Why Arduino users so agressive?

[SiliconWizard]

The guys that know their stuff aren't out there preaching; they're tired of dealing with dummies. It's the lower level disciples that go out and do the preaching.

Isn't that how that also works in sects?

[TK]

I don't think they are more aggressive than EEVBLOG members, or the whole internet as a matter of fact

[rstofer]

The guys who really know their stuff (not me), tend to help when asked. And their answers are infuriatingly like they came out of a datasheet. If I could understand the manual, I wouldn't be asking! The reason for this is they understand the stuff so well, they give answers that are carefully crafted to be unequivocally correct. And this is how technical documents are written.

And the alternative is an answer that is sloppy, easy to understand and wrong under multiple conditions.  This stuff isn't easy to learn and to know it to any depth takes a lot of time and a whole lot of mistakes.  How many hours does it take to learn that you have to apply power to a certain peripheral in the PCON register?  Unless you read the datasheet, and read it very carefully, you can be stuck on this for days.

Who has time to digest a multi-thousand page User Manual just to blink an LED?  Think about the difference between blinking an LED on an Arduino and doing it from bare metal (no HAL) on an ARM.  There's a HUGE difference and it is nearly insurmountable without a good deal of experience.  If the only thing you have is the User Manual for the ARM, it's going to be weeks before you get it to run.  This is especially true if you need to understand crt0.S (from ARM7TDMI like the LPC2106 or LPC2148) and then the ever popular linker script.  Try creating that from scratch.  Heck, just try to create a realistic Makefile that includes a target for device programming.

Nope!  The Arduino, within its capability, wins hands down!

[aandrew]

Has anyone seen an agressive Arduino user? Usually I notice the bare-metal zealots dribbling. My prediction is the anti-Arduino crowd will gather here. Which might have been the intention of the invitation.

I scoffed at Arduino for quite a while when it first started picking up. While I don't consider myself a "bare-metal zealot" I do do this professionally and I think that's what was causing my disdain: it was all hacked together kludge upon kludge to do things that could have been done so much better and with less. Then I started to realize that Arduino really brought a lot of people to electronics and computer programming, and in good ways, too. They were learning embedded, not just programming or electronics. Kids were building things. Adults were building things. Old mechanical guys who were afraid of electronics/programming were building things. Arduino took a lot of the initial pain away and for a lot of applications you don't really need more than an Arduino and some kind of shield to get the I/O where you need it. This isn't a bad thing at all; it is in fact a great way to solve a lot of problems. Those who want to take things to the next level can now do so, where before they'd be stuck behind the start line, unsure or overwhelmed with how to begin.

I really do appreciate what Arduino has done for my industry.  I do still roll my eyes a bit at the "I used an Arduino to blink an LED" type projects when that's all the goal of the project was, but I tend to do it only when I'm alone. :-)

[aandrew]

4. Try and build a fucking warp capable space ship because you know, you can flash an LED.

So much this. Flashing the LED isn't the hard part, it's learning how to manage your own expectations and not get discouraged that turns someone into a budding engineer.

[langwadt]

Based on what I see on the Arduino forum at https://forum.arduino.cc/index.php#c2
only a minority of Arduino users are doing it as a hobby.

As of today I think the breakdown is roughly as follows:
90% of the posts are from students who have been given a project that they need to get done ASAP so they can pass their course. They just want an answer on a plate, preferably source code they can copy and past into their project. Having to understand things is a low priority. For this audience spending 10 minutes doing a Google search is too much effort.

I think you give them too much credit assuming they know how to use google or that answer on a plate isn't too complicated

[Howardlong]

Best part about Arduino is

That’s right for pretty much the entire user base. And I don’t mean that in a derogatory way, I do exactly that when I want a quick and dirty simple automation one-off Arduino solution for my own use. I can’t imagine me using one in a commercial product though, although I know many have.

[james_s]

I also scoffed at Arduino for a long time, but then it got to where there are SO many ready to use libraries that I had to admit it has value. I can get any number of random widgets from China and more often than not I can find a library, wire it up and have it doing something in minutes.

My main complaint about these libraries is that they are often poorly documented. I can try to decipher the example programs but they are often poorly commented and may only use a few of the capabilities of the library.

[KL27x]

Rstopher, I know, right? Assembly (and more specifically, the device-specific trappings) are so complicated that spitting out near verbatim datasheet answers becomes natural, efficient, correct. I'm not criticizing. It's like at that level, you aren't even speaking in a language that normal human beings can recognize. Your saying exactly what you're thinking, the way you understand it.

Weeks? Yep. Took me months to blink a LED on a PIC. That's an obvious drawback. The only good thing about it is I know (or I knew) a process to do things on the lowest level if/when it mattered. For latency or power consumption or w/e. The problem is this knowledge is very device specific, and the devices change. Even the dev tools change. Heck, even the datasheets change. (Microchip stopped putting an index in their datasheets; and this was a big part of how I managed information). So it becomes almost the same challenge all over again to repeat the feat.

So to this day, some of the technically awesomest projects I have done? Super low power consumption, super low latency multitasking to get everything working on a single chip, automatic battery cutout, deep sleep, working like a peach? It's too much trouble to incorporate these features fully, all the time. It has to be very important and/or a very huge project (in production numbers) to bother. So for many projects I agree that Arduino can be awesome. I don't take a side on this. Nothing wrong with Arduino. But I'm sure it comes with its own problems, as many have noted.

[NorthGuy]

... I do exactly that when I want a quick and dirty simple automation one-off Arduino solution for my own use. I can’t imagine me using one in a commercial product though, although I know many have.

I do this all the time too. Except I copy from my past projects. And I never use Arduino. Works better that way

[bd139]

I also scoffed at Arduino for a long time, but then it got to where there are SO many ready to use libraries that I had to admit it has value. I can get any number of random widgets from China and more often than not I can find a library, wire it up and have it doing something in minutes.

My main complaint about these libraries is that they are often poorly documented. I can try to decipher the example programs but they are often poorly commented and may only use a few of the capabilities of the library.

Yes and no. Last two times I used Arduino libraries I got so pissed off with the utterly awful fucking bugs and bloated code output I rewrote my own in AVR-libc. I've written libraries for Si5351 and HD44780. Also it's a million times easier to throw together event and interrupt driven code without being punched in the balls over and over again. Debouncing when handling other interrupts is an example of kicking you in the danglies.

The hardware and toolchain underneath it is pretty good. I use avr-gcc toolchain and avrdude. Rest of it is like attempting to date a hungry crocodile.

[KL27x]

One thing that made me go "hmmm" about arduino is the boards. Yes, you can remove some unused pins from the header, arrange them in a consistent order, and with a nice silkscreen. But seems like a DIP chip would be almost the same thing. And more versatile.

I once worked with a project on which a test gear was done by an Arduino-based tool. The PSU died on it, and my client came to me to get production back up and running. I rigged up a USB supply with integral li ion battery and DC boost to 5V, cuz why not?

Well, one day, I run into the guy that developed this gear. He had seen it, and he asks me about the nifty solution I came up with. I'm explaining what I did, and he says, "How does it run on 5V? It needs 7, doesn't it?" That's when I see, for a "software guy," Arduino board just handles things, and they don't have to think about it. But... where's the line between making things easier and adding layers of obfuscation?

[Mr. Scram]

It seems OP hasn't bothered to specify what he means when he takes about agressive Arduino users. He hasn't bothered replying at all in fact.

[NorthGuy]

Who has time to digest a multi-thousand page User Manual just to blink an LED?

I start every project from blinking LED. However, the blinking LED is never the end goal. The more you want to accomplish, the harder it gets. In most cases the cookie-cutter solutions don't work for me.

If I want to hang a huge painting above my bed, I don't care how to do it easily - say without drilling, or without any tools, or with fewer wacks of a hummer. What I care about is that the painting is straight, solid, and will never fall on my head while I'm sleeping. And I'm prepared to do whatever is necessary to accomplish my goal.

I have similar approach to programming. I have requirements which must accomplish certain goals. So, I work on that. If I need something from the datasheet, I have no problem finding it there. If I need to use assembler, I'll do it. If I need to do a research to find a better MCU, I'll do it. Whatever it takes to meet the requirements. Would I save time if I used Arduino? I don't see how.

[rstofer]

One thing that made me go "hmmm" about arduino is the boards. Yes, you can remove some unused pins from the header, arrange them in a consistent order, and with a nice silkscreen. But seems like a DIP chip would be almost the same thing. And more versatile.

The Arduino Nano comes in a 'stamp' format with headers more suited to installation on daughter cards:

https://www.amazon.com/ELEGOO-Arduino-ATmega328P-Without-Compatible/dp/B0713XK923

I especially like the mbed LPC1768 for this kind of thing.  All I have to do is design the daughter card with project specific components and plug in the MCU

https://os.mbed.com/platforms/mbed-LPC1768/

In fact, all I need to add is a MagJack and ethernet is up and running.  The mbed project furnishes the lwIP stack.

[rstofer]

It seems OP hasn't bothered to specify what he means when he takes about agressive Arduino users. He hasn't bothered replying at all in fact.

Nearly all of his posts are provocative.  He gets scolded from time to time.

[KL27x]

Rstopher, I don't get it, though. If you're gonna do that, why not do away with the Arduino board, entirely, and put the nekkid chip on your board? And add the few external components (the ones that you need).

If you are repeatedly designing stuff within Arduino, and you want to recycle that footprint, you could even make the entire thing a library component, perhaps? (I guess it depends on your PCB software). So you plop that into your schematic and and then drag and drop it onto your board.

Humm.. maybe Arduino boards are so cheap it saves on assembly time/cost to buy and install the entire boards? 

[KL27x]

I have similar approach to programming. I have requirements which must accomplish certain goals. So, I work on that. If I need something from the datasheet, I have no problem finding it there. If I need to use assembler, I'll do it. If I need to do a research to find a better MCU, I'll do it. Whatever it takes to meet the requirements. Would I save time if I used Arduino? I don't see how.

I think it could save time on your next project... as long as Arduino can meet the requirements. Learning Arduino, that knowledge is somewhat consistent going into the future. Knowing how to program a chip to the bare metal is useful when it's useful. But unless you want to use Z80 forever, that knowledge is not quite as portable. You slog through it to meet the requirements of a given project. On the next project, you slog through it all over again. As AVR continue to evolve, someone over at Arduino Co will incorporate newer, faster, cheaper micro into more or less stable platform, I think. So as the user, you will mostly find the switches to be in the same position, with the same names, and the same functions. With only a few new ones. 

My POV is obviously super biased towards my own narrow experience.

[bsfeechannel]

Best part about Arduino is

When you write

#include <stdio.h>

Aren't you doing the same thing?

[rstofer]

Rstopher, I don't get it, though. If you're gonna do that, why not do away with the Arduino board, entirely, and put the nekkid chip on your board? And add the few external components (the ones that you need).

If you are repeatedly designing stuff within Arduino, and you want to recycle that footprint, you could even make the entire thing a library component, perhaps? (I guess it depends on your PCB software). So you plop that into your schematic and and then drag and drop it onto your board.

Humm.. maybe Arduino boards are so cheap it saves on assembly time/cost to buy and install the entire boards? 

There are complications when dealing with high speed signals (LPC1768 not Arduino) and, in the case of the LPC1768, there are components on both sides of the board.  How do I reflow that?  I'm a hobbyist, not a manufacturer.

Shorter answer:  I don't want to design the MCU part of the project.  I'll leave it to others.  Another short answer:  I consider my time to be worth $100/hour and I make all buy versus build decisions on that basis.  Lots of other decisions too!  I'll design something if I have to but not if I can buy it.  Arduinos are particularly cheap.

Consider too that when a project is no longer meaningful, I can recover the MCU board and move it to another project.

Arduinoo Nano's are especially cheap at around $4.30 with free shipping.  How much design can I do for $5 - about 3 minutes at my billing rate.  There's a reason my billing rate is so high!  It keeps me from making dumb mistakes like misusing what little time I have left.  My "Best Used By Date" is approaching and there's no point in wasting my time trying to reinvent the wheel.  Heck, I can't even order the parts in 3 minutes!

https://www.amazon.com/s?k=arduino+nano

[bd139]

Best part about Arduino is

That’s right for pretty much the entire user base. And I don’t mean that in a derogatory way, I do exactly that when I want a quick and dirty simple automation one-off Arduino solution for my own use. I can’t imagine me using one in a commercial product though, although I know many have.

Yep it’s why we write and share programs.

I’ve used one for automated testing of germanium transistors to palm off on eBay to guitar pedal builders. Literally saved me hours and hours of work. One of them, a cheap UniT Ut61E and python program that told the arduino to switch on a relay then give a go or no go.

But yes in a commercial project or something where you need better control over system interrupts or actually need to understand how broken the arduino libraries are then is run a mile.

Best part about Arduino is

When you write

#include <stdio.h>

Aren't you doing the same thing?

If you’re using glibc then yes. Bloated POS 

[rstofer]

Another strategy for dedicated ATmega328 (DIP package) projects is to program the device on the Arduino board and just plug the chip into the main project board.  As long as you don't need the USB->Serial Terminal support, this works well.

[NorthGuy]

Learning Arduino, that knowledge is somewhat consistent going into the future.

Many years ago I though that my "knowledge" was going to go with me into the future. Since then, so many things changed on me, so I learned to embrace the change. Besides, I found out that if you figure out how to work with one MCU, you can easily transfer your skill to others. I'm pretty sure that if I get a new MCU which I've never seen before, I can figure it out rather quickly. Things are still changing very quickly - if you get fixated on something particular, you're destine to live in the past.

[KL27x]

^The problem, though. Do I get wiser with age? Maybe. Slower and dumber? Definitely.

[NorthGuy]

Arduinoo Nano's are especially cheap at around $4.30 with free shipping.  How much design can I do for $5 - about 3 minutes at my billing rate.

For some projects $4.30 might be huge overkill as under-$1 PIC will work just as well (or better). For other projects you may need something for $10. There are thousands if different MCUs which are used for different purposes.

...  there's no point in wasting my time trying to reinvent the wheel.

But a tricycle wheel may not fit on Ferrari.

Another strategy for dedicated ATmega328 (DIP package) projects is to program the device on the Arduino board and just plug the chip into the main project board.  As long as you don't need the USB->Serial Terminal support, this works well.

You can program almost anything in-circuit. What's the point of dragging DIP chips back and forth?