Haskell Platform Proposal: HLint

[Neil Mitchell]

Hi,

>> The only problem is that it will take some work to adapt
>> haskell-src-exts to Haskell 20nn. I spoke to Niklas about this.
>> He has in principle volunteered to take on the task, but
>> he can't start on it for a few months. And I don't think it
>> would be wise to make the package name changes and
>> end-of-life the current haskell-src until this is done.
>
> Yes, I agree with both Duncan and Yitz here. It would be some monkey
> work to introduce a "simple" AST, but nothing tricky in principle. I
> wouldn't mind adding such an AST to haskell-src-exts, and I think
> that's the best solution long-term.

Niklas has got a lot on his plate without doing this simple AST - I've
filed at least 3 bugs against his libraries in the last week!

I'm not entirely sure what the intended audience is for this simple
AST. Unless someone's going to use it, I don't see much point in it
existing. I regularly use haskell-src-exts without caring about lots
of extensions, and it just works. Having all the extensions there just
means that when I do need them I don't need to switch library/module
at all.

>> So unless someone else can do the work, we may have
>> to wait until the following HP release.

I see no reason to even suggest this work be a precondition for
inclusion in the platform. I use HSE more than most people, and I'd
consider a simple AST to just be adding bulk. Without a compelling use
case it seems you're burning the universes entropy for no good reason
:-)

Thanks, Neil