[Perl5-syntax] Method::Signatures, now with types

[Buddy Burden]

Michael,

>> Labels and placeholders I
>> hadn't considered, because I don't use them personally, but I could
>> take those on.  Labels seems easy.  Placeholders perhaps a _bit_
>> trickier, but doable.  Unless you think they shouldn't be added, of
>> course.
>
> I haven't seen a compelling need for them, they all seem to solve really
> trivial edge cases, and I'd rather not add syntax that we don't need but will
> be stuck with.  Also I don't know if they're forward compatible with Perl 6.
> I think labels are, but I don't know about placeholders.

Yes, I'm fairly positive I saw the labels syntax in the Perl6 synopses
somewhere, but the placeholders one seems odd to me (hadn't even
noticed it  before you pointed it out).

> And I don't know how much they're really used.

Yeah, the idea behind placeholders in particular seems like just to
avoid a variable declaration, which doesn't seem that useful.

> If that's your itch, ok, but there's more pressing things to do.

Heh.  Well, I've added labels to my list, since it's something that
Perl6 will allow, but it's at the bottom.  Placeholders I will
probably not add unless someone makes a compelling case for them.

> I don't know of a specific incompatibility.  I just know it's different and
> some folks reeeeeeeally want methods to be declared at runtime.  It's trivial
> to give them that option in MS.

If you say so. :)  I'm not 100% sure how to tackle it, but we can
always talk about that later.

>>> Finally, MS and MXMS don't seem to play well together in the same file.
>>
>> Yeah, I definitely think it should be either-or.  Trying to mix them
>> would engender similar problems to trying to mix Mouse and Moose,
>> IMHO.
>
> This turned out to be a Devel::BeginLift bug.
> https://rt.cpan.org/Ticket/Display.html?id=66765

Ah, interesting.  I wonder if that's the cause of the MXMS bug on 5.12
(https://rt.cpan.org/Public/Bug/Display.html?id=61070 and
https://rt.cpan.org/Public/Bug/Display.html?id=63594).  Although MS
doesn't seem to have this bug, and it also uses Devel::BeginLift.


            -- Buddy