- Haskell-Cafe - Haskell.org

[ANN] extended-categories
by Ian Milligan 19 Aug '14

19 Aug '14

extended-categories <github.com/ian-mi/extended-categories> is an implementation of categories designed to make use of GHC's recently enriched kind system. This project is a work in progress (only a few basic constructs are implemented), but your comments are welcomed. extended-categories requires a recent version of GHC (I have only tested it with 7.8.3). Categories are poly-kinded and have an Object constraint family: class Category m where type Object (m :: k -> k -> *) (a :: k) :: Constraint id :: Object m a => m a a (.) :: (Object m a, Object m b, Object m c) => m b c -> m a b -> m a c This means in addition to our usual categories of kind * -> * -> * we can define categories of other kinds, for example (:-) from the constraints package forms a category with kind Constraint -> Constraint -> * (whose instance is used in this package)! Functors are named (tagged) as done in the data-category package. This allows a much wider variety of functors to be defined. Additionally functors must provide proof that they are closed under the object constraint, i.e. forall a. Object (Domain f c1) a :- Object (Codomain f c2) (FMap f a). For any two categories c1: k1 -> k1 -> * and c2: k2 -> k2 -> *, the product category (c1 :><: c2) has kind (k1, k2) -> (k1, k2) -> *. The functor (Diag c) is defined to be the functor from c to c :><: c which sends the object X to '(X, X) and morphism f to the morphism f :><: f. Categories which contain their products (the class ProductCategory) are defined by the universal property described here < http://en.wikipedia.org/wiki/Universal_morphism#Products>, i.e. by a terminal morphism, for every object a and b of c, from (Diag c) to the object '(a, b) of c :><: c. From this terminal morphism, the functions fst, snd, (&&&), and (***), as well as the product functor from c :><: c to c mapping '(a, b) to a >< b and f :><: g to f *** g, are defined. Here are some examples of how this works with the standard Haskell functors: *Functor> fmap (CanonicalF :.: CanonicalF) P.succ [P.Just 2, P.Nothing] [Just 3,Nothing] *Product> fmap (CanonicalF :.: ProductF) ((P.+1) :><: (P.+2)) [(1,2), (3,4), (5,6)] [(2,4),(4,6),(6,8)] or alternatively fmap CanonicalF ((P.+1) *** (P.+2)) [(1,2), (3,4), (5,6)] Next I will likely be adding more instances as well as some documentation. Thanks for reading, Ian Milligan

2 3

Reducing CPU load when using Helm
by Kaspar Emanuel 19 Aug '14

19 Aug '14

Forwarding this to haskell-cafe as it failed before. On 18 August 2014 20:24, Kaspar Emanuel <kaspar.bumke(a)gmail.com> wrote: > Hi, > > are there some simple tricks to reduce the CPU load? This simple > square rendering sits at 20% CPU even when the window size remains > constant. > > import FRP.Helm > import qualified FRP.Helm.Window as Window > > render :: (Int, Int) -> Element > render (w, h) = collage w h [move (100, 100) $ filled red $ square 64] > > main :: IO () > main = do > engine <- startup defaultConfig > run engine $ render <~ Window.dimensions engine > > The binary fractal tree I submitted in another thread on helm-dev > sits at 100% CPU when the mouse isn't moving. It seems to me that > Helm keeps re-drawing and re-computing when it doesn't need to. > > I started digging through the source a bit but it will take me a while > to understand Elerea and how it samples the inputs and causes things > to update. > > Are there any quick fixes? > > Ciao, > > Kaspar

1 0

Is there something for emacs to view project module structure?
by Magicloud Magiclouds 19 Aug '14

19 Aug '14

Hi, I think a structure viewing in editor could be very help for long file or big project. Something like java/c# IDEs do. But I could not find a hint on google. Any suggestions? -- 竹密岂妨流水过山高哪阻野云飞 And for G+, please use magiclouds#gmail.com.

3 3

Type-level numbers problem
by Konstantine Rybnikov 18 Aug '14

18 Aug '14

Hi! I wrote one thing with type-level numbers that were represented as simple data Zero data Succ a Then, in order to not invent my own peano numbers, I decided to go look for ones, and found GHC.TypeLits. But unfortunately, it shows a weird error for me. Here is a piece of code: https://gist.github.com/k-bx/a7b49165399658f87358 The one that's below "main" is working properly, while the one on top of it gives this error: typenats.hs:17:16: No instance for (Op D (0 + 1)) arising from a use of `f' Possible fix: add an instance declaration for (Op D (0 + 1)) In the expression: f (VJ i :: VJ D (0 + 1)) In an equation for `f': f (VJ i) = f (VJ i :: VJ D (0 + 1)) In the instance declaration for `Op D 0' Maybe I worked too much on type-level numbers today to not understand this simple error I'm getting, but could someone be so kind and explain me what am I doing wrong? Thank you very much!

2 2

tree insert and laziness
by Semen Trygubenko / Семен Тригубен ко 18 Aug '14

18 Aug '14

Dear Haskell-Cafe, I'm presently reviewing a translation of the following paragraph (taken from LUaH, chapter 8): http://learnyouahaskell.com/making-our-own-types-and-typeclasses "In languages like C, we'd do this by modifying the pointers and values inside the tree. In Haskell, we can't really modify our tree, so we have to make a new sub-tree each time we decide to go left or right and in the end the insertion function returns a completely new tree, because Haskell doesn't really have a concept of pointer, just values. Hence, the type for our insertion function is going to be something like a -> Tree a - > Tree a. It takes an element and a tree and returns a new tree that has that element inside. This might seem like it's inefficient but laziness takes care of that problem." If it is really the laziness that makes this approach efficient? If it isn't, what did the author want to say in the last sentence? Thank you, Semen -- Семен Тригубенко http://trygub.com

4 4

Haskell SDL2 FRP games working on Android
by Ivan Perez 18 Aug '14

18 Aug '14

Hi Café Answering recent questions posted on this list regarding FRP, I would like to announce that we have a working Android game written in Haskell, using SDL2 and Yampa. A short article describing the implications that this could have, together the libraries and backend we used, is here: http://keera.co.uk/blog/2014/08/13/most-inspiring-green-screen-you-will-eve… Changes made to existing libraries will be published as we move towards releasing the game on Google Play, and so will examples of Android programs written in Haskell. All the best Ivan -- Facebook: https://www.facebook.com/keerastudios Twitter: https://twitter.com/KeeraStudios Web: http://www.keera.co.uk

2 1

Fully funded PhD position on Computational Logic/Functional Programming/Interactive Theorem Proving, University of Dundee, Scotland
by Ekaterina Komendantskaya 18 Aug '14

18 Aug '14

One fully funded PhD position on Computational Logic/Functional Programming/Interactive Theorem Proving is available at the University of Dundee, Scotland. For more information, please email katya(a)computing.dundee.ac.uk or refer to: https://docs.google.com/document/d/10Cnlws_XPKrimo9m_CKPSXUZFCkP8FhBfD9a9xA… The funding is provided by EPSRC Doctoral Training Partnership Scheme, and is available for UK residents only. Katya Ekaterina Komendantskaya Senior Lecturer, Head of PhD Studies Room 1.04, Queen Mother Building School of Computing, University of Dundee Scotland, DD14HN Tel: (+44) 01382384820

1 0

Wish list for GHC API tooling support
by Alan & Kim Zimmerman 18 Aug '14

18 Aug '14

At the moment the GHC API is a sort of poor relation in the haskell world, where it could be a significantly useful resource for the growing list of haskell tool providers. Based on my experiences with HaRe, I have started putting together a wish list of features I would like to see in it, which is here https://github.com/fpco/haskell-ide/wiki/GHC-API I welcome feedback / discussion on this. Regards Alan

3 2

Haskell developer roles at Standard Chartered
by Don Stewart 17 Aug '14

17 Aug '14

Hi Cafe, The Strats team at Standard Chartered is hiring expert typed FP developers for Haskell dev roles in London. This is a "front office" finance role - meaning you will work directly with traders building software to automate and improve their efficiency. The role is very development focused, and you will use Haskell for almost all tasks: data analysis, DSLs, market data publishing, databases, web services, desktop GUIs, large parallel tasks, quantitative models, solvers, everything. There may be a small amount of C++ or F# on occasion. This is a fast paced role - code you write today will be deployed within hours to hundreds of users and has to work. You will join an expert team in London, and significant, demonstrable experience in typed FP (Haskell, OCaml, F# etc) is strongly preferred. We have more than 2 million lines of Haskell, and our own compiler. In this context we look for skill and taste in typed API design to capture and abstract over complex, messy systems. Experience writing typed APIs to external systems such as databases, web services, pub/sub platforms is very desirable. We like working code, so if you have Hackage or github libraries, we want to see them. We also like StackOverflow answers, blog posts, academic papers, or other arenas where you can show broad FP ability. The role requires physical presence on the trading floor in London. Remote work isn't an option. Ideally you have some project and client management skills -- you will talk to users, understand their problems, and then implement and deliver what they really need. No financial background is required. If this sounds exciting to you, please send CVs to dons00 at gmail.com

3 3

Issues with Marshalling arrays of user defined types
by lonetiger＠gmail.com 17 Aug '14

17 Aug '14

Hi All, I’ve been scratching my head for weeks trying to figure out what goes wrong with the code pasted below. It’s compiled to a shared library (dll) on windows. I know the Storable instance is correct because calls to testFooA works, but calls to testFooListA get an invalid pointer back. Lists of Ints do work fine, so it seems that newArray is also working. In GHCi it can dereference the pointer. And this used to work before in earlier GHC 7.0 and later 6.6 builds. Using WinDBG and gflags I see that the pointer is indeed invalid, but that the *valid* data is there in the heap. The example code can be found at http://lpaste.net/6148884045213204480 Anyone have an idea what I’m doing wrong? Kind Regards, Tamar

1 0