CubeIDE forcing you to login to ST [...] - Follow up

[CBerg]

The original Thread "CubeIDE forcing you to login to ST ...", opened by @JDoe.2 in August 2023 with 37 Kudos for JDoe

got shut down (flagged as read-only) today with a last statement by @Amel NASRI which is marked as "Solution".

Congratulations ST, now I am really upset. Now I know you read the community postings, you just prefer to ignore them if they are annoying. Flagging a thread as read only and posting an "alibi" technical solution where the thread clearly shows that thread was not about technical issues in common, but your customers do not want to be forced to log in to use your product is an affront against all those people who spoke out against those new policies.

Please take this posting as my formal protest against your action.

Edit: shortened Title to reflect the note from Amel

[Pavel A.]

What can be better than a good rant? )) Only a good fact checking.

Sorry I haven't time for thorough test right now, so did just a quick one. Like some other people on this forum I was scared by the rumors and stayed with older CubeIDE version sufficient to my work.

So here's quick informal results:

* Downloading v.1.15.1 for Windows (with logon, of course) - no issues.

* Unzip the downloaded file and install on Win11, side by side with v.1.12: no issues.

No logon requirement during install.

On the last page of the setup wizard there's a notice that telemetry can be disabled in the options (don't remember exact wording). Immediately I went there and unchecked "Help ST improve its products". No issues.

Also, in the Firmware updater option page I've checked Connection mode: OFFLINE MODE.

 * Started the new version. It opened as usual, no logon requirements again.

* Imported some existing CubeIDE projects and existing repository with installed Cube Packages for my projects: no issues, and no login prompts.

* Opened the .ioc in "Device configuration" view - again no issues and no login requirements.

* Build the imported projects - ditto, it all works (except of warnings due to the new toolchain).

Conclusion: You can install CubeIDE 1.15 in isolated environment and use it with pre-loaded files, without a single logon prompt.

[CBerg]

All joking aside, I 100% think the same. The newest announcements from M$ are an orwellian 1984 nightmare and this is definitively no exageration. I am also sick and tired of all that "corporate B$" (just for the censors, B$ stands for Belgian Dollars 😁 ) nagging, nudging and forcing me into total survilange and obedience. I will not tolerate this and I will e.g. also kick out all Microsoft Products from my company and replace them by Linux.

I came to ST because it was free. This was the main attractor.

[LCE]

> Conclusion: You can install CubeIDE 1.15 in isolated environment and use it with pre-loaded files, without a single logon prompt.

Maybe @Pavel A. 's post above on "how to install with isolation" should be made a sticky in this forum.

For now I'm staying with 1.10 anyway - never change a running system or so...

[JDoe.2]

> Conclusion: You can install CubeIDE 1.15 in isolated environment and use it with pre-loaded files, without a single logon prompt.

You're exactly right. But that's only true as of now. Who's to say the clowns in charge won't decide that v1.16 or v1.17 won't require an always-on internet connection?

Several years ago, OnShape was the first free parametric CAD based in the cloud. Only requirement is your designs were hosted on their cloud and could only be used with OnShape. It was free so I started using it, and it was legit (made by the people from SolidWorks). Then when they thought the users were hooked, it became SaaS with a hefty subscription fee. Fusion 360 pulled the same move. I have exactly zero trust that it will always be possible to install and use CubeIDE offline.

Granted, you can (and everyone does) save copies of the installers for the versions they use to make their products. We'd be insane not to. But here's the thing: each new version of CubeIDE brings support for new STM32 devices. Tomorrow, say ST releases the STM32X9 family of Edge-AI crypto-mining optical microcontrollers... well you won't find them in your old CubeIDE v1.15.

> On the last page of the setup wizard there's a notice that telemetry can be disabled in the options (don't remember exact wording). Immediately I went there and unchecked "Help ST improve its products". No issues.

And I maintain it should be opt-in, not opt-out. They hid that setting way down a dialog box into which nobody routinely goes. That's not innocent. If you don't do it immediately, chances are you'll forget this is even active. That is what I call hostile software design.

Now I may be insane, but if that telemetry was actually beneficial to us, ST could make it opt-in only, during setup, with a clear explanation of what we have to gain by providing that telemetry. The fact they try to do it behind our backs is, as the kids say, very sus.

I'll go even further: not only should telemetry be opt-in, but it should be reviewable by the user before sending. Two reasons:

A/ I don't trust anyone to measure only what they say they measure. I want to know exactly what I share.

B/ If that telemetry is actually valuable to the user, then the user should know what it says and be able to make their own conclusions.

[CBerg]

>>hostile software design<<

that's a very good term and nails it down. This is what it is: a hostile action against the (small) users.

Yes - sure - the "fact check" by Pavel is valid ATM at the first look. "oh why are you complaining so much, it's just a little inconvenience you can easily work around". In the best tradition of the "Fact checkers" - which turned out to be the worst and most maleficent propaganda spreaders we have in "corporate media" these days - @Pavel A. picks out the one point that he could prove and ignores the whole rest and the most important context. If that was not your intention, I apologize if i was to hasrh with you.

Yes, this point viewed separately and isolated may just be a "little inconvenience". But I have seen a lot and know, that this might be the first step of a series of "little inconveniences" that lead to a toxic hostile environment. And it will be the first step if the community does not speak out with all the necessary determination.

Every big corporation is on this path and one would be absolutely lame brained to assume the mercenaries in the middle and top management were not instructed by the stake-/stock-holders to follow the same path to make more money. It’s up tp us – the users  - to speak out and make clear: your job as a corporation / supplier / solution provider is to provide good products and services at a reasonable price, so users can decide to use it or not. And it is NOT your job to gain so much power over the users, that the supplier can force the user into certain behavior.

If you want to know where this ends, just take a look at M$ and their "CoPilot+" which takes a screenshot of every action you make on your computer to train an AI. Yes of course, they claim this is a "controversial" feature but they assure you not to use this against you. Haha, whow, really nice joke, tell me more of these! They just forgot to mention the footnote, which states that they wouldn't use an AI trained by you, being able to reflect all your skills and knowledge, against you ... until someone asks for exactly that and offers enough money. You put the slave collar on, by your own ignorance, because they tell you the collar will be smooth and will - of course - never used to force you to do something that you would not do, unless you are forced to.

This is, what this thread is about: freedom of software, freedom of work, freedom of decisions and foremost not being controlled by Corporations. And any step that takes away previously granted / available freedom is highly suspicious, because ATM we are living in times where some “elites” want you to believe that you will own nothing and you will be happy. They forgot to mention that you will be happy because you get some drugs in your pod and have some electrodes in your brain, that trick your brain into believing the metaverse would be real. Not with me. I will resist.

[unsigned_char_array]

And just after MS announced their new spyware "feature" Facebook just launched AI for European users that will train on all their posts unless they opt out. Such features should be opt-in, not opt-out.

My bank just announced it switched from their own wireless payment implementation to Google pay, it will now share all my wireless payment transactions I make with my smartphone with Google, because that's a requirement of Google pay. So I will just have to stick to card payments and cash(while that's still allowed) from now on. Let's hope banks will still allow cards in the future and not require smartphones for all payments(they already restrict the number of cards I can have per account to 1, and cannot have an extra bank card for my wife due to her having a different nationality, so it's not a far stretch they will eventually stop giving out cards).

And now ST with the requirement for logging in for downloads and pack updates without a means to opt out or use alternative mirrors (we archive all downloaded tools and packs for this reason).

I feel like there is a huge disconnect between marketing people, CEOs and actual end users. They do not realize how fragile companies can be if they drive off customers.

I used to love Atmel AVR microcontrollers because at the time they were the only microcontrollers with free end open source c compilers that actually worked and had no crazy restrictions (functionality or binary size limits). The affordable debuggers and development boards were a nice bonus. It took a while before other manufacturers followed and started using free and open source compilers and made their tools free and their development boards affordable. Then I moved to NXP ARM microcontroller because they had affordable high performance microcontrollers at the time. But the quality of their software libraries was so bad we were constantly fixing bugs in their drivers. So I switched to ESP32, NRF and STM32.
Once I am invested in a platform I won't switch on a whim. But if things get really bad I will and then there is no looking back. Risc-V is starting to look attractive.

My embedded software teacher often said "all microcontrollers are the same". And he was right. They all can run machine code and all have memory mapped peripherals and GPIO pins. They slightly differ in features, performance, architecture, development boards and tools. So once a microcontroller manufacturer, such as ST, starts screwing up in some of those aspects I will start looking around a bit. I won't switch right away since I have invested a lot of time in learning STM32. But once I make the switch I won't go back.

[Pavel A.]

@unsigned_char_array 

> And now ST with the requirement for logging in for downloads and pack updates without a means to opt out or use alternative mirrors (we archive all downloaded tools and packs for this reason).

Could you explain please how  ST can limit your ability to make backups of the software in any way that you prefer?

@CBerg 

>  @Pavel A. picks out the one point that he could prove and ignores the whole rest and the most important context.

Correct, I've picked only one scenario that is vital (and mostly sufficient)  for my workflow.  This was exactly my intent. Others can consider ability to create new projects and update online as most important for them.

Now, is whistleblowing an important part of our culture or necessary evil... this is way off-topic here IMHO.

[unsigned_char_array]

"Could you explain please how ST can limit your ability to make backups of the software in any way that you prefer?"

I did not claim that at all. Backups work fine like I wrote in the same sentence you quoted of me. My main problem is that you need to log in to download packs. So after you've installed all STM32CubeMX you need to download packs and those require login. It's an unnecessary hurdle. And as we saw recently it actually led to problems due to ST having problems with their servers or users with strict firewalls. There are also privacy concerns since ST knows which users use which packs, what their IP address is and where they work. You gave an example of already installed packs in your earlier post, but that scenario is not a problem. If ST provided users to use download mirrors to download packs (such as a git repo) that would make everything so much easier.
I do not plan to leave STM32 microcontrollers any time soon, but I think it's heading in the wrong direction. It seems a bit of an overreaction, but I simply don't want it to get further then this. More and more products and services have mandatory logins and accounts, require phone number verification or require you to install apps. This is a problematic trend as these things make us more depended on our smartphones, on big tech and also leads to potential privacy concerns.

[JDoe.2]

@CBerg I feel you, dude. That "you will own nothing and you will be happy" schema is creeping me out. The amount of control over everyone and everything that corporations crave is off the charts. ST has been around for a very long time, it managed to survive without requiring us to log in to download their free tools. We are given zero reason why it suddenly needs us to create accounts.

@Pavel A. are you being obtuse on purpose ? I see they call you an "evangelist". That's like a religious fanatic, isn't it ? The reason we don't want logins isn't just about the login (which is actually handicapping us in many use-cases), it's because it's one step towards a truly evil endgame which I call the "streaming platform model". I'm sure you've heard of people who (thought they) bought movies on streaming services only to find out they can no longer watch those movies because the streaming platform sold the rights to some other company. Now imagine a world, not too distant, where ST moves all its tools "in the cloud". You need an internet connection to compile everything, and your source code gets sent to ST every time. You can never backup the compiler because it's never on your premises. And then ST goes out of business and all those tools die with them, leaving you with nothing but your *** in your hand.

NXP tried it. MBED was cloud-only in the beginning, before lots of people demanded an on-premise option. Those *** practices are already happening in a lot of industries. And when a bad idea takes root anywhere, you can be certain that brainless executives everywhere will be lured by the buzzwords and do the same thing whether or not it makes any sense.

@unsigned_char_array you made me remember the good old days of AVR... I used to be an AT90S8515 expert. Knew that chip inside and out. I used assembler instead of C, though. Good times. But you're very right: it's all about convenience. AVR were the cheapest and easiest option back in the late 90's. Early in the 2000's Cypress sent me a PSoC kit that is, for all intents and purposes, a Nucleo-32 module : it was so easy to get into, it motivated me to switch to Cypress. When NXP released the first MBED module, I switched to NXP. Wasn't happy about the online compiler of course, but I knew this was only to get started and eventually used IAR for work. Switched to STM32 because of the Nucleo modules. There's a pattern. If tomorrow a competitor releases a nice cheap module for their MCU and tools are free... well, ST, perhaps it'll be time to say goodbye. Unless you give me (us, really) additional reasons to stay.

Anyway, ST, I don't know who runs your forums, but this isn't a Britney Spears fan club or a social network. I've just been informed that I've received yet another "badge" for all the "kudos" I've been given recently. WTF do you want me to do with that ? Brag about it on LinkedIn ? Who do you think I am, an engineer or a tiktok influencer ? Take all your badges; if you value me that much why don't you send me free development boards instead ? Stuff I can actually work with and that might even generate some sales for you. Let me assure you, if I were to tell my boss "wow ST gave me a badge", you know he'd just say "IDGAF", so why do it ?

And if dev kits are too expensive, may I suggest saving money by firing the people who "gamified" this forum ? Also, the "community managers". I'm your customer, I don't want you to manage me. A forum should only need moderators, I have no idea what a "community manager" even does.

And another rant done... then again, I don't have much better to do while I keep an eye on that Jetson Nano recompiling OpenCV from source. Getting CUDA acceleration to work is like pulling teeth...

[Techn]

Hi, 

I  had an issue that when logout situation one older work space was not visible. If I  change to my recent work space I can login. Then if I change the work space back to the older one the projects started visible.  

Can you please explain the reason why it happens like this