November 2012 - Libraries

Changes to Typeable
by Simon Peyton-Jones 26 Nov '12

26 Nov '12

Friends (Ian pointed out that I should have sent this to the libraries list): The page describes an improved implementation of the Typeable class, making use of polymorphic kinds. Technically it is straightforward, but it represents a non-backward-compatible change to a widely used library, so we need to make a plan for the transition. http://hackage.haskell.org/trac/ghc/wiki/GhcKinds/PolyTypeable Comments? You can fix typos or add issues directly in the wiki page, or discuss by email Simon

25 54

Re: [Haskell] ANN: time-1.4.0.2
by José Pedro Magalhães 26 Nov '12

26 Nov '12

Hi Ashley, Sorry to repeat this question, but is there any particular reason why you give handwritten instances of Typeable, instead of deriving them? Thanks, Pedro On Sun, Nov 25, 2012 at 12:14 PM, Ashley Yakeley <ashley(a)semantic.org>wrote: > time-1.4.0.2 is now on Hackage. > > 1.4.0.2 fixes bugs and infelicities in the formatting and parsing in > Data.Time.Format > > For formatting: > > * %0Y, %_Y, %0G and %_G now pad to four characters. > > * Likewise %0C, %_C, %0f and %_f pad to two characters. > > For parsing: > > * Runs of n whitespace characters now match n or more whitespace > characters. Formerly it matched zero or more whitespace characters. > > * %Y, %G, %C and %f match one or more digits > > * %0Y and %0G match exactly four digits. %0C and %0f match exactly two > digits > > General: > > * Now uses mkTyCon3 instead of deprecated mkTyCon (contributed by Chris > Moline) > > Testing: > > * Various infrastructure improvements (Chris Moline and me) > > * QuickCheck properties now tested > > I tried using Cabal's "detailed" test-suite, but it doesn't play well with > QuickCheck and is very slow with large numbers of small test cases. I'm now > making use of the test-framework package, which does play well with > QuickCheck. > > -- Ashley Yakeley > > > ______________________________**_________________ > Haskell mailing list > Haskell(a)haskell.org > http://www.haskell.org/**mailman/listinfo/haskell<http://www.haskell.org/mailman/listinfo/haskell> >

2 2

can't understand the Cabal manual. flag, manual
by Roman Beslik 24 Nov '12

24 Nov '12

Hello. http://www.haskell.org/cabal/users-guide/developing-packages.html#package-d… {{{ manual: boolean (default: False) By default, Cabal will first try to satisfy dependencies with the default flag value and then, if that is not possible, with the negated value. However, if the flag is manual, then the default value (which can be overridden by commandline flags) will be used. }}} Can I interpret this as {{{ First, Cabal tries to satisfy dependencies with the default flag value. Second, if the flag is manual, Cabal tries to satisfy dependencies with the negated flag value. }}} ? {{{ This file should contain of a number global property descriptions and several sections. }}} Maybe it's my English, but I can only think of "descriptions of global properties of a number" which does not make sense.

4 6

Adding swap/compose functions to Data.Tuple and Data.Function
by John Wiegley 23 Nov '12

23 Nov '12

Moving discussion from http://hackage.haskell.org/trac/ghc/ticket/7435: I propose adding the following three functions to Data.Tuple and Data.Function, respectively: swap :: (a,b) -> (b,a) swap = snd &&& fst Fairly obvious, but strangely missing. compose :: [(a -> a)] -> a -> a compose = foldr (.) id composeM :: [(a -> m a)] -> a -> m a composeM = foldr (<=<) return This can be done with Endo, of course, but the result is awkward. The idiom of composing a list of endomorphisms occurs often enough that you can Google for it. -- John Wiegley FP Complete Haskell tools, training and consulting http://fpcomplete.com johnw on #haskell/irc.freenode.net

10 11

Proposal: Allow gunfold for Data.Map, Data.IntMap, etc.
by Edward Kmett 22 Nov '12

22 Nov '12

I would like to propose improving the Data instances for a number of currently completely opaque data types in the containers package, by using virtual constructors. The instance for Data.Map already uses fromList for gfoldl, it just stops there. Extending it to be able to gunfold and mention the name of that constructor would enable generic traversal libraries like uniplate, etc. to work over the contents of the Map, rather than bailing out in fear or crashing at the sight of a mkNoRepType. An example of the changes for Data.Map are highlighted below. instance (Data k, Data a, Ord k) => Data (Map k a) where gfoldl f z m = z fromList `f` toList m toConstr _ = fromListConstr gunfold k z c = case constrIndex c of 1 -> k (z fromList) _ -> error "gunfold" dataTypeOf _ = mapDataType dataCast2 f = gcast2 f fromListConstr :: Constr fromListConstr = mkConstr mapDataType "fromList" [] Prefix mapDataType :: DataType mapDataType = mkDataType "Data.Map.Map" [fromListConstr] I've used this approach for years on my own libraries to great effect. Discussion Period: 3 Weeks (I added a week for ICFP) -Edward

5 10

Re: Proposal: Add (&) to Data.Function
by John Lato 21 Nov '12

21 Nov '12

> > From: David Menendez <dave(a)zednenem.com> > > On Tue, Nov 20, 2012 at 11:59 AM, Yitzchak Gale <gale(a)sefer.org> wrote: > > > It is a common idiom to write a sequence of composed combinators in > > reverse order to the way they would be written with ($) or (.). That > > naturally expresses the idea of the combinators as operations being > > applied in the given order. > > > > I think the bar should be pretty high before we add a third way to write > function application to base. How common is this? Common among whom? Why > can't the existing idioms of (.) and ($) work just as well? > I guardedly agree with Dave Menendez. There are a very limited number of convenient 1-character operators available, and I'm not certain that function application should take up 3 of them. On the other hand, this is a common enough idiom to have emerged in both diagrams and lens at least. (I'm not taking a position at this time) John L.

1 0

Proposal: Add setEnv/unsetEnv to "base"
by Simon Hengel 19 Nov '12

19 Nov '12

Hi, setting environment variables is possible on both POSIX and Windows systems. Currently this functionally is missing from System.Environment; the "unix" package provides an implementation for POSIX systems. I propose to add setEnv/unsetEnv to System.Environment, so that it is easier to write applications that behave the same on POSIX and Windows systems. I implemented both in a way, so that they behave exactly the same on POSIX and Windows (modulo case sensitivity on Windows). For now I prepared a standalone implementation [1] and corresponding tests [2]. That implementation depends on the "unix" package. If the proposal is accepted, I will prepare a patch against "base" that includes the relevant code from "unix". Discussion Period: 2 Weeks Cheers, Simon [1] https://github.com/sol/setenv/blob/master/src/System/SetEnv.hs [2] https://github.com/sol/setenv/blob/master/test/System/SetEnvSpec.hs

8 13

intersectionSet, differenceSet
by Henning Thielemann 16 Nov '12

16 Nov '12

When using the new Data.Map.fromSet function I encountered that I use it many times in connection with Map.intersection and Map.difference like so Map.intersection m (Map.fromSet (const ()) s) Map.difference m (Map.fromSet (const ()) s) Actually the element types of the second operands of Map.intersection and Map.difference are completely independent and the values of these maps are not touched. Thus I think that intersection and difference with a Set are the more natural operation. How about adding Map.intersectionSet :: Ord k => Map k a -> Set k -> Map k a Map.differenceSet :: Ord k => Map k a -> Set k -> Map k a to Data.Map?

3 3

Cabal: download dependencies for later offline build
by Bram Neijt 14 Nov '12

14 Nov '12

Dear reader, I'm experimenting with creating my own little build server using disk images as job input/output and KVM to run them[1]. As a security measure, the build process does not have any network access during the build, so I need to make sure everything is available before the build starts. What is the easiest way to "take a cabal build offline" as it were, so I can make sure a cabal build will work without a network connection? Greetings, Bram [1] https://github.com/bneijt/construir/

2 2

mtl-2.1 severly broken, cabal needs blacklisting
by Andreas Abel 14 Nov '12

14 Nov '12

After 2 days of shrinking 251 modules of source code to a few lines I realized that modify in MonadState causes <<loop>> in mtl-2.1. http://hackage.haskell.org/packages/archive/mtl/2.1/doc/html/src/Control-Mo… The bug has been fixed, apparently seven month ago. https://github.com/ekmett/mtl/pull/1 However, the "malicious" mtl-2.1 still lingers on: it is available from hackage and installed in many systems. This calls for a means of blacklisting broken or malicious packages. cabal update should also pull a blacklist of packages that will never be selected by cabal install (except maybe by explicit user safety overriding). I think such a mechanism is not only necessary for security purposes, but also to safe the valuable resources of our community. Cheers, Andreas -- Andreas Abel <>< Du bist der geliebte Mensch. Theoretical Computer Science, University of Munich Oettingenstr. 67, D-80538 Munich, GERMANY andreas.abel(a)ifi.lmu.de http://www2.tcs.ifi.lmu.de/~abel/

5 8