More phantom breaking in the last month or so ...

[marsarbu]

Since the last update before the Android Auto one, I am having a lot of phantom breaking. Not the throw-you-through-the-windshield type, but it taps the breaks a lot. It happens maybe 30% of the time when I am overtaking a truck that is on the right. It also happens - not as much - with non-trucks, and it seems it's always when the other car is in the lane to my right. I don't have any error for LIDAR or cams. The other vehicle doesn't venture in my lane, or doesn't even ride the divider. It's happening regardless if the road winds or it's straight.

Is it just my luck or is anyone else having the same issue lately?

 

[lucidtouringman]

I just drove about 400 mi yesterday and experienced this at least half a dozen times on the road. It was extremely irritating at best and dangerous at worst but now definitely happening with me and I can confirm this is not just you

 

[Genghis]

Was it while you’re driving normal, with DA on , or with ACC on? I haven’t noticed it but my GF was complaining about why I was braking so much when I know I wasn’t on the freeway driving normal and with DA on.

 

[lucidtouringman]

DA on, leftmost lane, mostly happened while passing a truck to my right. Very disconcerting. It wasn't a hard brake, more like I tapped just hard enough to disengage DA, only thing being my foot was nowhere near the pedal.

 

[RAF42]

DA on, leftmost lane, mostly happened while passing a truck to my right. Very disconcerting. It wasn't a hard brake, more like I tapped just hard enough to disengage DA, only thing being my foot was nowhere near the pedal.

Same for me. DA only, not ACC. On a recent trip up I-5 in CA in the morning, every time I passed a truck on my right the shadow from the truck would trigger a slowdown. Not overly aggressive, but enough that I changed to ACC after it happened a number of times.

 

[lucidtouringman]

Same for me. DA only, not ACC. On a recent trip up I-5 in CA in the morning, every time I passed a truck on my right the shadow from the truck would trigger a slowdown. Not overly aggressive, but enough that I changed to ACC after it happened a number of times.

Same here.

 

[marsarbu]

So it is a problem. Classic Lucid: fix one thing, break another. I'm wondering when is this going to stop. So frustrating.

 

[JohnAStebbins]

This was also discussed in the "Update 2.6.16" thread https://lucidowners.com/threads/update-2-6-16.12164/page-4#post-271811

 

[RAF42]

FWIW, my post above was when I was on 2.6.16, so this was definitely before 2.7. Thus they did not “break” this with this update. And when it happened with my Tesla, it was more often and much more aggressive, So not unique.

 

[marsarbu]

FWIW, my post above was when I was on 2.6.16, so this was definitely before 2.7. Thus they did not “break” this with this update. And when it happened with my Tesla, it was more often and much more aggressive, So not unique.

All I know is that I did not have this problem before, now I do.

One of my teams (140 people) manage a very large, complex, 25-year old application. It has been built on, mangled, fixed, patched by multiple generations of developers. The development "style" is CI/CD which basically means that they don't have a QA group, they are fully relying on automated testing, and they are pushing code into production on a daily basis. Despite all that, it does not break existing functionality.

To all of you that keep saying "this is normal", I am telling you, it is not. If I can pull it off, Lucid should be able to. There are well-established, even boring ways, to do this right. Why Lucid can't do it right is beyond me unless they are reasonably incompetent when it comes to managing this process.

 

[L-Dude]

Comparison is the thief of joy

 

[borski]

All I know is that I did not have this problem before, now I do.

One of my teams (140 people) manage a very large, complex, 25-year old application. It has been built on, mangled, fixed, patched by multiple generations of developers. The development "style" is CI/CD which basically means that they don't have a QA group, they are fully relying on automated testing, and they are pushing code into production on a daily basis. Despite all that, it does not break existing functionality.

To all of you that keep saying "this is normal", I am telling you, it is not. If I can pull it off, Lucid should be able to. There are well-established, even boring ways, to do this right. Why Lucid can't do it right is beyond me unless they are reasonably incompetent when it comes to managing this process.

Okay, that’s definitely it. You should apply, I guess. They need you.

Also, as a reminder, this did not change with this release. This has been the case for a while.

Good luck writing automated unit tests and integration tests for ADAS software. You would need to build, at least, full simulations. Lucid has done and is continuing to do this, but it isn’t whatever “complex” webapp or desktop app you think it is.

Not all software is equally difficult or complicated to build.

(Complex software is hard, CI/CD is simply a testing process but does not replace QA, etc)

 

[marsarbu]

Okay, that’s definitely it. You should apply, I guess. They need you.

(Complex software is hard, CI/CD is simply a testing process but does not replace QA, etc)

I guess they already have you.

 

[borski]

I guess they already have you.

Nope; I’m starting a company in the clinical trial space and have never been paid a dime by Lucid.

But I’ve been writing software since I was 9 years old.

This isn’t a competition. My point was that a lot of people look at the forest and go “it can’t be that hard,” and then dive in and realize they neither have the tools nor the will to trudge through the bush, and that a lot of problems are unseen when you’re looking at them from above the canopy.

That’s all.

 

[marsarbu]

Nope; I’m starting a company in the clinical trial space and have never been paid a dime by Lucid.

But I’ve been writing software since I was 9 years old.

This isn’t a competition. My point was that a lot of people look at the forest and go “it can’t be that hard,” and then dive in and realize they neither have the tools nor the will to trudge through the bush, and that a lot of problems are unseen when you’re looking at them from above the canopy.

That’s all.

I was kidding. Good luck with the start-up.

 

[borski]

Just to be clear: Lucid’s software team is far from perfect, and I’m not saying I’d hire all of them. I can’t say that; I don’t know all of them.

It is true that they somewhat regularly make mistakes that have me placing palm upon face, but that is mostly a sign of *immaturity* in the software team and process, more than a sign of incompetence.

I guess that’s where I take issue; I don’t like it when anyone paints a group of people with a very broad stroke, especially without the context necessary to do that.

Calling the software team and process immature is accurate. Calling them incompetent isn’t.

 

[marsarbu]

Just to be clear: Lucid’s software team is far from perfect, and I’m not saying I’d hire all of them. I can’t say that; I don’t know all of them.

It is true that they somewhat regularly make mistakes that have me placing palm upon face, but that is mostly a sign of *immaturity* in the software team and process, more than a sign of incompetence.

I guess that’s where I take issue; I don’t like it when anyone paints a group of people with a very broad stroke, especially without the context necessary to do that.

Calling the software team and process immature is accurate. Calling them incompetent isn’t.

You are way more forgiving that I am. My personal standards are different. If we fail the end goal, we are not competent at it. It's as simple as that, no need to hide behind words and semantics. It doesn't mean we cannot get better, failing is normal. But REPEAT failure is driving me absolutely mad. It shows you don't have the capacity to learn from your mistakes or you don't care. There's always a way if you want to change things.

And I didn't call the developers specifically incompetents. They need to be enabled to do their job right. I don't know who cannot fix this repeat-fail thing, but I would get rid of them.

 

[borski]

You are way more forgiving that I am. My personal standards are different. If we fail the end goal, we are not competent at it. It's as simple as that, no need to hide behind words and semantics. It doesn't mean we cannot get better, failing is normal. But REPEAT failure is driving me absolutely mad. It shows you don't have the capacity to learn from your mistakes or you don't care. There's always a way if you want to change things.

And I didn't call the developers specifically incompetents. They need to be enabled to do their job right. I don't know who cannot fix this repeat-fail thing, but I would get rid of them.

What repeat failure? Again - this is not a new issue, and did not arrive with this update.

What you’re talking about is regression testing; you don’t want the same mistake made twice. I agree.

I haven’t seen Lucid make the same mistake twice.

 

[HC_79]

So it is a problem. Classic Lucid: fix one thing, break another. I'm wondering when is this going to stop. So frustrating.

Get used to it. If you’re not prepared to put up with this then stay away from the likes of Lucid, Tesla & Rivian. These cars are so software driven, bugs will always exist.

 

[marsarbu]

What repeat failure? Again - this is not a new issue, and did not arrive with this update.

What you’re talking about is regression testing; you don’t want the same mistake made twice. I agree.

I haven’t seen Lucid make the same mistake twice.

Repeat failure of releasing updates that break previous functionality.