- Haskell-Cafe - Haskell.org

Thoughts about redesigning "Num" type class
by DreamApart AtHaskells 15 Sep '15

15 Sep '15

The class Num defines the method functions like (+), (-), (*), negate, ... But they are actually different concepts. We often meet situations that we want an addition for some type, but not a multiplication. For example: *data* Vec3 = Vec3 Float Float Float The additive (and subtractive) operation is obvious, but not for a multiplication. * Note* (*) :: a -> a -> a -- not avaliable for a Vec3 scale :: Float -> Vec3 -> Vec3 dot :: Vec3 -> Vec3 -> Float We cannot define a (+) alone for a type, so we got many different functions from different libs, like mappend, mplus, plus, (<+>), (.+.), (+.), which all reads "plus". My opinion is that the methods of Num should be seperated into different type classes, so we don't need to invent a new symbol for a new lib. A feasible design is: *import* Prelude *hiding* ( Num (..), Monoid (..), sum ) *class* SemiGroup a *where* (+) :: a -> a -> a *class* SemiGroup a => Monoid a *where* zero :: a *class* Monoid a => Group a *where* negate :: a -> a (-) :: a -> a -> a x - y = x + negate y negate x = zero - x *class* Group a => Ring a *where *-- not sure about the name ... (*) :: a -> a -> a *class* Ring a => Num a *where* abs :: a -> a signum :: a -> a fromInteger :: Integer -> a sum :: (Foldable t, Monoid a) => t a -> a sum = foldl' (+) zero -- for compatibility mempty = zero mappend = (+) mconcat = sum

11 21

Strange problem with inference
by Nikolay Amiantov 15 Sep '15

15 Sep '15

Hi Cafe, I've been playing around with lens and stumbled upon strange GHC behaviour. Consider this source (using lens package and GHC 7.10.2): {-# LANGUAGE TypeFamilies #-} import Control.Lens class Test a where type TestT a myiso :: Iso' a (TestT a) test1 :: Test a => a -> TestT a test1 = view myiso test2 :: Test a => TestT a -> a test2 = view (from myiso) GHC would emit this error: /tmp/test.hs:13:9: Could not deduce (Control.Monad.Reader.Class.MonadReader (TestT a) ((->) (TestT a))) arising from a use of ‘view’ from the context (Test a) bound by the type signature for test2 :: Test a => TestT a -> a at /tmp/test.hs:12:10-31 In the expression: view (from myiso) In an equation for ‘test2’: test2 = view (from myiso) Failed, modules loaded: none. However, `MonadReader r ((->) r)` is defined for any and all `r`! Furthermore, `test1` has no problem with this and `view` there uses this instance too. The only difference that I see is the presence of a type family: * `test1` needs `MonadReader a ((->) a)` * `test2` needs `MonadReader (TestT a) ((->) (TestT a))` , but I don't understand how can this result in a different behavior. Notice that this likely may be reproduced somehow without lens -- I've spent some time trying to minify this example further but alas to no avail. Thanks in advance! -- Nikolay.

4 5

Darcs Hacking Sprint #9 (September 18th-20th, Paris)
by Guillaume Hoffmann 15 Sep '15

15 Sep '15

Dear Hackers the next Darcs Sprint that will be in Paris, on September 18-20 at IRILL (near Place d'Italie). Please check out the details at: http://darcs.net/Sprints/2015-09 Here are three things to know 1. Everybody is welcome to join us. We'd love to have you, whatever your Haskell or Darcs hacking experience. Also, if you've got a wacky idea for the future of version control, or a cool use for the Darcs library, you should join us too :-) 2. Please let us know if you're attending. Just add your name to http://wiki.darcs.net/Sprints/2015-09 and you're good to go. You can also send us an email. 3. We can reimburse travel costs (within reason!). Let us know if you'd like a reimbursement, and save your receipts. Many thanks to everybody who participated in our fundraising drives or who gave money on the side. Thanks also to the Software Freedom Conservancy for making fundraising and reimbursements so painless! If you can't join us in person, but you'd like to cheer us on, say hello at http://darcs.net/Donations ! see you in one month! Guillaume

1 1

Powerset of a set
by JORGE MALDONADO 15 Sep '15

15 Sep '15

The powerset of set s is a set containing all subsets of s. I need a clue on how to write Haskell code to get the superset of a set using direct recursion and list comprehension. Best regads.

7 9

the library of beautiful instruments implemented in haskell / csound
by Anton Kholomiov 14 Sep '15

14 Sep '15

Status update for my haskell synth csound-expression. The main point is presence of many cool instruments. They are implemented in the package csound-catalog. All packages are compiled with GHC-7.10 So the hackage fails to build them and unfortunately docs a broken too. But you can look at the source code of the module Csound.Patch to now the names of the instruments. The usage is pretty straightforward. It's described here: https://github.com/spell-music/csound-expression/blob/master/tutorial/chapt… There is an mp3 file to listen to the instruments. http://ge.tt/1jNETqN2/v/0 *The 4.8.3 is out! New features:* This is a very important release to me. It tries to solve the problem present in the most open source music-production libraries. It's often the pack of beautiful sounds/timbres is missing. User is presented with many audio primitives but no timbres are present to show the real power of the framework. This release solves this problem. See the friend package csound-catalog on Hackage. It defines 200+ beautiful instruments ready to be used. The csound-expression defines a new type called Patch for description of an instrument with a chain of effects. It's good place to start the journey to the world of music production. There are new functions for synchronized reaction on events. The triggering of events can be synchronized with given BPM. The library is updated for GHC-7.10! github repo: https://github.com/spell-music/csound-expression hackage: http://hackage.haskell.org/package/csound-expression Cheers!

3 5

The semiring of Scott reals (Thoughts about redesigning "Num" type class)
by Olaf Klinke 14 Sep '15

14 Sep '15

This departs somewhat from the thread's topic. Below I detail a semiring R that is not a ring because negation is not definable in Haskell. I further sketch integration in Haskell which I ripped off papers by Gordon Plotkin, Steve Vickers, Anders Kock and others. Infinity comes in as a limit, of course. Approximate a real number by an ascending stream of rational numbers. (The ascending can not be enforced by Haskell's type system, of course.). What you have is a so-called lower real or Scott real. Still addition and multiplication is continuous on such reals and there is a unique way of extending these operations to infinity, represented by any strictly ascending stream. Neither subtraction not division are computable for the Scott reals. Indeed, to know a lower bound of x-y one needs to know upper bounds of y. Denote this type by R: type R = Stream Rational Further consider ascending streams of Booleans, where any two streams containing True are considered semantically equivalent. type S = Stream Bool Consider the following types: type ScottOpen x = x -> S type V x = (ScottOpen x) -> R Certain elements of the type V x are called valuations, assigning to any Scott open subset[*] of x a lower real number. The type constructor V has a monad instance. If there was a one-element type in Haskell (say 1), then V 1 is isomorphic to R. Hence a lower semicontinuous real-valued function f :: x -> R is the same as a function f :: x -> V 1. Now observe that (=<<) :: (x -> V 1) -> V x -> V 1 :: (x -> R ) -> V x -> R has the type of an integral operator. In other words, the bind of the monad V implements integration. I have an implementation that lets you do integration over any Haskell type, but it is very slow. An analogue of this is the monad []. Indeed, total elements of type N = [()] are natural numbers. Disregarding permutations, an element of [x] assigns to finitely many elements of x a weight from the natural numbers. Then (=<<) :: (x -> [()]) -> [x] -> [()] :: (x -> N ) -> [x] -> N is the integral operator integrating natural-number-valued functions against a finite natural-number-valued measure. Further observe that liftM2 (flip const) :: Monad m => m () -> m x -> m x is a monoid action that implements multiplication on N. It distributes over (++) for categorical reasons: (++) arises from the isomorphism (up to permutation) ([a],[b]) -> [Either a b]. As an exercise, define a monad instance for type FinDist x = [(x,Rational)] Hint: FinDist x = WriterT (Product Rational) [] x and observe that FinDist () is isomorphic to Rational (up to some equivalence relation). Voila: You have integration of Rational-valued functions against finitely supported distributions. For many combinatorial applications, this is sufficient. [*] For u :: x -> S, consider the set of elements of x for which u produces a stream containing True. This is the Scott open subset. Valuations are required to send the empty open to the constant zero stream and further obey the modular law you know from measures: v(u) + v(u') = v(union u u') + v(intersection u u') where union = liftA2 max and intersection = liftA2 min. -- Olaf

2 1

ANNOUNCE: ioctl 0.0.1
by Maciej Piechotka 13 Sep '15

13 Sep '15

A package for type-safe I/O control. Currently only ioctl is supported. Currently simply a extract from my tuntap fork TODO: - Return the integer as well as structure (will break the API) - Port for Windows Network.Socket.IOCtl (as soon as I manage to setup some sane environment on this platform) - Wrapping around DeviceIoControl Example (in hsc): data NotRead = NotRead instance NotRead Int where ioctlReq _ = #const FIONREAD notRead s = ioctlsocket' s NotRead Regards

2 1

Re: [Haskell-cafe] haskell est il compatible avec mon projet graphique ?
by Daniel Díaz 13 Sep '15

13 Sep '15

Pardonnez mon mauvais français. Il existe des liaisons pour SDL2: http://hackage.haskell.org/package/sdl2 Aussi jeter un oeil à "diagrams" http://projects.haskell.org/diagrams/gallery.html On Saturday, September 12, 2015 at 6:17:04 AM UTC+2, Vincent Douce LR wrote: > > salut à tous > je dois réaliser des créations graphiques, avec deux destinations : des > animations sur l'écran, et des images fixes destinées à être publiées sur > papier > ces créations reposent sur des algorithmes de mon crû > je souhaite me créer une bibliothèque de séquences de code que je pourrai > étoffer au fur et à mesure et utiliser dans mes programmes > je cherche quel langage adopter, et quel pendant graphique lui adjoindre > j'envisage des pistes comme C+sdl ou python +? > utiliser haskell serait un défi parce que c'est moins intuitif pour moi un > langage fonctionnel et aussi tout simplemetn parce que je n'y connais rien > en haskell, je sais juste un peu de Scheme > mais peut être qu'il y a dans l'univers haskell un moyen de générer des > graphiques de qualité ? je susi ouvert > qu'en pensez-vous ? >

2 1

Re: [Haskell-cafe] Thoughts about redesigning "Num" type class
by Olaf Klinke 13 Sep '15

13 Sep '15

I sencond the statement that semirings might be a useful typeclass to have. Another example: Non-negative reals with infinity, the values that a measure can take. Nevertheless the biggest issue is what _symbol_ to use for the operations. We know that a ring is a semigroup in two ways, and no matter what semigroup structure we think the ring to "originate" from, we're going to annoy the other half of users who thought it the other way around. Some semigroups are thought of as (+), others as (*). Picking one of these symbols for all semigroups will be counter-intuitive for some structures. As to affine structures, Martin Escardo and Alex Simpson have a paper showing that the unit interval is the free cancellative midpoint algebra over two generators. They have Haskell code for this [1]. And yes, a typeclass is not a mathematical category, but Haskell is a language that lets you formalise quite a bit of category theory, and this is a cool thing to have. You can read a paper based on category theory and obtain executable code with very little effort. There are now entire Ph.D. theses written in Agda. To me it seems that the solution to the Num typeclass problem would be to have sub-typeclass statements that let you introduce aliases for the operations, as so: class Semigroup a where (<>) :: a -> a -> a class Group a where (*) :: a -> a -> a unit :: a inverse :: a -> a superclass Semigroup Group where (<>) = (*) But wait, we do have a mechanism for aliases: It's called newtype! Conclusion: If you want to express that any Num instance is a Group, do so with a newtype. Data.Monoid is a good example for this. -- Olaf [1] http://www.cs.bham.ac.uk/~mhe/.talks/map2011/martin_map.pdf

4 3

Re: [Haskell-cafe] lazily updating dependencies, git submodules and cabal
by timmy tofu 11 Sep '15

11 Sep '15

> > Using the 'cabal sandbox add-source' solution now requires that I > "add-source" the MarketModel package too. Otherwise, any change I make > to this package will not be tracked. > then… > Cabal sets the paths for "import"ed > modules in the "hs-source-dirs:" and that is global to the whole > package. > I can't tell if you do or do not want ProfitCalculator and Persistence to track changes made in MarketModel. If you do, and both should be in lock-step with it (unlike your situation with Tax), then you don't have a problem. If not, then the problem is Trader would depend on ProfitCalculator which might depend on a different version of MarketModel than Persistence (upon which Trader also depends). That is a classic dependency problem and not a function of submodules or add-sourcing, per se. The Trader module seems to be at the top of this heap - why would you want the Trader to (indirectly) use one version of MarketModel through ProfitCalculator, and another through Persistence?

3 2