On 31 January 2006 17:48, Andrew Pimlott wrote:
On Tue, Jan 31, 2006 at 10:17:57AM -0000, Simon Marlow wrote:
On 30 January 2006 21:49, Andrew Pimlott wrote:
In the present case, people aren't (only) opposing the M-R out of principle, but because they actually use overloaded variable definitions and (at least sometimes) want to leave off the signature.
So I don't see how one could claim, as on the wiki, the warning "wouldn't happen much". I suspect it would happen, and annoy people, and defeat the reason that people want to remove the M-R.
The assertion that it "wouldn't happen much" is based on the observation earlier in this thread that it was actually difficult to write some code that illustrated the problem.
This indicates that the warning "wouldn't happen much" _when you want sharing_. But it would happen all the time when you don't want sharing, eg. in the case Malcolm Wallace just posted.
So you either add a type signature, or turn off the warning. What's the problem? I suspect you're saying that you don't want a warning by default, and you don't want the langage to recommend that compilers issue a warning by default, right? If so, your objection is duly noted and I'll add the point to the wiki. Cheers, Simon
On Wed, Feb 01, 2006 at 10:32:26AM -0000, Simon Marlow wrote:
On 31 January 2006 17:48, Andrew Pimlott wrote:
This indicates that the warning "wouldn't happen much" _when you want sharing_. But it would happen all the time when you don't want sharing, eg. in the case Malcolm Wallace just posted.
So you either add a type signature, or turn off the warning.
Right (provided you understand the warning). Though I don't think many people would turn off a default warning--it's just too much trouble. (I always add dummy definitions of coarbitrary to my Arbitrary instances, rather than figure out how to suppress the undefined method warning.) So in practice, I think most people would add a type signature.
What's the problem?
I've learned from this sub-thread that people oppose the M-R for at different reasons. Some dislike it on principle, or on aesthetic grounds, and they would be happy to add a type signature. But others think it would be useful to define polymorphic variables without a type signature, and they would not be happy to add one.
I suspect you're saying that you don't want a warning by default, and you don't want the langage to recommend that compilers issue a warning by default, right? If so, your objection is duly noted and I'll add the point to the wiki.
Cool. Andrew
participants (2)
-
Andrew Pimlott -
Simon Marlow