[e-lang] Count of complex guards

[Dean Tribble]

Mark Miller wrote:

>> nullOK[foo] doesn't work
>
> What do you mean, "doesn't work"? It doesn't work computationally? Or 
> it doesn't read well? If the latter, well, after being corrupted a bit 
> by ML, I've considered

I'm glad you didn't see it either.  The "K" is capitalized, so it 
doesn't name a guard.

>     optional[foo]
> It's more declarative perhaps, but its operational meaning is less 
> clear. It also doesn't mean what it means in ML, since the non-null 
> case doesn't need to be unwrapped. I'm inclined to stick with nullOk.

I thought of proposing "optional", but that usually means that the arg 
doesn't need to be present at all, which is not what we mean here.

> > Yup (that was easier than I expected :-)
> Much. We can't possibly be converging this quickly. We must be 
> misunderstanding each other. ;)

:-)