Possibly of interest, possibly for correction:
http://www.elektor.com/news/microchip-launches-first-32-bit-arduino.1835545.lynkx
// Christian
"Microchip launches first 32-bit Arduino compatible development platform"
(16 posts) (6 voices)-
gbulmer
ModeratorI went to the 'feedback' page and wrote a complaint.
I realise things like Elektor get loads of press releases to print, but IMHO, it is abysmal that they did not check the validity of the headline. That is dreadful. I hope other people go to Elektor and complain.
I have no quarrel with the product, but I do dislike the misleading marketing.
Here is a link to the Digilent page:
http://www.digilentinc.com/Products/Catalog.cfm?NavPath=2,892&Cat=18I thought it might have very close hardware compatibility, but it is clear that it is only compatible with 3.3V shields.
I looked at the forum, and found this link to a compatibility analysis at:
http://ruggedcircuits.com/blog/2011/05/22/microchip-chipkit-and-arduino-pin-compatibility-analysis/
which identifies some of the pin-function incompatibilities.Interestingly, they have used jumpers to allow you to select which pin is connected to the header. I'm not at all sure about that. It seems more complex to me. Loses the spirit of Arduino simplicity. But, that's their choice.
The software is at https://github.com/chipKIT32/chipKIT32-MAX
It appears to be a fork of Arduino IDE 0022.Maybe LeafLabs should follow up?
Posted 4 years ago # -
poslathian
LeafLabsLeafLabs has been in touch with some of the publishers, who realize the error. We are preparing a response to print up in EETimes. If you feel so inclined to write to the publishers, we appreciate your support!
Mbolivar is spearheading our response. To be honest, we are focused less on "who was first," that is obvious ;), but instead what the right approach to 32 bit adoption should be. Chipkit is focused on "100% arduino compatibility" to the extreme extent they can achieve this goal. This is not our goal.
A developer from chipkit has been in touch with us, and there is not any drama there. He is a friendly and helpful guy. I expect that Maple support in their new super arduino IDE is a distinct possibility.
Posted 4 years ago # -
mbolivar
LeafLabsThanks to everyone for their support. And thanks to gbulmer for the useful links; they will be of great help as I prepare our response.
As poslathian said, we don't have any problems with the chipKIT product or its developers. They've done a ton of great work, and we don't want to detract from that. However, we dispute that their goals are the right way to bring the convenience that Wiring/Arduino brought to the AVR into a 32 bit context.
Here's a quick preview of our main points:
1. We believe that "100%" compatibility is pure marketing and not really achievable. There are always going to be differences when you change architectures, and it will always be possible to write code which works on Arduino boards but not on 32-bit architectures.
2. Achieving the maximum degree of compatibility which is possible would constitute a deliberate crippling of the capabilities of the STM32, which in our minds does a grave disservice to you, our users. Who wants a 10-bit ADC reading when the hardware is capable of 12 bits? Similarly, who wants crippled PWM resolution? Etc.
3. Achieving maximum compatibility would be misleading. The most glaring example of this to me is requiring that "word" be a typedef for a 16-bit quantity, which is an outright lie on the 32-bit Cortex M3.
Basically, we think that you are smart and talented people who are fully capable of understanding such differences. There's no need to detract from the power of the STM32 to make it easy to use. All that's really needed is solid documentation and supporting tutorial material, which (at least IMO) is the real reason for Arduino's success, not the number of bits of resolution in their PWM.
Again, we thank you all for your support, and as poslathian said, we'd really appreciate it if you wrote to any publishers you notice repeating the claims made in chipKIT's press releases.
Posted 4 years ago # -
gbulmer
ModeratorI counted the number of 32bit non-ATmega Arduino boards on the Wikipedia 'Arduino' page.
5 are ARM, 1 Parallax Propeller, and one PIC32.I wrote to Elektor. It seemed extremely sloppy that they didn't even type 'Arduino' into google to check the facts.
I tried writing to dangerousprotoypes, but, for some reason (maybe the URL to ruggedcircuits) it got silently rejected.I've written to ARM too.
I have no quarrel with the chipkit board.
I wouldn't have done it that way, but they have made their design decisions, and it looks like a worthy piece of work.I am annoyed at what I read as false advertising, and the laziness of the publishers.
Posted 4 years ago # -
poslathian
LeafLabsWe appreciate the support guys! We have been in touch with a lot of the publishers, and with microchip. The publishers are surprisingly supportive, generally they just reprint the press releases lock stock and barrel and trusted microchip to do the right thing. They were generally pretty irked to learn of the error.
One publisher in particular even cited their "no firsts" policy to avoid this issue. A smart policy, but unfortunately it was not followed here ;)
Posted 4 years ago # -
gbulmer
ModeratorI was able to post at my comment about 'Not First' at Dangerous Prototypes, maybe because I didn't include any URLs.
While there, I and stumbled across this article:
http://dangerousprototypes.com/2011/05/31/chipkit-pic32-compiler-qa/I looked at PIC32 in 2009 as a possible MCU to base an improved Arduino on, but I abandoned any plans on the basis that some of the source was not freely available from Microchip.
When I read the "First ..." press release for the chipkit, I assumed, I think very reasonably, that it is all Open Source.
According to the person "Jason Kajita", who is described at that link as "a compiler developer" it isn't.
It seems that some of the code is not Open Source, specifically "At the moment, the source for the Standard C library is not provided. This is because it contains some code licensed from other companies."That makes that marketing even more misleading. Sorry to whine, but that marketing is a travesty of "truth in advertising".
Posted 4 years ago # -
gbulmer
ModeratorI have updated the Arduino section on Wikipedia:
http://en.wikipedia.org/wiki/ArduinoIt took a while because I wanted to check the level of compatibility of the other non-ATmega boards to ensure it is fair.
I also thought it was helpful to try to explain the types of compatibility, and hence why there are two lists of boards.
While checking, I noticed that one of the FEZ boards was missing, so I added that too.
I hope people can find some time to check my words for accuracy and fairness.
Posted 4 years ago # -
gbulmer
Moderatormbolivar
1. We believe that "100%" compatibility is pure marketing and not really achievable. There are always going to be differences when you change architectures, ...
Agreed. This is true for Arduino UNO vs Arduino Mega 256 where the I2C appears on different pins.
... and it will always be possible to write code which works on Arduino boards but not on 32-bit architectures.
I'm not sure what you mean. Is this about the differences in peripherals on an ATmega vs other microcontrollers, which I accept, or something else?
2. Achieving the maximum degree of compatibility which is possible would constitute a deliberate crippling of the capabilities of the STM32, which in our minds does a grave disservice to you, our users. Who wants a 10-bit ADC reading when the hardware is capable of 12 bits? Similarly, who wants crippled PWM resolution? Etc.
I think there are variations in requirements, and hence variations in solutions.
I can agree with you, because I want the extra facilities in an STM32F103. I think it is inappropriate to over constrain what is deemed compatible. If that were done, there might be no Arduino Mega, LillyPad, Nano, Pro Mini, Mini or Fio.
I also can accept that some users may only want an Arduino compatible with more memory, that runs faster, but otherwise behaves identically. Maybe someone will rework the FPGA-based Papilio to make a 'perfect', high-performance Arduino clone.
I believe that part of the attraction and success of the Arduino boards is the ease of use. IMHO the chipKIT pin-selection headers compromise that. But, that is Digilent/Microchip's choice.
I am concerned to ensure people have accurate and balanced information which helps them choose the solution that suites their needs. Hence I am very concerned that Microchip has used its significant financial advantages over LeafLabs to propagate a marketing message which I believe is factually inaccurate. I hope this sad and worrying start to the Microchip marketing approach does not continue. We can hope that Microchip understand that part of the ethos of the Open Source community is fair and accurate information.
Posted 4 years ago # -
poslathian
LeafLabsTo be fair, I imagine there was no mal-intent on the part of Microchip. They are a big company, and I imagine their PR campaign is NOT run by engineers. In fact, the PR campaign for chipkit IS being run by Michelle (dot) Ragsdale (at) microchip who I am guessing is simply misinformed by those working downstream.
More surprising than Microchip is Digilent, whose president - Clint Cole - was quoted as saying: "The chipKIT platform is the first and only 32-bit solution of its kind in the industry...Academics and hobbyists can plug the boards into their designs, download the open-source software and have their projects up and running in minutes." - http://www.soccentral.com/results.asp?CatID=191&EntryID=33909
Posted 4 years ago # -
My guess is that Digilent doesn't really research into other competitor's products when given a task such as implementing a 32-bit PIC into an Arduino Mega and Arduino Duemilanove form factor.
As for Microchip, they probably know of other 32-bit Arduino-compatible boards, however marketing does like to exaggerate.
Posted 4 years ago # -
poslathian
LeafLabsExactly. The real solution to this problem is to put ourselves on others' radar by making a great product, doing right by our users, and getting user projects as much internet facetime as we can. It may be annoying, but not at all surprising that big companies wouldn't go looking for any toes they might be stepping on when launching an expensive marketing campaign.
Posted 4 years ago #
Reply »
You must log in to post.