Firmware 7.1

[dirkhh]

I don't think this has anything to do with Model 3. Model 3 is several years out (yeah, yeah, 2017, yada yada, whatever. Model X was 2013 based on that.. moving on)
I think they are chasing a lot of bugs across their increasingly complex and diverse hardware footprint. I hear from insiders that their grasp of "software distribution discipline" is still at a startup level. With >100k cars on the road that's a problem.

Anyway, just saw the first report for 2.4.35. I'm calling this a wild and active beta test, chasing way too many bugs in way too many directions. My recommendation would be to decline updates until this is settling down. There is no way they have done any reasonable Q&A on the six different builds that they have pushed to end users in the last week. They don't have the man power to do that. This is completely out of control.

 

[dirkhh]

Any idea why this is possibly happening? I don't remember seeing anything like this before with this many builds.

No, we've never had anything even remotely close. As I said. Be careful out there. You are being used as testers for experimental software. I am now declining updates (again) on both of my cars.

 

[Lex]

Hmm, no .13 or .14 builds yet for my MS either, but 2.12.126 is solid. I like the new cold regen (for me it is more regen overall and when it's hovering around freezing there's some now, vs. none before).

Are any of these below addressed in the newer builds ? If not I'd say they are testing something or finding issues, either way I can wait

- driver-side handle present code still needs some work as it's more than just present but locking / unlocking 1 door vs. all. (eg. I can sneak through trunk -- but I am tortured by this as I like the 1-handle thing more than the 1-door unlock thing so maybe the answer here is detaching locking from presenting).

- music seems to constantly skip or degrade -- be it on Slacker, TuneIn playing FM stream, or FLAC from my USB. It should not be a surprise to get through a song without a hiccup.

- 1-point nav system is laughable / terrorizing.

- sad few 3 or 4 voice commands barely ever work properly.

Guide me to.. Navigate to... Drive to... which is it today, I give up.

These are minor to me though.. the car is a delight, nice long drive today with an SC stop that was purely to aid comfort and speed. Given the choice I am not a hypermiler

 

[LetsGoFast]

No, we've never had anything even remotely close. As I said. Be careful out there. You are being used as testers for experimental software. I am now declining updates (again) on both of my cars.

For all we know, they are testing changes to the audio player. We don't really have enough evidence to suggest any effect on any driving systems.

 

[dirkhh]

For all we know, they are testing changes to the audio player. We don't really have enough evidence to suggest any effect on any driving systems.

100% agree. We don't no one way or the other. I'll leave the beta testing to the people who signed up for it. Thankfully the firmware tracker makes this easy. Once I see a build with hundreds of installs I will wait for the next "new upgrade" notification on my phone and then upgrade. Until that time I'm out.

 

[SomeJoe7777]

I agree completely, they need to get a discipline and formality with regards to the software release trains. Major tech companies like Cisco have a very formalized release process that keeps QC at a high level.

Cisco produces two software trains: General Deployment (or Mainline) release, and Early Deployment (or Tech release). ED releases have the latest features that may still be under development, whereas GD releases are supposed to be mature, nearly bug-free releases.

Any release that is found to have a bug of significant impact in any feature is moved to a "Deferred" status, where it is no longer available at all.

Within each software train, maintenance releases are put out to address very small, non-impactful issues -- these releases do not go through full QC but only through verification that the issue is addressed.

Main releases within each software train go through full QC, which includes regression testing and full testing suites on every hardware configuration.

Tesla needs to adopt a similar system. The only releases that should be pushed OTA are General Deployment, Main releases that have gone through full QC. Service centers should be able to push GD maintenance releases that address specific bugs like the dual-charging bug in 2.12.126, addressed in 2.13.77.

ED releases should only be in the hands of the early release program participants.


Right now, Tesla appears to be pushing many maintenance releases OTA, and none of them have gone through full QC. Each maintenance release changes more code, so the entire release's code base moves further and further away from the last build that was subject to full regression testing and all hardware platforms. This ADDS bugs to the release instead of squashing them.

 

[dgpcolorado]

Any idea why this is possibly happening? I don't remember seeing anything like this before with this many builds.

Could it have anything to do with scrambling to redact the future model info that was leaked?

 

[ccharleb]

You're talking about the 17" (I presume). That remains white-outlined-in-blue (I think, I'll check tonite) in 126-with-new-maps.


I'm talking about the instrument cluster. Mine is now blue instead of the yellow/orange shown in the (from the web) image below:

View attachment 113224
http://cdn.teslarati.com/wp-content/uploads/2014/07/IMG_4306-e1406339678562.jpg


It looks more like this now:
View attachment 113225


I'm not sure which I like better, but it's definitely different from what I've driven with the last 3+ years.

Got my first MAP (Garmin/Navigon) update this morning. Anyone wondering if they currently are with the latest, the line telling me to exit an highway, for example, is now blue instead of orange.

 

[Mark Z]

7.1 (2.14.35) does have one major change for Model X:

The passenger door no longer closes when pressing the brake pedal to start the car. The touch screen can be used to close any door just as before. The driver door still closes when the brake pedal is pressed.

This was a needed improvement as passengers were occasionally getting a "love tap" by the door.

 

[BertL]

Received notice a couple hours ago and just applied 7.1 2.14.66 to my S90D. It took less than 30mins, although the estimate was the standard 1hr 20min. Release notes do not appear any different than before. There are no specific new things called-out as improvements.

Unlike some owners, I'm electing to continue applying all firmware releases provided by Tesla, especially after having seen some improvement in my several-months-old 90KwH Rated Range issue in the last couple of drops, and after reading yesterday that Tesla plans to introduce proactive changes to prevent charging if a potential short is determined like was the case in the isolated Norway (Supercharger) fire. I personally am not technically involved enough to know better than Tesla on such things, and Tesla does not provide detail on everything included in each update for me to really know what's there anyway, so I'll just take the possible good with any bad that comes my way. (I can always vent here on TMC, and make myself feel a little better, right?)

 

[GlmnAlyAirCar]

I agree completely, they need to get a discipline and formality with regards to the software release trains. Major tech companies like Cisco have a very formalized release process that keeps QC at a high level.

I agree that Configuration Management has to be a nightmare right now at TM. It's tough enough at a more mature software company and by all evidence we have seen their process is not that mature. Given the number of configurations (two major models, many various option configurations within each model, and frequent hardware changes) means there are probably very few identical cars on the road.

Hopefully they are bringing in experienced people with software process and CM experience.

 

[msnow]

Received notice a couple hours ago and just applied 7.1 2.14.66 to my S90D. It took less than 30mins, although the estimate was the standard 1hr 20min. Release notes do not appear any different than before. There are no specific new things called-out as improvements.

Unlike some owners, I'm electing to continue applying all firmware releases provided by Tesla, especially after having seen some improvement in my several-months-old 90KwH Rated Range issue in the last couple of drops, and after reading yesterday that Tesla plans to introduce proactive changes to prevent charging if a potential short is determined like was the case in the isolated Norway (Supercharger) fire. I personally am not technically involved enough to know better than Tesla on such things, and Tesla does not provide detail on everything included in each update for me to really know what's there anyway, so I'll just take the possible good with any bad that comes my way. (I can always vent here on TMC, and make myself feel a little better, right?)

Any change (up or down) on your Rated Range with the new update?

 

[Todd Burch]

Also received 2.14.66 this morning. No changes noticed yet.

 

[Doug_G]

I received 2.14.66 this morning. Weird thing is I received 2.14.50 yesterday. Never had back-to-back updates before.

 

[taldric]

I just received 7.1 (2.14.66) this morning as well. Maybe they are standardizing on this build?

 

[BertL]

Any change (up or down) on your Rated Range with the new update?

No. It began doing a top-off after the reboot (MS was plugged into my HPWC but didn't charge last night because I had not driven it the day before), and it came back to 251 RR which is within the 251-252 I've been seeing since the previous improvement we saw.

 

[BertL]

Received notice a couple hours ago and just applied 7.1 2.14.66 to my S90D. It took less than 30mins, although the estimate was the standard 1hr 20min. Release notes do not appear any different than before. There are no specific new things called-out as improvements.

Unlike some owners, I'm electing to continue applying all firmware releases provided by Tesla, especially after having seen some improvement in my several-months-old 90KwH Rated Range issue in the last couple of drops, and after reading yesterday that Tesla plans to introduce proactive changes to prevent charging if a potential short is determined like was the case in the isolated Norway (Supercharger) fire. I personally am not technically involved enough to know better than Tesla on such things, and Tesla does not provide detail on everything included in each update for me to really know what's there anyway, so I'll just take the possible good with any bad that comes my way. (I can always vent here on TMC, and make myself feel a little better, right?)

I will add that the Automatic Connection to WiFi problem that was introduced with 2.12.126 (see back around Mar 1, post #1691 and following in this thread) appears to have been resolved or at least improved after applying 2.14.66. Upon automatic reboot after this firmware release was installed, my MS is connected to WiFi without manual intervention once again.

It really does appear Tesla is (finally) ticking off some of the outstanding firmware bugs. YEAH!!! (But like many of us I suspect, I want more!!!) TU Tesla!

 

[garygid]

Under 2.12.126 installed at the SvC, and with the LTE/4G upgrade done there about a week ago, my auto connection to my WiFi has continued to work with no apparent problems.

 

[MarcG]

Just got 2.14.66 as well (tracker updated). Also took just about half an hour to complete. No changes that I can see so far.

 

[JPP]

Just got 2.14.66 as well (tracker updated). Also took just about half an hour to complete. No changes that I can see so far.

Rumor has it that the 14.66 build has a fix for the one-off fire at the Norwegian supercharger that apparently occurred due to a failure in a HV module under the back seat.