here's a possible theory... (please note the fixes are made-up just for this example)

Version:    Fix:                                 Cars Pushed To:
17.32       Bug with the UHFS                    Only cars with UHFS
17.32.6     Bug with Ludicrous                   Only cars with Ludicrous
17.32.17    Bug with heated wipers               Only cars with heated wipers
17.34       Bug with Pano & vented seats         Only cars with Pano OR vented seats
17.34.x     Bug with Rev2 inverter & Rev1 TPMS   Only cars with Rev2 inverter OR Rev1 TPMS
17.34.xx    Bug with SAS on RWD only             Only cars with RWD that have SAS

You get the pattern here...

There's definitely something to that.
For example, 722130c currently has 77 AP2 cars and 0 AP1 cars. That can't be by chance.

 

[csanders90D]

I am wondering what is happening at the Tesla Firmware team. Generally Tesla has been issuing firmware updates every 4-8 weeks. But this month (August 2017), according to ev-fw, they have issued 4 firmware updates:

This month's behavior doesn't look much different than their past history. Your 4-8 week number isn't right at all, Tesla has always done far more versions than that. Here's a count of how many firmware versions were released each month this year:
January: 6
February: 4
March: 5
April: 6
May: 3
June: 6
July: 3
August: 4

August is slightly under their average of 4.6 versions per month. An individual car doesn't see most of those updates, so it feels like they are much less frequent, until you pull the full data from ev-fw.

 

[wdolson]

Looking at the firmware tracker, it looks like some versions are regionally targeted. I just got 2448cfc this morning (I have an AP1, refresh S90D in the US).

I did notice a couple of things that are different. One issue that was a minor annoyance was when I back the car out of the garage, I fold the mirrors and I kept having to remember to unfold them again. I noticed today they auto deployed when I put the car into Drive. I also noticed the mirrors unfolded when the car unlocked instead of when I opened the door.

The last is a perception thing, I'm not sure if it's true, but it feels like regen is a little stronger now. For minor slowing, it feels the same, but when I completely take my foot off the accelerator, it feels stronger.

 

[Josh88]

I got 16.32.6 last Wednesday when I took my MX2.0 in for service. By Saturday my mirrors stopped folding in when I locked the car and I could not fold them in with the button on drivers side door either. Took the car into the service center today and they said it appears to be a new firmware issue. They had to reset the "module" to fix the issue. Tech said another MX was coming in tomorrow with the same issue. Of course I left with a new firmware update (still downloading). He also stated owners could not reset the module on there own.

Here is my concern, I used to think updates were a plus however I asked the service advisor what this would have cost out of warranty and he said it would be $175 because they had to diagnose the issue. My problem is why should we have to pay Tesla to submit bugs in the firmware? Isn't being unconvinced enough?

Also I called tech support first to verify it wasn't a known firmware issue and he stated that since the button didn't work it had to be hardware.

To be clear I am still under warranty so I didn't have to pay the $175 but I would have been very upset had I been out of warranty.