[sdiy] STM32 Nucleo + mbed

Nils Pipenbrinck n.pipenbrinck at hilbert-space.de
Mon Dec 14 13:26:30 CET 2015


On 12/13/2015 06:23 PM, Vladimir Pantelic wrote:
> On 12/13/2015 06:02 PM, Robert Spencer wrote:
>> "compared to a simple "make program" that has my code compiled and
>> uploaded in seconds.”
>>
>> How long did it take you to get to that point though? ;)
> 
> uhm, give or take a decade or so :))

I've worked with mbed for 8 month and can say a thing or two about this
issue because I'm from the makefile family as well. Even the idea of a
web IDE makes me nervous. I want to have access to my code even if the
net is down.

The good thing is, you don't have to use the web IDE. You can set up
your project on the web IDE once (aka just add the libraries you want)
and then export the project as a whole. It worked for me on the first
try when I exported to vanilla ARM-GCC. I just exported, unzipped the
archive, typed make and got my binaries.

The auto-generated makefile is even well written and nice to work with.
There are no obscure scripts that do undocumented things or anything
like that. Just pure, clutter free make.

Mbed itself becomes just a well tested peripheral library and
(optionally) a wrapper library around CMSIS-RTOS that way.


Regarding mbed in general: It's easy to throw together simple stuff. C++
really simplify things here. That is a good and a bad thing: Bad because
as soon as you want to use chip specific features you're on your own and
you have to write your own peripheral drivers.
The good thing on the other hand is, that Mbed lets you do that just
fine. There is no hidden layer in mbed that does unexpected things like
reprogramming the peripheral clocks behind your back or so.


I'd use it again, especially for fast prototyping.

/Nils



More information about the Synth-diy mailing list