Software Versioning Explained

[RM-S8]

Like Bobby said, 10 is higher than 4.
Again class, 2.4.10 is higher than 2.4.4.
What's so confusing?

 

[Buffalo Bob]

Here's a crazy thought... Could Lucid have simply called this 2.4.5 instead of 2.4.10? And then taken on their numbering schemes in such a manner as to minimize the number of times that this LEGITIMATELY CONFUSING (to a non-software engineer) phenomenon occurs?

 

[Bobby]

Here's a crazy thought... Could Lucid have simply called this 2.4.5 instead of 2.4.10? And then taken on their numbering schemes in such a manner as to minimize the number of times that this LEGITIMATELY CONFUSING (to a non-software engineer) phenomenon occurs?

The first number is the major release number. The second number is indicative of minor releases, usually incorporating known improvements. The third number is usually for bug releases or extremely minor tweaks of the others. This is typical in all software products.

 

[joec]

Bottom line: Release numbers are part convention, part company specific. I've seen it done lots of different ways. But usually once a company establishes a system, they keep it consistent.

It might be fun to try and figure out exactly what they are doing with these numbers over time. If I'm correct that .10 is always the "final bug release" then we should almost never see a .11 release.

 

[borski]

I don't get the seeming anger over people's confusion with Lucid's OTA release numbering scheme. Most people are not software engineers, so when they are exposed to an apparent numerical inconsistency for the first time, they are bound to be confused. I, like many of us, have been there in past instances, so I'm sympathetic when some newcomers are scratching their heads. It's a teaching/learning opportunity!

Fair. We've had this conversation on this forum at least five times that I can count, which is why it starts to grate, especially when people argue about it. You're right, though, that I should have been kinder in my responses because though *I've* had that converastion a dozen times, that doesn't mean they have. Feedback taken. Thanks.

Here's a crazy thought... Could Lucid have simply called this 2.4.5 instead of 2.4.10? And then taken on their numbering schemes in such a manner as to minimize the number of times that this LEGITIMATELY CONFUSING (to a non-software engineer) phenomenon occurs?

The 'skips' are simply unreleased versions. In short, think of it this way: any time an engineer makes a change and is 'ready to test it' either on a car or in some simulator, they need to make a release. Of course they will have tested locally, but the only way to get a proper test is to run all of the integration tests, get it on a vehicle or simulator, etc.

Sometimes, those tests fail. When they do, that version doesn't get released to the 'public' channel; people name these channels differently, but the meaning is the same. However, that version isn't deleted; it is simply 'internal'. That way, they can always track what happened when, and they don't lose data. Version number re-use is extremely rare and very looked down upon, the only exception usually being if it was truly a botched release and didn't include things it should have, etc.

So, between 2.4.4 and 2.4.10 there may have been a bunch of versions that didn't work. Or perhaps 2.4.5 needed to be there before they could apply 2.4.6. Or, perhaps they skipped a bunch so that they could start with a new "ten."

It doesn't really matter, but all of those are likely candidates for why there seem to be 'version skips.'

 

[levsha]

https://semver.org/ was mentioned already here. And it explain the purpose of every number right at the beginning:

Given a version number MAJOR.MINOR.PATCH, increment the:

1. MAJOR version when you make incompatible API changes
2. MINOR version when you add functionality in a backward compatible manner
3. PATCH version when you make backward compatible bug fixes

 

[Wolfman]

I'll be glad to tell you that. But I'm afraid it won't help. 2.4.4 isn't a decimal number to begin with. Decimal numbers don't have multiple decimal points. So I don't see why anybody would think that the part after the second decimal point represents a fraction of anything. The logic simply isn't there. It also defies sense to think that a release number could somehow be fractional. Something could be the 4th release or the fifth release, but can't be the 4.4th release. So thinking of it as a decimal number makes zero sense to begin with. What Lucid is doing is entirely consistent with what other companies do with software releases or versions.

Mine isn't, but I don't have the update yet. I look forward to it. There have been lots of threads related to the TCU, even though they didn't mention the TCU explicitly.

Well, this is how software revisioned wether it’s a Lucid or an iPhone.

The process is always the same.
Major version is the first number.
Minor version is the second number
The batch version is the third number.

Edit: somebody beat me to it

 

[AlignTrex]

I just wasted 5 mins of my life reading this thread about people debating if 4 or 10 is larger.

 

[borski]

I just wasted 5 mins of my life reading this thread about people debating if 4 or 10 is larger.

And what did you learn?

 

[Halodde]

Indeed a learning opportunity for me. I have learned so much from this forum and the incredibly intelligent people here. Although I am definitely smarter now than when I was 2 years ago when I joined, I feel a whole lot drummer because of all the brilliant people who surround us here. I appreciate all the help and information!

I don't get the seeming anger over people's confusion with Lucid's OTA release numbering scheme. Most people are not software engineers, so when they are exposed to an apparent numerical inconsistency for the first time, they are bound to be confused. I, like many of us, have been there in past instances, so I'm sympathetic when some newcomers are scratching their heads. It's a teaching/learning opportunity!

 

[ccarroz]

Bottom line: Release numbers are part convention, part company specific. I've seen it done lots of different ways. But usually once a company establishes a system, they keep it consistent.

It might be fun to try and figure out exactly what they are doing with these numbers over time. If I'm correct that .10 is always the "final bug release" then we should almost never see a .11 release.

You can have a .11 release. It would be easier easier to read and understand since its easy to drop the zero and read it as .1 vs .10. However, I think the bigger problem is when we see 11 now, we think its going to supercharge something in the car since this patch "goes to 11".

 

[Buffalo Bob]

Indeed a learning opportunity for me. I have learned so much from this forum and the incredibly intelligent people here. Although I am definitely smarter now than when I was 2 years ago when I joined, I feel a whole lot drummer because of all the brilliant people who surround us here. I appreciate all the help and information!

It must be pretty cool to have a Lucid in Buffalo. I'm guessing that you don't see many there, and that your car garners quite a bit of attention. If I ever get my car that far east, we should do a Buffalo-style version of Lucids & Lattes... Perhaps a meet-up for Beef on Weck at Charlie The Butchers! (My wife and I recently had to cancel a September trip to WNY and Ontario, so I had to dash a number of local dining plans!)

 

[Halodde]

You can have a .11 release. It would be easier easier to read and understand since its easy to drop the zero and read it as .1 vs .10. However, I think the bigger problem is when we see 11 now, we think its going to supercharge something in the car since this patch "goes to 11".
View attachment 23806

{patiently waiting for the inevitable joke about the robustness level going to 11}

 

[Haggy]

You can have a .11 release. It would be easier easier to read and understand since its easy to drop the zero and read it as .1 vs .10. However, I think the bigger problem is when we see 11 now, we think its going to supercharge something in the car since this patch "goes to 11".
View attachment 23806

The Tesla Model S and X have volumes and fan levels that go to 11. The Model 3 and Y do not. Those poor folks!

 

[levsha]

The Tesla Model S and X have volumes and fan levels that go to 11. The Model 3 and Y do not. Those poor folks!

Well they need to have something to justify the price difference

 

[Haggy]

Well they need to have something to justify the price difference

I for one was glad. The Model S and X were out first. People who bought the Model 3 didn't deserve anything that went to 11. The sound system is loud enough and the climate control works well enough, but they didn't even get numbers. It's just a bar that goes from off to maximum. So they can't tell people at parties what setting they use, which is probably also a good thing, because who would want to hear it anyway?