Looks like I will soon have yet another distraction - Arduino UNO Q

KurtE

Senior Member+
1759851102533.png
1759851139848.png


Should be interesting, with two different processors and looks like there will be another development platform for it...
 
The two 60-way connectors looks interesting. Maybe something like that for the Teensy 5 ?
 
The end of Arduino
Rest in peace , you were a great system that allowed many none SW people to use computers for their projects , hobbies , art work etc all
 
I'm not a fan and I think the buyout of arduino is not good news.
Personally, I have no idea if this will be a good thing or a bad thing. Time will tell. I have seen several different changes during my
time in this hobby, like:

LynxMotion purchased by RobotShop (I had more fun with LynxMotion)
PJRC and Sparkfun - will be interesting to see where that goes.
now Arduino-> Qualcomm could be good, could be bad.

Could be good as maybe it will allow them to have more resources, to better flesh out some of the warts. Could be bad if ...

That's my sentiment too. Don't need Linux or any other OS. Bare metal is so nice. Teensy 1GHz + SDRAM is all what I could dream of.
If it is not for you, it is not for you... I also really like playing with MicroControllers and I still spend a lot of time doing stuff with the Teensy.
But for example when I was doing more stuff with Robotics, like Lynxmotion or Trossen Robotics Hexapods or Bipeds, I did a lot more
of that stuff using RPIs or UP boards, BeagleBone Black, etc. Most of the time when I was playing with this, I had one or more Teensy boards
doing some of the peripheral work, like controlling the servos...

And this board has both. How well it will work, time will tell. Will I like it? I don't know but I do have one pre-ordered. Again another
distraction.

When it comes to the Microcontroller, if you wish to see and potentially change a lot of code, it is all open source and will be, using Zephyr.
@mjs513 and I posted earlier in this thread links to the Pull Requests to add first Zephyr support for the board and then the
one to add the Arduino wrapper around the Zephyr.

The good thing and the bad thing about Zephyr is there are a lot of people who do a lot of work to keep all of the code going into it
being consistent and builds and follows a lot of standards, code reviewed almost all of it is straight C... Which is both good and
also can be a royal PIA :LOL:

Things like: source lines can not exceed 100 characters, no blanks/tabs at ends of lines, no spaces followed by tabs,
... Commit messages lines must not exceed N characters, ... Don't have more than one blank line between functions.

(Begin Ignore section)
For example: I thought it would be good to add support for the LED controller that is on the Arduino Giga Display shield.
It is similar to the controller (IS31FL3197) they used in a different Arduino board, so I started with that one, made a copy and converted to the
capabilities of the IS31FL3197, I had it working, Did one git commit and created a PR for it. Well they decided that it should be merged
with the IS31FL3194 driver and the driver should be renamed to IS31FL319x, And the PR needs to broken up into multiple commits, where
each commit only does one specific thing, and everything needs to be able to build and pass coding standards and the like... So the one
commit now turned into:
a) Rename the IS31FL3194 driver to IS31FL319x - This includes the source, the document links, Sample code, the config links, etc.
b) During our discussion it was found current driver codes model for how to deal with color mapping was not correct - So I needed to update
that part next.
c) Finally I could merge in the IS31FL3197 specific code into the driver, sample, document.
d) Finally part 2, could add it to the board, in particular the arduino_giga_display_shield into the arduino_giga_r1//m7 config under this shield.

The rational for breaking it into steps like this, is suppose some wishes to add a another similar chip like: IS31FL3196
They can look through this PR see what I did for the 97 and have a good idea what they need to do, and likewise if
they have another board that supports one of these display chips they can look at the last commit to see maybe what they
need to do there.

Now waiting for the next round of code reviews requiring at least two people to signoff before it is merged. Hopefully soon. All just to
drive the RGB LED that is on the front of the display shield.

(End Ignore :LOL: )

But it is also good that it hopefully keeps the quality of code and documentation reasonably good. For example if you want to
know about the current released led driver, you can do a google for it and it will take you to: https://docs.zephyrproject.org/latest/samples/drivers/led/is31fl3194/README.html

Will also be fun to see how well this new board works with displays and Cameras.
Again just another diversion.
 
Personally, I have no idea if this will be a good thing or a bad thing. Time will tell

Massive corporation buying up small company never ends well. Large companies are always profit driven.

Theyre already releasing a completely new different ide. Soon you will be having to create a qualcom account. Any new boards will have to go through the new ide. 'Qualcominos' will probably be using all propirety chips which will probably not be sourceable for the average person.

Arduino was built on simplicity. Simple to get into. Barebones. Yet powerful.

I think its a kick in the guts to the thousands of people that have spent years and years contributing to arduino. eg, Paul has contributed so much to the arduino ecosystem, well before Teensyduino.

I feel like Teensy is now the last true hold out on easy to use barebones modern microcontrollers. No RTOS's or background operating systems. I hope it stays that way.
 
I'm trying to keep an open mind about Arduino's future, but it's not so easy with a company like Qualcomm.

PJRC and Sparkfun - will be interesting to see where that goes.

Until just a couple weeks ago, I was feeling... well, going to hold my tongue.

Good news is SparkFun made huge improvements starting Sept 22. Those changes were in the works earlier but delays with materials pushed everything out. Robin and I traveled to Denver on Sept 30 to meet with them in person. We also visited back in June. I really shouldn't discuss their internal details, but I can tell you we were both really relieved to see they've made so many positive changes. You can also see the effect on their website, with the main Teensy page showing 8 of the 9 models in stock. Hopefully soon they'll keep all models continuously in stock.

I'm still working on a big project that's meant to eventually transition the bootloader chips to be programmed, sold, shipped by SparkFun too. The way we currently program and package them just won't work for them. Again, I shouldn't go into details.

I will admit a huge backlog of software and documentation has built up. I had hoped to be working on it every day shortly after SparkFun took over manufacturing. Still want to get there, but we really need to transition the chip programming. This thread about Arduino & Qualcomm probably isn't the place to talk so much about Teensy's future. And to be honest, until we've got the chip programming moved the Sparkfun, it probably won't be the time (at least for me) to talk about it either.

Hopefully things will go well for Arduino and the whole maker community and Teensy can continue to be a part of that bright future.
 
Last edited:
This board is interesting at first, and many long-time makers will jump on it. They will look at it, do some cool things with it, and then put it away, only taking it out to “play” with it every now and then.
The thing is too expensive, too big, and too complex to be permanently integrated into a maker “play” project. It is unsuitable for real applications that you might want to monetize on a small scale. There are many other solutions for small Linux applications. Today, ESP+a few others are used for smaller applications.
It will be fine for a few years, then Qualcomm will resell Arduino. I don't see how Qualcomm intends to earn enough money with it in the long term to make it worthwhile.
Unfortunately, my gut feeling has been right too often.
However, PJRC also has to consider how long it can continue to rely on the Teensy 4. Any add-ons can only delay the inevitable. I believe that there will be no new processor board. And what will happen if NXP decides to stop manufacturing the chip? You know.
 
Last edited:
I don't think this new board is fit to be called an Arduino unless we can see every little piece of code that it runs, which includes everything that runs on the Dragonwing MPU - no magic binary boot or gpu blobs, etc. I can't see that happening, instead it will be a proprietary linux board with a few common peripherals dangling off the side which is just like what is already available from ODROID or the various raspberry pis.
 
I don't think this new board is fit to be called an Arduino unless we can see every little piece of code that it runs, which includes everything that runs on the Dragonwing MPU - no magic binary boot or gpu blobs, etc. I can't see that happening, instead it will be a proprietary linux board with a few common peripherals dangling off the side which is just like what is already available from ODROID or the various raspberry pis.
I can not say for sure, but a lot of this stuff is being discussed over on the Arduino Site including:

Also several other threads over on that forum about Qualcomm.
 
Exactly, it doesn't take much to see that the MPU has no open code now and doesn't look like it will in the future.
The linked datasheet is just electrical specifications, not a programming guide with register descriptions and such.
Qualcomm's site at first glance has a lot of documentation, but it's all sealed off unless you're a business partner. That's not how an Arduino should be.
 
Exactly, it doesn't take much to see that the MPU has no open code now and doesn't look like it will in the future.
The linked datasheet is just electrical specifications, not a programming guide with register descriptions and such.
Qualcomm's site at first glance has a lot of documentation, but it's all sealed off unless you're a business partner. That's not how an Arduino should be.
I agree - with the added caveat - this is pretty early in the process, so keeping fingers crossed they are still working all of this out.
 
What Qualcomm is doing is what Arduino has been doing for a long time anyway. They seem to feel like the BeagleBone and RaspberryPi sort of boards have eaten into their niche so they went there too. I have a VICOR Arduino with an FPGA on it. Over time they seem to have started to come out with hybrid boards so this is nothing new. Still, I share the trepidation about Qualcomm. They are absolutely not open source friendly. If you have a MOQ of 10000+ and you work at a fortune 500 company they're more than willing to get you under NDA and give you all the documentation you want. But, I'm not so sure they want to do that for open source developers. Of course, NDA and open source don't go together but it seems like this same trouble happened to PJRC with the current STM32 chips. Didn't they have details of the chips that are under NDA?

At worst, Qualcomm can't really stop any of us from continuing on with the existing boards and keeping them and the software infrastructure around more or less indefinitely. But, they could lock out anything newer they make (and I suspect they will do so)

Qualcomm is interesting, they make great chips but they are firmly in the closed source big business mindset and that's worrisome. But, sometimes companies change. One could argue that Microsoft went from "open source is cancer!" to moving a lot of their products to open source or partially open source. So, sometimes things change. We can dare to dream.
 
what Arduino has been doing for a long time anyway

Yes, Arduino has wanted to marry a Linux SOC with microcontroller since at least 2013 when they announced Arduino Tre (which never came to market).

1759964331224.png


Of course, NDA and open source don't go together but it seems like this same trouble happened to PJRC with the current STM32 chips. Didn't they have details of the chips that are under NDA?

Also yes to some degree, except we're using NXP iMXRT rather than STMicroelectronics STM32. But I wouldn't say the situation is at all similar to most of the Qualcomm, Broadcomm, Marvell chips where you can't get pretty much any register level programming info without signing a NDA.

NXP does publish a reference manual with nearly all the hardware registers fully documented. No NDA is required. On their site you do have to log in to download it, which is annoying, but anyone can do so. Since there's no NDA, we have that annotated copy on the PJRC site which you can get by just clicking the link.

However, NXP does require a NDA for details about some of the security features. It's small part of the chip's total features. Much of the security info is not secret, as you can see at the Teensy code security page. But sadly, since inner workings of the security features are secret. I can't talk about these parts, since PJRC did sign the NDA. I can tell you it's a very small piece and if you do sign the NDA to get the special security manual, the info is pretty underwhelming.
 
With the NXP security stuff there's ways to at least figure out how it works without having to go down the NDA route yourself, e.g. there are SDK samples and application notes floating around that show where the registers are and how to poke them to make things happen.
For something like Qualcomm's Adreno GPU, you get given a binary firmware blob and get told "this has to be loaded to make it go." What's in the blob or what on earth it does is never explained anywhere. Is it safe/secure? Does it have a gaping flaw that would let someone hack your device by connecting a maliciously crafted display? No idea because it can't be reviewed.
 
Back
Top