January 2011 - Libraries

Proposal: Applicative => Monad: Call for consensus
by John Smith 28 Jan '11

28 Jan '11

The patches attached to http://hackage.haskell.org/trac/ghc/ticket/4834 make Applicative a superclass of Monad. Default definitions are provided for backwards compatibility. Advantages: The class hierarchy correctly models the logical relationship between Applicative and Monad. Boilerplate Applicative instances that duplicate Monad functions will no longer be required. Disadvantage: All Monads will be Applicatives, even if you're only interested in the Monad interface. This is akin to all Ords being Eqs, even if you're not using ==. Backwards Compatibility: Existing Monad definitions will have to be changed to declare Applicative. The original function bodies can be used unchanged, but they are now declared in a different class. Calling code should work unchanged. This ticket has already been discussed on the mailing lists. The purpose of this message is to call for consensus by the deadline of 1 February. Please note that although the ticket references the wiki page at http://haskell.org/haskellwiki/Functor-Applicative-Monad_Proposal, the patches attached to the ticket are much more conservative than the more ambitious reforms on the wiki. (I would like to change the ticket description to make this clearer, but I can't edit it.)

29 104

Use explicit language extensions in base
by Bas van Dijk 27 Jan '11

27 Jan '11

Hello, I just wrote a patch that removes the extension fields from base.cabal and adds LANGUAGE pragmas where needed: http://hackage.haskell.org/trac/ghc/ticket/4925 This has the advantage that modules become more modular (since they don't need to depend on external make or cabal files which specify how to build them). For the record: this is not a library proposal. It's just an internal patch for base. It would be great if someone could review and apply it. Regards, Bas

2 1

Proposal: Add missing Foldable/Traversable instances for Prelude types
by Edward Kmett 27 Jan '11

27 Jan '11

I'd like to propose adding the following missing instances to Data.Foldable and Data.Traversable respectively, since there isn't a canonical location that they could be added outside of those packages without orphans, their definition is unambiguous, and they are quite useful (plus, I happen to need them for a monad transformer in my adjunctions package). instance Foldable (Either a) where foldMap f (Left a) = mempty foldMap f (Right b) = f b instance Foldable ((,)e) where foldMap f ea = f (snd ea) instance Traversable (Either a) where traverse f (Left a) = pure (Left a) traverse f (Right a) = Right <$> f a instance Traversable ((,)e) where traverse f ~(e,a) = (,) e <$> f a Discussion period: 2 weeks -Edward Kmett

8 9

transformers
by Edward Kmett 27 Jan '11

27 Jan '11

Do changes to the 'transformers' package fall under the purview of the libraries change control process? -Edward

4 7

question about cabal
by Kathleen Fisher 26 Jan '11

26 Jan '11

Hi all, I have another question about cabal. (Thanks for the information about whether cabal supports two cabal files in one directory). I have a library pads-haskell with a cabal file pads-haskell.cabal. If I ask to install this library from the current directory: > ~/.cabal/bin/cabal install -w ~/sw/ghc-7.01/bin/ghc --package-db=/Users/kfisher/sw/ghc-7.01/lib/ghc-7.0.1/package.conf.d I get the following error messages: > Resolving dependencies... > cabal: cannot configure mainland-pretty-0.1.0.1. It requires containers >=0.2 > && <0.4 > For the dependency on containers >=0.2 && <0.4 there are these packages: > containers-0.2.0.0, containers-0.2.0.1 and containers-0.3.0.0. However none of > them are available. > containers-0.2.0.0 was excluded because regex-posix-0.94.4 requires containers > ==0.4.0.0 > containers-0.2.0.0 was excluded because regex-base-0.93.2 requires containers > ==0.4.0.0 > containers-0.2.0.0 was excluded because containers-0.4.0.0 was selected > instead > containers-0.2.0.0 was excluded because regex-bytestring-0.0.1 requires > containers ==0.4.0.0 > containers-0.2.0.1 was excluded because regex-posix-0.94.4 requires containers > ==0.4.0.0 > containers-0.2.0.1 was excluded because regex-base-0.93.2 requires containers > ==0.4.0.0 > containers-0.2.0.1 was excluded because containers-0.4.0.0 was selected > instead > containers-0.2.0.1 was excluded because regex-bytestring-0.0.1 requires > containers ==0.4.0.0 > containers-0.3.0.0 was excluded because regex-posix-0.94.4 requires containers > ==0.4.0.0 > containers-0.3.0.0 was excluded because regex-base-0.93.2 requires containers > ==0.4.0.0 > containers-0.3.0.0 was excluded because containers-0.4.0.0 was selected > instead > containers-0.3.0.0 was excluded because regex-bytestring-0.0.1 requires > containers ==0.4.0.0 (even though mainland-pretty-0.1.0.1 is already installed: > /Users/kfisher/sw/ghc-7.01/lib/ghc-7.0.1/package.conf.d: > Cabal-1.10.0.0 > ... > mainland-pretty-0.1.0.1 ) That's question number one: why is it trying to configure a library that is already installed? If instead of asking cabal to install the library, I compile Setup and then do a configure, build, and register in place: > >./Setup configure -w ~/sw/ghc-7.01/bin/ghc --with-haddock=/Users/kfisher/ghc-head/bin/haddock --package-db=/Users/kfisher/sw/ghc-7.01/lib/ghc-7.0.1/package.conf.d > >./Setup build > >./Setup register --inplace Everything works just fine. That's the second question: why does cabal install not work while breaking it into pieces does work? I've run into this in working with other libraries on Hackage, but have never understood why sometimes I had to do the installation by hand. Can someone help me understand what is going on here with cabal and how to fix it so that cabal install will work for my library? If this is already documented somewhere that I missed, please just point me to the relevant documentation Thanks! Kathleen

2 1

Proposal: Add Applicative/Alternative instances to Text.Read.ReadPrec
by Edward Kmett 24 Jan '11

24 Jan '11

I propose adding the missing Applicative/Alternative instances to Text.Read.ReadPrec, as I've wanted to use them now on more than one occasion when manually building Read instances. Discussion Period: 2 weeks

2 1

Text.PrettyPrint.HughesPJ -- has a ByteString version ever been considered?
by Ryan Newton 24 Jan '11

24 Jan '11

Hello libraries [= maintainer of 'pretty' package], I'm a happy user of the pretty printing library, with which I emit large quantities of nicely indented C++ code. Sometimes I append large strings with "text x <>". In this case performance improvements would not go unappreciated. Has there been any thought of a ByteString version? I notice that the source is only 1K lines. While I imagine that you wouldn't want to break the interface in a backwards compatible way, would it be possible to simple give the "TextDetails" type a parameter and then use a StringLike class (with length, append, what else?). Text.PrettyPrint.HughesPJ could keep the same interface, and the generic version could go in Text.PrettyPrint.HughesPJGeneric or Text.PrettyPrint.HughesPJ.Generic or whatever. Cheers, -Ryan P.S. I notice that other packages like mainland-pretty and ansi-wl-pprint are similarly committed to "String".

5 4

IEEE-754 abs and sign
by Christian Maeder 20 Jan '11

20 Jan '11

Hi, should the following hold? abs (-0.0) = 0.0 sign (-0.0) = -0.0 The generic definitions for abs and signum in ghc produce: Prelude> abs (-0.0) :: Float -0.0 Prelude> signum (-0.0) :: Float 0.0 The outcome for "abs (-0.0)" looks irritating. I know it does not matter for comparisons and the documented properties hold: <cite> The functions abs and signum should satisfy the law: abs x * signum x == x For real numbers, the signum is either -1 (negative), 0 (zero) or 1 (positive). </cite> Yet, I think, for numbers with a negative zero (class RealFloat) the top two equations above should hold. Cheers Christian

1 0

Proposal: keep Data.Map.foldWithKey
by Christian Maeder 17 Jan '11

17 Jan '11

Hi, I've now created a proper library proposal to keep Data.Map.foldWithKey. Exchange arguments and make up your mind until Jan. 15th. http://hackage.haskell.org/trac/ghc/ticket/4842 Cheers Christian

8 45

darcs patch: Expand and clarify MVar documentation.
by Edward Z. Yang 16 Jan '11

16 Jan '11

1 patch for repository http://darcs.haskell.org/packages/base: Sat Jan 15 00:06:43 GMT 2011 Edward Z. Yang <ezyang(a)mit.edu> * Expand and clarify MVar documentation.

3 3