On 24/09/2006, at 1:46 AM, Michael Shulman wrote:

On 9/23/06, Bernie Pope wrote:

...

If a pattern binding is not simple, it must have a data constructor on the lhs, therefore it cannot be overloaded. So the (dreaded) MR only applies to simple pattern bindings.

I thought it was simple pattern bindings that could be *exempted* from the MR by supplying an explicit type signature, while non-simple pattern bindings are always subject to it.

Actually, I realised after I posted my message that what I wrote was dubious. I'm sorry about that, please ignore it. This section from the Report seems to clear things up: In "Motivation", section 4.5.5: "Rule 1 prevents ambiguity. For example, consider the declaration group [(n,s)] = reads t Recall that reads is a standard function whose type is given by the signature reads :: (Read a) => String -> [(a,String)] Without Rule 1, n would be assigned the type forall a. Read a =>a and s the type forall a. Read a =>String. The latter is an invalid type, because it is inherently ambiguous. It is not possible to determine at what overloading to use s, nor can this be solved by adding a type signature for s. Hence, when non-simple pattern bindings are used (Section 4.4.3.2), the types inferred are always monomorphic in their constrained type variables, irrespective of whether a type signature is provided. In this case, both n and s are monomorphic in a." Sorry for the confusion, Bernie.

On 26/09/2006, at 12:19 AM, Christian Sievers wrote:

Bernie Pope answered:

...

...

1. Why do the rules of the monomorphism restriction explicitly mention *simple* pattern bindings? Where is the difference, especially as there is a translation to simple pattern bindings? Why should

p | "a"=="b" = 2 | otherwise = 3

be treated different than

p = if "a"=="b" then 2 else 3

They are the same (both are simple pattern bindings). The report says in section 4.4.3.2 that the first can be translated into the second.

Indeed, I meant to allude to this translation.

...

A simple pattern binding is one where the lhs is a variable only.

That's consistent with the second reason for rule one of the MR.

However, the mentioned section 4.4.3.2 defines it differently:

A simple pattern binding has form p = e.

And if there is any doubt about what p stands for, it goes on:

The pattern p ...

Contrasting to that:

The general form of a pattern binding is p match, where a match is the same structure as for function bindings above; in other words, a pattern binding is:

p | g1 = e1 | g2 = e2 ... | gm = em where { decls }

So according to this definition, a pattern binding is simple iff there are no guards (unless they are in the expression). Also the translation to a "simple pattern binding" only gets rid of guards.

So there seems to be an error in the report, which can be fixed by either redefining "simple pattern binding", or using a differnet description in the MR.

Aha, Christian I see what you mean. It seems I did not read section 4.4.3.2 carefully. In fact I think I was interpreting that section in light of the MR. So I am now as puzzled as you are. Anyway, thanks for persisting on this point. Cheers, Bernie.