Pixl.JS 'D0' pin is false-y?

I think the subject says all.
Here is the output I got testing couple of pins:

>reset(1)
Erasing saved code.
Done!
=undefined
 ____                 _
|  __|___ ___ ___ _ _|_|___ ___
|  __|_ -| . |  _| | | |   | . |
|____|___|  _|_| |___|_|_|_|___|
         |_| espruino.com
 2v02 (c) 2018 G.Williams
>!!A0
=true
>!!A1
=true
>!!D0
=false
>!!D1
=true

// and here comes the fun part:
>!! (D0.getMode)
=true

The built-in pins are a special Pin type, but they can also be converted to numbers.

For instance: digitalWrite(1,0) does the same as digitalWrite(D1,0) (on many platforms) to behave the same as on Arduino.

That means that D3==3 and D0==0 - and when you convert a pin to a boolean, there's no specific conversion for bools so it's converted to an integer first, which is then converted like an integer would be (eg !!D0==!!0==false).

If you want to check if a pin is defined, it's best to do pin!==undefined

Having said all that, JS does have some hacks in it - for instance "0"-0==0 but !!"0"==true - so it may make sense to add similar functionality to Espruino?

Thanks for the explanation, so it's expected and by design?

But that means, that a couple of libraries that support optional pins, just "accidentally" work? IIRC in most cases the check is usually if(options.whateverPin){...}
For example in the CCS811, it's

if (ccs.options.int) {
...

Thinking back, I may have had an issue with A0 on the Pixl, but just moved to another pin without thinking too much. Or maybe not :)

Consistency with JS I think usually would be beneficial, for example setting the length of an array in a browser empties that array. But creates a length property with value of 0 in Espruino :)

Don't know how much work would be to add a special handling from Pin to bool.

Thanks - I've just changed it, so it should be in 2v03 - as you say, it seems there's a bunch of stuff that relies on it