October 2012 - Haskell-Cafe

arr considered harmful
by Ryan Ingram 21 Dec '12

21 Dec '12

I know it's a bit of an 'intentionally provocative' title, but with the recent discussions on Arrows I thought it timely to bring this up. Most of the conversion from arrow syntax into arrows uses 'arr' to move components around. However, arr is totally opaque to the arrow itself, and prevents describing some very useful objects as arrows. For example, I would love to be able to use the arrow syntax to define objects of this type: data Circuit a b where Const :: Bool -> Circuit () Bool Wire :: Circuit a a Delay :: Circuit a a And :: Circuit (Bool,Bool) Bool Or :: Circuit (Bool,Bool) Bool Not :: Circuit Bool Bool Then :: Circuit a b -> Circuit b c -> Circuit a c Pair :: Circuit a c -> Circuit b d -> Circuit (a,b) (c,d) First :: Circuit a b -> Circuit (a,c) (b,c) Swap :: Circuit (a,b) (b,a) AssocL :: Circuit ((a,b),c) (a,(b,c)) AssocR :: Circuit (a,(b,c)) ((a,b),c) Loop :: Circuit (a,b) (a,c) -> Circuit b c etc. Then we can have code that examines this concrete data representation, converts it to VHDL, optimizes it, etc. However, due to the presence of the opaque 'arr', there's no way to make this type an arrow without adding an 'escape hatch' Arr :: (a -> b) -> Circuit a b which breaks the abstraction: circuit is supposed to represent an actual boolean circuit; (Arr not) is not a valid circuit because we've lost the information about the existence of a 'Not' gate. The arrow syntax translation uses arr to do plumbing of variables. I think a promising project would be to figure out exactly what plumbing is needed, and add those functions to a sort of 'PrimitiveArrow' class. All of these plumbing functions are trivially implemented in terms of 'arr', when it exists, but if it doesn't, it should be possible to use the arrow syntax regardless. -- ryan

9 13

Why Kleisli composition is not in the Monad signature?
by damodar kulkarni 30 Nov '12

30 Nov '12

Hi, The Monad class makes us define bind (>>=) and unit (return) for our monads. Why the Kleisli composition (>=>) or (<=<) is not made a part of Monad class instead of bind (>>=)? Is there any historical reason behind this? The bind (>>=) is not as elegant as (>=>), at least as I find it. Am I missing something? - Damodar

17 32

Re: [Haskell-cafe] Executables got bigger using dynamic link.
by Magicloud Magiclouds 26 Nov '12

26 Nov '12

Sorry, I left the profiling option on, which seems to suppress the dynamic options. On Mon, Oct 29, 2012 at 1:29 PM, Magicloud Magiclouds <magicloud.magiclouds(a)gmail.com> wrote: > Hi, > I set "shared: True" in ~/.cabal/config, and using ghc 7.6.1. Then > clear user space hackages and reinstall them. > First of all, I think comparing to static link, dynamic linked > executables file should be smaller. And the libraries (.so) could > benefits from caching to save memory usage. > > Well, I installed cabal-install and alex. I got this: > cabal 9.9M -> 24M > alex 2.7M -> 5.1M > Is this correct? > -- > 竹密岂妨流水过 > 山高哪阻野云飞 > > And for G+, please use magiclouds#gmail.com. -- 竹密岂妨流水过山高哪阻野云飞 And for G+, please use magiclouds#gmail.com.

2 1

JavaScript (SpiderMonkey, V8, etc) embedded in GHC?
by Bob Hutchison 13 Nov '12

13 Nov '12

Hi, I've looked around with no success… this surprises me actually. Has anyone embedded SpiderMonkey, V8, or any other relatively decent JavaScript interpreters in GHC (using the FFI)? I did find http://justinethier.github.com/husk-scheme/ which is a scheme R5RS implementation (I could make this work). There's also some work done embedding Lua. I also found a number of packages that compile javascript to Haskell, or the other way around, but I don't need that kind of thing. All I really need is to allow users to write some JavaScript that accepts a single JSON 'file/string' from my Haskell program and produces another JSON 'file/string' that my Haskell program will accept. Thanks, Bob

4 7

Call for discussion: OverloadedLists extension
by Michael Snoyman 06 Nov '12

06 Nov '12

(Prettier formatting available at: https://gist.github.com/3761252) Many of us use the OverloadedStrings language extension on a regular basis. It provides the ability to keep the ease-of-use of string literal syntax, while getting the performance and correctness advantages of specialized datatypes like ByteString and Text. I think we can get the same kind of benefit by allowing another literal syntax to be overloaded, namely lists. ## Overly simple approach The simplest example I can think of is allowing easier usage of Vector: [1, 2, 3] :: Vector Int In order to allow this, we could use a typeclass approach similar to how OverloadedStrings works: class IsList a where fromList :: [b] -> a b instance IsList Vector where fromList = V.fromList foo :: Vector Int foo = fromList [1, 2, 3] ## Flaws However, such a proposal does not allow for constraints, e.g.: instance IsList Set where fromList = Set.fromList No instance for (Ord b) arising from a use of `Set.fromList' In the expression: Set.fromList In an equation for `fromList': fromList = Set.fromList In the instance declaration for `IsList Set' Additionally, it provides for no means of creating instances for datatypes like Map, where the contained value is not identical to the value contained in the original list. In other words, what I'd like to see is: [("foo", 1), ("bar", 2)] :: Map Text Int ## A little better: MPTC A simplistic approach to solve this would be to just use MultiParamTypeClasses: class IsList input output where fromList :: [input] -> output instance IsList a (Vector a) where fromList = V.fromList foo :: Vector Int foo = fromList [1, 2, 3] Unfortunately, this will fail due to too much polymorphism: No instance for (IsList input0 (Vector Int)) arising from a use of `fromList' Possible fix: add an instance declaration for (IsList input0 (Vector Int)) In the expression: fromList [1, 2, 3] In an equation for `foo': foo = fromList [1, 2, 3] This can be worked around by giving an explicit type signature on the numbers in the list, but that's not a robust solution. In order to solve this properly, I think we need either functional dependencies or type families: ## Functional dependencies class IsList input output | output -> input where fromList :: [input] -> output instance IsList a (Vector a) where fromList = V.fromList instance Ord a => IsList a (Set a) where fromList = Set.fromList instance Ord k => IsList (k, v) (Map k v) where fromList = Map.fromList foo :: Vector Int foo = fromList [1, 2, 3] bar :: Set Int bar = fromList [1, 2, 3] baz :: Map String Int baz = fromList [("foo", 1), ("bar", 2)] ## Type families class IsList a where type IsListInput a fromList :: [IsListInput a] -> a instance IsList (Vector a) where type IsListInput (Vector a) = a fromList = V.fromList instance Ord a => IsList (Set a) where type IsListInput (Set a) = a fromList = Set.fromList instance Ord k => IsList (Map k v) where type IsListInput (Map k v) = (k, v) fromList = Map.fromList foo :: Vector Int foo = fromList [1, 2, 3] bar :: Set Int bar = fromList [1, 2, 3] baz :: Map String Int baz = fromList [("foo", 1), ("bar", 2)] ## Conclusion Consider most of this proposal to be a strawman: names and techniques are completely up to debate. I'm fairly certain that our only two choices to implement this extension is a useful way is fundeps and type families, but perhaps there's another approach I'm missing. I don't have any particular recommendation here, except to say that fundeps is likely more well supported by other compilers.

10 19

Help: Main: thread blocked in MVar operation
by "José A. Lopes" 04 Nov '12

04 Nov '12

Hey everyone, I changed my code I now I get the following error message Main: thread blocked indefinitely in an MVar operation Before the change, I was using the State monad with runState. Then, I changed the code to use the StateT monad transformer wrapped around IO monad and runStateT. And this change introduced the error message. BTW I am using DoRec extension, maybe it is the source of the problem, but I don't know. Here are some specs: ghc 7.4.2 mtl 2.1.1 If you need any other information please don't hesitate to ask. Any help is greatly appreciated! Cheers, José -- José António Branquinho de Oliveira Lopes Instituto Superior Técnico Technical University of Lisbon

2 2

[Security] Put haskell.org on https
by Niklas Hambüchen 02 Nov '12

02 Nov '12

(I have mentioned this several times on #haskell, but nothing has happened so far.) Are you aware that all haskell.org websites (hackage, HaskellWiki, ghc trac) allow unencrypted http connections only? This means that everyone in the same Wifi can potentially - read you passwords for all of these services - abuse your hackage account and override arbitrary packages (especially since hackage allows everybody to override everything) I propose we get an SSL certificate for haskell.org. I also offer to donate that SSL certificate (or directly create it using my Startcom account). Niklas

17 34

Haskell User Group Hamburg (Germany) - Meeting Thursday November 8th 2012
by Stefan Schmidt 31 Oct '12

31 Oct '12

Hello cafe, there's an upcomming meeting of the Haskell user group Hamburg next week on Thursday, November 8th. We'll start 19:00 at the fortytools GmbH in Hamburg (Georgsplatz 10, 20099 Hamburg). We don't have a program or talk yet. Maybe we'll discuss our experiences from the last ICFP and the Haskell Symposium. If you are interested to join us, have a look at our doodle (only in german) http://www.doodle.com/gvdsbsyfz7a7427g Be careful: the Google Maps link in the Doodle may direct you to the old address. Georgsplatz 10, 20099 Hamburg is the new one. Further announcements will be made via Twitter, so you may want to follow https://twitter.com/hug_hh.

1 0

ANNOUNCE: btree-concurrent
by Johan Brinch 31 Oct '12

31 Oct '12

Greetings Haskellers, We are pleased to announce this new package, featuring a concurrent b-tree developed as part of our master's thesis on backup systems. The package features code related to STM, caching, QuickCheck and of course concurrency and parallelism. And it's also on Hackage: http://hackage.haskell.org/package/btree-concurrent In case the build fails on Hackage and the documentation doesn't work, it's also available here: http://hindsight.dk/doc/html/btree-concurrent Credits to Thomas Conway of Melbourne University, for discussions related to the implementation, and to our supervisor Ken Friis Larsen of University of Copenhagen. And now, the description of the package: ===== A backend agnostic, concurrent BTree with relaxed balance[1] written in Haskell using a mix of IO operations and pure STM. Although the code does work, it is neither production-ready nor complete. Features include: * Caching: While nodes are periodically saved on persistent storage (e.g. disk) they are cached in-memory during operations. * Live flushing: It is possible to save the current version of the tree to disk and run modifying operations at the same time. The nodes are updated buttom-up to ensure a valid tree in the case of a crash. * Backend agnosticism: A simple API is used as an abstraction for the persistent storage. * Verification: The test-suite uses QuickCheck to compare the trees behaviour to that of Data.Map. Deficients include: * Memory greedy. Nodes are not stored in a compact fashion and are constantly being serialised/deserialised. * findMin and findMax are incomplete and may fail (see the TODO for details). * The implementation does not parallelise well. [1] B-trees with relaxed balance, K. S. Larsen & R. Fagerberg, Parallel Processing Symposium, 1995. Proceedings., 9th International -- Johan Brinch & Morten Brøns-Pedersen

2 2

ANNOUNCE: plat
by Miguel Mitrofanov 31 Oct '12

31 Oct '12

My current project at work has a web interface. Therefore I needed an HTML templating library. I've tried several and found them all unsatisfactory. Most of them generate Haskell code from templates (with Template Haskell or a separate utility), and I don't like metaprogramming; and recompiling the program every time you make a small change in the template kinda defeats the purpose. Other libraries make it PITA to implement basic things, like iterating over an array. So, I'm proud to announce the first release of another template library: plat-0.1.0.1 (http://hackage.haskell.org/package/plat) An example of what it does. If your data looks like do "name" =: "Mac's tools" "staff" =: [ do "index" =: "1" "name" =: "Alice" , do "name" =: "Bob" "index" =: "2" "bad" =: True , do "index" =: "3" "name" =: "Nicolás" ] and your template is List of employees at @name: @staff[person @person.index@.. @person.name@{@!person.bad (going to be fired)@|@}@ @]@# There won't be an empty line in the result Then the result you get would be this: List of employees at Mac's tools: 1. Alice 2. Bob (going to be fired) 3. Nicolás Keep in mind, that it's actually a string template library, and the resulting string doesn't have to be valid HTML, or any HTML at all. I haven't benchmarked this library - since I intended to use it for web interface in an program which won't be released outside, speed was not in my priorities; but I'll work on that in future versions. Also, templates are supposed to be in UTF-8, as I've successfully forgot all about other encodings. So, if you want binary templates, I'm afraid, this library is not for you.

1 0