What stops your Arduino project?

"...Sometimes beginners use a hundred lines of code when they could have collapsed it all into a 5-line 'for' loop for instance, or creating functions..."

I think that sometimes happens because beginners don't realise how quickly things tend to grow out of control. Also because they are often working from small snippets of example code from beginners guides and chaining it together into something a bit larger without always considering the overall program design. After several hours and once they get to a few 1,000 lines of something that appears to be working, it can be daunting to hack it all apart to try and tidy it up, with a good chance of breaking it in the process.

For a beginner it can also often be difficult to firstly identify well-written code and then secondly to understand it.

I think there is room for a more structured course based around Arduino as the platform. The official starter kit is great for getting people interested in Arduino but there ideally needs to be something that follows on from that which helps develop better programming and interfacing skills.

If anyone has seen David Beazley's Python Programming Language Live Lesson course, then a lot can be learned from that style. He starts with the basic way of doing stuff but then is always revealing a better way, whilst talking you through as to why it is better each time. He also adds context by using useful everyday examples where you can use these techniques. 

Perhaps slightly controversial, but another course worth mentioning is Atmel's Getting Started with AVR by Glen Nilsen. It uses the Atmel ATmega 328P Xplained/Atmel Studio instead of the Arduino Uno/IDE but it leverages the hardware side of the ATmega 328P. Most of the main program loop is empty as the on-chip peripherals have been set up to do all the heavy lifting in the initial setup routine.

Another question just to balance the flames is.....why do you choose to use an arduino in your project. There are lots of great reasons, usually more pros than cons.

Good point!  Like many engineers, I like to talk about and work on problems so those tend to get my attention.  They aren't intended to be flames, but they might come across that way.  I use the Arduino libraries and boards because they get me started faster and are often suitable for my needs.

What typically kills my Arduino projects is finding something super cool that I would love to build - only to find out the hardware hasn't been available for a few years - or more.  My next greatest challenge is how to package it.

It sure would be nice to know if the code examples are the most recent methods or are from IDE 1.0.

Hi Enrique,

I agree, but there comes a point when some extra help is needed, otherwise even the basics for real-time, such as button-debouncing, display updating and so on, are in a long unmaintainable loop. The single loop() is fine to begin with, without the clutter.

But for next steps, without a template or reminder of how to organize code may be hard.

I suppose I should add, technically a long loop can be fine : ) I worked on a communications product where the code (not written by me) was in fact a long loop.. one function call after the next, there must have been 40 of them in the loop : ) It was in equipment intended for mission-critical purposes : ) It was certainly hard adding new features, and today things can be done differently (there was no OS in use in that product) but good code can be written in a long loop, my comment above was very general.

With all these additional comments you want to add then your loop could potentially get even longer

However it sounds like a good topic for further discussion.

Are you sure that isn't more like pouring gasoline onto the fire  ? 

Some factors may be:

What you already have to hand and how much resource you have already invested in that platform be it digesting the reference manual, buying shields, writing libraries, etc.

How much support/documentation/training is readily available, especially in the area you are working in. Has someone published a similar project that you can use as a starting point to speed things up. If starting from scratch, is there sufficient documentation to do so.

Platform roadmaps - is this platform still likely to be supported and actively developed in twelve months time.

Platform features - does it have all the features that the project requires. Is the platform easily scalable if additional features are required.

Costs - there may be cheaper competition now but how much will it cost overall to move to a new platform.

Development environments - does the IDE still satisfy the programming requirements.

Form factor - will it physically fit into the project space.

Oh, and what did the Element14 team send you in the post to experiment with...

I personally don't like the IDE it kinda sucks!! I wish the thing had just used C so we could have just used Eclipse. I have tried to build a protocol converter with it.  But running CAN and Ethernet is a pain. So for this project, I chose one of my  Raspberry Pies. It has built-in Ethernet and I can attach a CAN module via SPI. Fins.

Oh, come on now. duh its called "printf ie printf("variables LAT %f6, LON%f6, ALT%f6, pitch %f6, roll f6, yaw f6",
lat, lon, alt, pitch, roll, yaw" );
if you want to be fancy you can create a function  void debug () { line above; return;}
in C even if you have a function of a void type you can still add the return statement.