June 2006 - Haskell-Cafe

Inferring types from functional dependencies
by Jeff.Harper＠handheld.com 16 Jun '06

16 Jun '06

Thank you. I followed your suggestion and implemented my default Divide as you've shown below. oleg(a)pobox.com wrote: > If we assume that we need Reciprocate only if we are going to use the > 'default' method, the solution becomes obvious. It does involve > overlapping and undecidable instances, sorry. These extensions are > really useful in practice. Here's the solution: > > > class Divide a b c | a b -> c where > > (/) :: a -> b -> c > > > Here's the most general instance. It applies when nothing more > specific does. It is in this case that we insist on being able to > take the reciprocal: > > > instance (Reciprocate b recip, Multiply a recip c) => > > Divide a b c where > > (/) x y = x * (recip y)

1 0

Rotating matrices
by Thomas Sutton 16 Jun '06

16 Jun '06

Dear list, I'm currently engaged in an attempt to wrap my head around type arithmetic. To make sure that I've understood, I plan to write a few operations on matrices a la Oleg's Number Parameterised Types. Before I get down to it, I've been making sure I know how to implement the operations (so that all I need to think about later is the type arithmetic). Today I've been looking at rotating matrices, i.e: taking a column- wise matrix and making it row-wise and, in the process, swapping the dimensions (thus a 3*2 matrix becomes a 2*3 matrix). I've been working with lists of lists for simplicity and have come up with: > rotate' :: [[a]] -> [[a]] > rotate' [] = [] > rotate' xs = (map (head) xs ):(rotate' $ filter (not . null) $ map (tail) xs) which seems to work just fine. While this solution is adequate (it seems to work for infinite structures as well, which is good), I originally set out to solve this problem with a fold or a mapAccum of some sort (I don't really care if it doesn't work with infinite structures). Can anyone suggest a way to write this with a fold (or tell me why it isn't a good idea)? Cheers, Thomas Sutton

4 3

what do you think of haskell ? (yes, it's a bit general ...:)
by minh thu 16 Jun '06

16 Jun '06

hi all folks, i'm diving into haskell for more than one year now. the reason for that is just that i like haskell. (i'm a computer science student) but i consider to move back to c/c++. here are my thoughts. i've no specific question but i'd like to have your opinion. here we go: haskell is really nice : short phrased, declarative, math-feeling, the type system ensure low bug amount (i surely forget other nice things). but haskell is quite ugly in some way : * array : if i want to write something involving array, i could use list, and a lot (too much!) of array types (io/st, mutable/immutable, c-friendly (storable). worst, code involving one type can need to be rewritten for another type. * laziness / array (again) (i am biased since i have no knowledge of haskell profiling / performance seeking) always with array code, i was forced (maybe it's because i dont know enough) to use iouarray : if not, performance|memory consumption were low|high. but with iouarray, i cant use an array of MyType; i have to use an array of (say) float (it's ok if i have only float in MyType). that kind of thing is what i think is *really* low-level. in c, i can have an array of what-i-want. * randomIO side-effect is nicely resolved with monad. and you have to thread your state. if you're writing your monad or use a transformer, things are quite explicitly (even if it's implicit in the do notation) threaded. but the threading of the randomIO argument is really not explicit for me : it just means that the underlying/threaded state in the IO monad can encapsulated a lot of things. it feels exactly like a c function. but usually, you cant design a c-like function in haskell (i.e. a function with state). * more things i dont remember... * generally my general feeling for haskell vs c is: in haskell i always have to learn new things to get my work done ; although haskell is really easy to learn in the first step, it's becoming increasingly hard to get what's the *trick* to do what i want. e.g. writing myfunction x1 .. xn | x1 `seq` ... False = undefined is not declarative (and i still have to learn to identify where it helps and where it doesn't) the c language take some more time to learn at the beginning but that's it! what you can learn by after is better c programming. there are memory management, pointers, side-effects, but you can do what you want, it behaves as expected and you dont have to learn (in a academic paper) how to use an array or how to do io. :) please don't tell me "you're stupid, go back to your c and leave this glorious mailing list" did you had the same feeling ? does it disappear ? how ? thanks a lot, vo minh thu

9 18

Everything but the lazyness - idea for force/delay lists
by Brian Hulley 16 Jun '06

16 Jun '06

Hi - I've been thinking about how to get an extremely fast language with all the benefits of Haskell ie completely pure with no side effects, but with monads, higher order functions, type classes etc, but without the lazyness. I know this is controversial, but having started to write a serious program in Haskell, I find that almost all of it doesn't need any lazyness at all, yet I have to constantly mess up my code by using Tuple2 a b instead of (a,b), ! in all my data types, and millions of brackets to use $! (which has the wrong associativity for passing multiple arguments strictly) etc and I can't do anything about the slowness and possible space leaks introduced by library functions which are lazy, not to mention the fact that afaiu the lifted type system necessary for any language which supports lazyness, and general undecidability results means that it will probably be impossible to ever compile lazy code to be as fast as OCaml etc. (This is not to say Haskell is too slow - I find the app I'm writing at the moment runs fast enough (given all my strictness annotations) it's just that lazyness is making my life more difficult rather than easier and is probably also costing something in terms of speed.) The one place where I'm using lazyness is where I need to glue the output of one computation to another by using a lazy list. In effect I'm using a lazy list as the equivalent of an iterator in C++ - the elements are only created when needed, but once read, I only read them once, so I don't need the memoization properties of lazyness. When I first started to learn Haskell, I thought lazyness was essential for monads and hence an acceptable price to pay for using them. However I now think that monads would work perfectly without lazyness, since they are usually always defined in terms of >>= (not >> as I'd thought when I didn't know Haskell). Other uses of lazyness are of course infinite structures and fixpoints etc. I think the use of a lazy list as a read-once-on-demand stream could be achieved just as easily by a strict language with some syntactic sugar by redefining the meaning of []: data GlueList a = Empty | Cons a [a] type [a] = () -> GlueList a force :: (() -> a) -> a force f = f () Pattern matching could then be changed so that p [a] = exp is short for: p x = case force x of Cons a y -> case force y of Empty -> exp and p [h|t] = exp -- Prolog style list sugar p x = case force x of Cons h t -> exp and in an expression, [exp1,...] would be expanded into the appropriate delayed construction eg: x = [a] x = \() -> Cons a (\() -> Empty) (List comprehensions could just be written using do notation). There could also be extra syntactic support for force/delay eg !exp to mean exp () and ~exp to mean \()->exp. With this extra sugar, it might be relatively painless to still use infinite data structures. Lazyness seems to sometimes be used to accomplish things that could easily be implemented in other ways, perhaps even better, and with more control (and certainly more explicitness) over the exact aspect of lazyness that is being used in a certain situation eg memoization, desugared attribute grammars, list-as-glue, etc. The advantages would be that the resulting language would be simpler to understand and could execute like lightning so there would be no need to use C or ML ever again. Afaik the full speed advantage can't be achieved just by syntactic sugar for regular Haskell since regular Haskell, even with seq etc, still needs to use a lifted type system, but syntactic sugar could certainly be used to implement such a language on top of Haskell to investigate if the complete absence of lazyness would cause any problems in practice. If I get time I might try to do this but I've shared the idea here in the vague hope someone else might want to do it first :-) There certainly seems to be a "gap in the market" between the perfection of lazy Haskell's monads, typeclasses etc and strict ML's side effects and less expressive type system. Regards, Brian. -- Logic empowers us and Love gives us purpose. Yet still phantoms restless for eras long past, congealed in the present in unthought forms, strive mightily unseen to destroy us. http://www.metamilk.com

3 3

do we have something like isDefined or isNull in Haskell?
by Vladimir Portnykh 15 Jun '06

15 Jun '06

Suppose there is a data definition in Haskell: data MyType = MyType { date :: Double, weight :: Double, height :: Double } deriving (Eq, Ord, Show) Is it possible to check if the field height, for example, is filled in(defined)? Can we give default values in Haskell? Many thanks and sorry fro so sily questions. Vladimir _________________________________________________________________ The new MSN Search Toolbar now includes Desktop search! http://join.msn.com/toolbar/overview

5 5

Learning C after Haskell
by Chad Scherrer 15 Jun '06

15 Jun '06

Ok, so I'm doing things somewhat backward. I've been using Haskell for a while now, whenever I get a chance to. But in order to become more involved in high-performance computing projects at my work, I need to learn C. I've heard a lot of people say that experience in Haskell can improve one's abilities in other languages, but I also wonder how different the C "way of doing things" is different from Haskell's. My question is, as I learn C, are there any particular Haskell concepts I should keep in the back of my mind, or is it better to approach C from scratch? Thanks in advance! Preparing for a foot-shooting, Chad

4 4

Re: [Haskell-cafe] do we have something like isDefined or isNull in Haskell?
by Bryan Burgers 15 Jun '06

15 Jun '06

Vladimir, I think you forgot to put Haskell-cafe as a recipient of this email, so first I'll repost what you wrote. On 6/15/06, Vladimir Portnykh <vportnykh(a)hotmail.com> wrote: > many thanks. i have the follwoing code: > > module MyType (DataContainer(..)) where > import Maybe > data DataContainer = > MyType { > date :: [Maybe Double], > weight ::[Maybe Double], > height ::[Maybe Double] > } > deriving (Eq, Ord, Show) > > myType = MyType { > date = > [38548,38562,38576,38590,38604,38618,38632,38646,38660,38674], > weight = [], > height = [] > } > > reportProblem = if isJust (height myType !! (length (date myType) - 1)) > then 0 else 1 > > > it failed to compiled saying "No instance for (Num (Maybe Double)) arising > from the literal 38674 . After reading your code for a while, I still don't know if I quite understand. Can you explain exactly what you are trying to do? My thought is that maybe you just want a list of objects, since you are indexing height at the same index as date. Then you just want a list of the structures that Duncan described. myObjects = [default {weight = 3.2}, default {weight = 7.9, date=38458},...] If not, hopefully somebody a little more knowledgable than I can help you. Bryan

1 0

Receiving multicasts
by Rich Neswold 15 Jun '06

15 Jun '06

Hello, Has anyone figured out a way to receive multicasts in a Haskell program? It doesn't appear that Network.Socket.setSocketOption provides enough information to join a multicast address. Any information would be appreciated. -- Rich AIM : rnezzy ICQ : 174908475

1 0

how to apply a function which returns IO() to a list?
by developer＠imaginando.net 14 Jun '06

14 Jun '06

Hi, I have this function which write a file: writeHtml b x y = do writeFile (b ++ ".html") (htmlCode b x y) i need to apply to each member of a given list: the way i always try to codify functions in haskell is allways the same, the most primitive one and i think i cant use it here. writeList [] = ??? writeList (x:xs) = (writeHtml x) ??? How should i do this? Many thx, Nuno SAntos

4 4

Closure trace?
by Michael Marte 14 Jun '06

14 Jun '06

Hello *, frequently we are seeing messages like "List.head: empty list" or "Maybe.fromJust: Nothing". It is clear what happened and where the messages were triggered yet the real cause is usually VERY hard to find unless the program is small and simple. I came to the conclusion that functions like head and fromJust are best to be avoided because their use may render large programs unmaintainable. Instead I use irrefutable pattern matching like (x : _) = l and Just bla = maybeBla whenever possible because when a pattern match fails the ghc runtime system gives me a nice error message naming the module, line and column. However, this procedure may become tedious when the error occurs in a call to a third-party library function. I wonder whether it is possible to print a "closure trace" similar to a stack trace in procedural programming. Say we have two modules A and B: module A where import B a = b module B b = c where c = error "go to hell" I would like to see something like this: sh$ ghci A <ghci startup messages> > a Runtime error "go to hell" in module B at line 2 Trace: B.b.c B.b A.a This way it would be easy to find the reason that actually caused, say, head to fail. I guess it would be sufficient to complement closures with information on the precise place of their definition. Michael

6 5