JRP3
Hyperactive Member
As he's explained, initially he thought there might be a safety issue but since he found the actual problem he realized it's not a safety issue. The words "If" and "I believe" denote conjecture not certainty.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Wiki Sudden Loss Of Range With 2019.16.x Software
- Thread starter Dutchmeeuw
- Start date
Thats not what he said in post #13653. There are a lot of word games being played here, which I equate with lies.
So, the internet forever has outdated speculation available. Got it.
I'll briefly point out that this is not backing up your misquote of me in any way.
As explained above, and as I'll go over here again for the hundredth time, this was posted when I believed Tesla was detecting and mitigating instances of "condition X"... which, if this had been the case as I'd thought at the time, would have been quite bad.
However, and as I pointed out many times, this was NOT the case. Tesla was instead detecting a different issue, Z, and mitigating it with the corrections intended to correct X. Once this disparity was realized by Tesla, and seriousness of X, they updated the software to basically lock down the battery if X was detected (0 known cases of this), and implement mitigations if Z was detected.
Again, I _never_ said any update or version was dangerous, never said any update prevented fires, etc. These are things you're fabricating and falsely attributing to me.
Fortunately I'm not the only one who recognizes this.
islandbayy
Active Member
Yes. Agreed. I think of it like this. Same as a inactive cell phone. A inactive cell phone can still dial 911 in an emergency. Similar feelings with the car. What would have happened if I had been taking someone to the ER, only to find the car in a software update basically bricked until it's completed. I do think they have stopped that practice though. I have a loaner M3 right now that has a update queued and it's not installing itself. thankfully. Always seemed off they would do this.
Can we please get back on topic with the issue at hand and stop all this childish he said/she said.
To me, safety issue, and dangerous go hand in hand. I may have not used the exact wording when I quoted you, but the end result is the same. You are entitled to your own opinion, just not your own facts.
I’m fairly certain that condition X (as you call it), has happened. Cars don’t just burst into flames with condition Z. Can anybody prove it(including yourself). Of course not, as neither of us have the logs from the crispy cars.
Furthermore, Tesla has and will stealth update those on old software, without consent. They absolutely freak out when they see pre 2019.16.1. Maybe not the techs(most of who are clueless), but engineering will have a caniption. Don’t believe me? Start flashing your customers cars with 2018.xx.x, and see what happens. I can guarantee all hell will break loose.
Addressing condition X... can I be 100% certain this hasn't happened anywhere? No, of course not. Can I be certain enough to bet the farm, so to speak, on it having never happened? Yeah.
Like I said, when the first updates that started hunting for condition X started hitting cars, and some false positives were happening (condition Z's), I believe it was pretty clear from the code's evolution that no one expected to see a positive at all. Looking back in hindsight at this point, it seems fairly obvious that whoever was doing these revisions was aware these were false positives, also.
Fast forward to when this was all somewhat finalized in the code (separating out X from Z), the error that is thrown (and the subsequent loss of use of the vehicle/battery) after detecting X is obvious and unambiguous. In fact, I'm pretty sure if anyone had actually ever gotten this particular note it would have made headlines. (I do find it amusing that this is one of the only errors where there's a specific code path for displaying it, and it's obfuscated enough that most would never encounter it in examining the UI code.) Since I've seen zero cases of this occurring anywhere, along with everything else I know about these issues (including convos with insiders)... it's a super safe bet to say there have probably not been any cases of X whatsoever.
Last edited:
As for the "freaking out" over older software, this is because all of their internal tools are setup around the latest or near-latest firmware. There's a lot of things they're just no longer able to do when a car is on an older firmware version, especially if that version is prior to the transition from VPN to Websockets connectivity to the mothership. In most cases all they have at their disposal is the ability to get a vehicle on to a newer version manually (with physical access) before proceeding with any other service.
And for the most part, I don't think it's an unreasonable way to handle it. Tesla shouldn't be expected to provide service and support for old versions indefinitely, as this would be a huge ongoing burden in order to deal with the small handful of folks who don't want to update.
That said, I don't think that they should be able to go the Windows 10 "YOUR COMPUTER IS RESTARTING FOR UPDATES RIGHT NOW WHETHER YOU LIKE IT OR NOT" forced update style of things... but I could easily see it getting to this point if they ever take the time to make it so updates can be applied with minimal downtime. Like, if an update can be installed while a car is charging, and complete with say 2-5 minutes of vehicle downtime... I could see mandatory updates becoming a thing. Again, I believe I should be able to control update (as well as see clearly all changes/revisions made BEFORE installing an update...) but, unfortunately I think we're in the minority on this and it'll just never happen.
And for the most part, I don't think it's an unreasonable way to handle it. Tesla shouldn't be expected to provide service and support for old versions indefinitely, as this would be a huge ongoing burden in order to deal with the small handful of folks who don't want to update.
That said, I don't think that they should be able to go the Windows 10 "YOUR COMPUTER IS RESTARTING FOR UPDATES RIGHT NOW WHETHER YOU LIKE IT OR NOT" forced update style of things... but I could easily see it getting to this point if they ever take the time to make it so updates can be applied with minimal downtime. Like, if an update can be installed while a car is charging, and complete with say 2-5 minutes of vehicle downtime... I could see mandatory updates becoming a thing. Again, I believe I should be able to control update (as well as see clearly all changes/revisions made BEFORE installing an update...) but, unfortunately I think we're in the minority on this and it'll just never happen.
I actually just looked into this a little further. Turns out they actually did update this behavior. Updates on Tesla-owned vehicles only happen automatically when they're in a geofenced area now (any Tesla service center, mainly). So, kudos to them for tweaking this.
Basically if a vehicle isn't marked as delivered (ie: owned by Tesla), the update will just stage as normal and schedule in the background (along with the normal UI popups). If at the background scheduled time (1AM local I think) the car is in a geofenced area, has sufficient charge, etc... the car will kick off the update. If not, it just remains staged as normal.
Pretty cool.
Software revisions that result in material changes that negatively impact a verified significant specification would surely be of great interest to all owners.
When I look at this from a non bias perspective, I just cannot see how we could be polar opposites on our opinions. There is no doubt that you see many more cars than I do, yet you claimed to have only experienced condition Z in a couple. My experience is that quite a few of old 60/85kwh cars have condition Z. I have personally seen a 60kwh car that shut down due to a battery error(no longer would function on new software), but when rolled back, it would work. I suspect it had condition X. I wasn’t the one working on it, so no logs/proof, but did see it with my own eyes. It did have a battery failure later(no fire). I guess I just cannot believe your current posts here, as they not only are counterintuitive, but exact polar opposites of my experience.
Tesla's treatment of me as a 90 pack owner has me looking closely at Lucid's offerings. I'll certainly be careful, but Tesla's duplicity is already apparent, making them completely unacceptable when it's time to purchase a replacement vehicle.
I was specifically talking about cars that will never see a service center, so there is absolutely no worry of Tesla techs having to monkey with anything. Tesla can, and still has access to these cars(and ability to make changes over the air), as long as one does not remove their super user access.
Does anybody actually believe that Tesla is paying for connectivity on all cars out of the goodness of their hearts? It’s just not happening.
wk057,
Have you any insights into the emasculation of charge rates for the 90 kw packs that happened a few years ago? Since 2018 (or maybe late 2017) I spend a full one third of my time on road trips at superchargers. Before throttling, I spent about half that amount of time charging.
Is this in any way related to the choking off of charge rates for the 85's?
Have you any insights into the emasculation of charge rates for the 90 kw packs that happened a few years ago? Since 2018 (or maybe late 2017) I spend a full one third of my time on road trips at superchargers. Before throttling, I spent about half that amount of time charging.
Is this in any way related to the choking off of charge rates for the 85's?
There's definitely other detectable issues that will cause a vehicle to shutdown in all versions, and a lot of these detections involve testing over an extended period involving data accumulated across the lifetime of the vehicle. Rolling back BMS firmware before certain points will clear parts of EEPROM, as the structures have changed over time. While forward compatibility is assured, backwards updates have to wipe out some data to install properly, which can cause some of these algos to have to start from scratch.
The condition X error is definitely not what you experience though. The error literally comes up on both screens and, in part, tells you to exit the vehicle immediately. It's pretty unambiguous. So this definitely isn't what happened to the car that had the issue that you're referring to.
Well, sure. They do it for all sorts of reasons. But I was referencing your "freaking out" note... which would be tied to Tesla needing to service the vehicle and being unable to.
Goodness of their hearts? Of course not. This was a paid feature bought as part of the car. There was nothing that said they can drop connectivity related to refusing updates or any other method "for the life of the vehicle". And they definitely get tons of diagnostic data from the fleet that they can use to improve things. It's definitely not useless to them.
This just doesn't really tie into forcing updates or anything to do with this thread, really.
The original 90 packs, as noted elsewhere, are garbage. They degrade more than twice as quickly as the 85-type, so would hit issues related to slower charging and the like sooner in life than their 85-type counterparts.
I haven't seen anything related to anything in this thread affecting 90 packs.
AACK!I didn't read all 684 pages, so I didn't see the part about the 90 packs being garbage. At 75,000+ miles, I haven't seen a lot of degradation in mine, but the charge rate was cut by about 20-25kwh, and the taper increased draconically, very early on (at maybe 20k miles on the car, with around 2/3 of that being supercharging).
Guess I should have waited a few months longer to order a car, but when I ordered in Jan 2016 we didn't know about the degradation issue (to be fair, I don't think Tesla did either), and had no idea a 100 kw battery would be coming out a few months after I took delivery.
Tesla's handling of the issue sure leaves something to be desired, though. They never sent any sort of alert or message when they discovered the issue - just killed the charge rate and hoped no one would notice.
Similar threads
