Re: Proposal: ValidateMonoLiterals - Initial bikeshed discussion

And no one of my proofreaders noticed that >.> I would propose to have the extension replace the 'fromString "foo"', 'fromIntegral 5' and 'fromList [1,2,3]' calls (for monomorphic cases) in the AST with the relevant Typed TH splice. I considered quasi-quotation initially too, but there's no quasi quotation syntax for Typed TH. I'm guessing that's just an oversight, but I'd really be in favour of adding a typed quasiquoter too. Similarly to thinking we should have an easier way to obtain Lift instances since, to me at least, it seems that the Lift instance for most ADTs should be fairly trivial? I'll quickly clarify the proposal on the wiki :) Cheers, Merijn

On 5 Feb 2015, at 22:48, Simon Peyton Jones wrote:

I'm all for it. Syntax sounds like the main difficulty. Today you could use quasiquotatoin [even| 38 |] and get the same effect as $$(validate 38). But it's still noisy.

So: what is the non-noisy scheme you want to propose? You don't quite get to that in the wiki page!

Simon

| -----Original Message----- | From: ghc-devs [mailto:ghc-devs-bounces@haskell.org] On Behalf Of Merijn | Verstraaten | Sent: 05 February 2015 14:46 | To: ghc-devs@haskell.org; GHC users | Subject: Proposal: ValidateMonoLiterals - Initial bikeshed discussion | | I've been repeatedly running into problems with overloaded literals and | partial conversion functions, so I wrote up an initial proposal | (https://ghc.haskell.org/trac/ghc/wiki/ValidateMonoLiterals) and I'd like | to commence with the bikeshedding and hearing other opinions :) | | Cheers, | Merijn