SparkFun has good stuff too — I’ve been using their ESP32 Thing for a while, and they have a whole ecosystem of little plug-and-play I2C peripherals that don’t even require breadboarding.
Pretty much every dev board I’ve looked at satisfies this article’s first two criteria, I.e batteries-included [sic] and easy flashing. And I’m doubtful about the other criteria; they seem aimed too much at beginners.
The description of the Arduino bootloader sounds incredibly primitive compared to the mBed boards. They appear as mass-storage devices and you just drag and drop the image. They contain a shortcut file that opens up a web-based IDE, which lets you compile for your board from a web browser and then download the final results as a file that you can drop onto the attached USB mass-storage device and then hit the reboot button. That’s far more batteries-included than the article describes (you don’t need any custom hardware or to install any software, you just need a USB port and a web browser), yet mBed boards don’t seem to be particularly popular. The libraries also look like the ones described in the article, the price-point is about the same, and there is a reasonably active community (though that’s a bootstrapping problem: Arduino presumably didn’t have one on launch, you only get a community once the board is popular and so it can’t be a prerequisite for a board becoming popular).
The whole article reads like something I’ve read a lot of times: someone who was successful by being in the right place at the right time wanting to justify their success as something other than luck.
The mbed flow is probably as integrated as something that requires no installation can get.
On the other hand, I don’t feel that it’s far more batteries included. With Arduino there is one more step: You download and run the Arduino IDE. For any typical board like the Uno there is no additional hardware to buy. Given that these boards show up as typical USB CDC devices, you don’t have to install any drivers either.
Of course, mbed came later. I also think that it’s not so much clearly better in the same niche as the Arduino which means that it’s hard to run up against the established default solution, i.e. the Arduino.
I’m not contesting that there was a component of luck to the success of the Arduino. However I remember quite well when I first came across the Arduino ca. 2006: It was a revelation. At that point I already had some embedded development experience. As a beginner you’d either flock to the various vendor dev boards or tear your hair out trying to design your own circuit. Even for vendor boards there was little community and certainly not ecosystem of shields. With the Arduino you got one board that had everything to run the code, made it possible to buy off the shelf hardware, use a C compiler for free, all with almost minimal setup.
There’s also an ecosystem of libraries and documentation with Arduino, which is hard to beat. Similar with the RPi versus the many other SBCs out there; pretty much every Linux tutorial is “how to do $X on pi”.
Yeah but I think this wasn’t the real point of the author. First of all adafruit is (now) already an established player in the space and they are known for above-average stuff[1]. I think this is more about the plethora of startups or people who have this kickstarter idea.
But yes, the tooling is important, otherwise you could just use the microcontrollers on their own, with all their horrible toolchains or other problems for hobbyists.
[1]: phrasing it as neutral as possible without having used it, I think it’s good stuff but I simply don’t know.
These are good points, but Adafruit is already creating multiple lines of dev boards and they catch on.
Benefits with Lady Ada’s boards are:
Solid libraries for each board component on release.
Plenty of tutorials for each board on release.
The best tech team I’ve ever interacted with.
The good points of Adafruit are easiest to see with a subscription to adabox, where every other month you get all the pieces for a working project.
I will happily furnish more citations and information on request, and I encourage you to buy adafruit for quality and library support.
P.S. my software developer workflow has been enhanced by building my process status project on an adafruit neotrellis.
SparkFun has good stuff too — I’ve been using their ESP32 Thing for a while, and they have a whole ecosystem of little plug-and-play I2C peripherals that don’t even require breadboarding.
Pretty much every dev board I’ve looked at satisfies this article’s first two criteria, I.e batteries-included [sic] and easy flashing. And I’m doubtful about the other criteria; they seem aimed too much at beginners.
The description of the Arduino bootloader sounds incredibly primitive compared to the mBed boards. They appear as mass-storage devices and you just drag and drop the image. They contain a shortcut file that opens up a web-based IDE, which lets you compile for your board from a web browser and then download the final results as a file that you can drop onto the attached USB mass-storage device and then hit the reboot button. That’s far more batteries-included than the article describes (you don’t need any custom hardware or to install any software, you just need a USB port and a web browser), yet mBed boards don’t seem to be particularly popular. The libraries also look like the ones described in the article, the price-point is about the same, and there is a reasonably active community (though that’s a bootstrapping problem: Arduino presumably didn’t have one on launch, you only get a community once the board is popular and so it can’t be a prerequisite for a board becoming popular).
The whole article reads like something I’ve read a lot of times: someone who was successful by being in the right place at the right time wanting to justify their success as something other than luck.
The mbed flow is probably as integrated as something that requires no installation can get.
On the other hand, I don’t feel that it’s far more batteries included. With Arduino there is one more step: You download and run the Arduino IDE. For any typical board like the Uno there is no additional hardware to buy. Given that these boards show up as typical USB CDC devices, you don’t have to install any drivers either.
Of course, mbed came later. I also think that it’s not so much clearly better in the same niche as the Arduino which means that it’s hard to run up against the established default solution, i.e. the Arduino.
I’m not contesting that there was a component of luck to the success of the Arduino. However I remember quite well when I first came across the Arduino ca. 2006: It was a revelation. At that point I already had some embedded development experience. As a beginner you’d either flock to the various vendor dev boards or tear your hair out trying to design your own circuit. Even for vendor boards there was little community and certainly not ecosystem of shields. With the Arduino you got one board that had everything to run the code, made it possible to buy off the shelf hardware, use a C compiler for free, all with almost minimal setup.
There’s also an ecosystem of libraries and documentation with Arduino, which is hard to beat. Similar with the RPi versus the many other SBCs out there; pretty much every Linux tutorial is “how to do $X on pi”.
Yeah but I think this wasn’t the real point of the author. First of all adafruit is (now) already an established player in the space and they are known for above-average stuff[1]. I think this is more about the plethora of startups or people who have this kickstarter idea.
But yes, the tooling is important, otherwise you could just use the microcontrollers on their own, with all their horrible toolchains or other problems for hobbyists.
[1]: phrasing it as neutral as possible without having used it, I think it’s good stuff but I simply don’t know.