December 2010 - Haskell-Cafe

threadWaitRead and threadWaitWrite on multiple fds
by Mathijs Kwik 13 Dec '10

13 Dec '10

Hi all, I read the paper about the new ghc7 event handling IO manager goodies. This is all very exciting stuff. I didn't know GHC's RTS had these smart async-IO facilities. The paper pointed me at threadWaitRead/threadWaitWrite. While very nice the way they are, I would also like to be able to wait on more than 1 fd until 1 of them becomes available. Would it be possible to create this functionality myself? Or do I need to request it and wait for a new GHC? It looks like the functionality I need is in System/Event/Thread. I think I can manage to write my own version of those functions which then loop through a list of fds, but the original file imports System/Event/Manager (registerFd, unregisterFd_) which I can't do because it's a hidden module. Is there a way around this? Thanks, Mathijs

4 6

Re: [Haskell-cafe] A home-brew iteration-alike library: some extension quiestions
by John Lato 13 Dec '10

13 Dec '10

From: Permjacov Evgeniy <permeakra(a)gmail.com> > > Hi. I Wrote a simple iteration library. It was not intensively tested, > so it MAY contatin bugs, but it is very unlikely. The library is > currently on github: https://github.com/permeakra/iteration > > I'm not ready to upload it to hackage, as some testing and extension is > really needed. However, I'd like to know about possible flaws. > > Current goal is addition of byte-stream (de)compression and IO functions > extenstion. After this package will be cabalized and uploaded to > hackage. So, while design is not frozen yet, I'm interested in criticism > -)/ > First, I haven't examined your code in sufficient depth to fully understand it, so I may be off the mark with this comment. However, I'm not convinced it's possible to safely implement zipping in iteration-style IO without another technique (e.g. monadic regions), by which I mean I suspect any code which exposes enough control to pause an enumeration (or alternatively allows a user-supplied termination check to the enumerator) will also not provide guaranteed finalization/garbage collection of the underlying resource (e.g. handle). I also require zipping of streams, and in fact it is possible with iteratee. I haven't included the code in the library because I was formerly not convinced of its safety. I do think it's safe now, but only by use of a monadic region. John

1 0

Re: [Haskell-cafe] A home-brew iteration-alike library: some extension quiestions
by oleg＠okmij.org 13 Dec '10

13 Dec '10

Just for the record: the library IterateeM.hs, uses NO extensions to Haskell98, let alone Haskell2010. The library as written requires LowLevelIO.hs, which uses FFI (which has been Haskell98 addendum and is in proper Haskell2010). The sample code, Wc.hs, for example, is Haskell98. So, the iteratee can be written with no extensions whatsoever. In particular, IterateeM does not use any monad transformer library (although it could have). I found that the trouble of writing a state monad for a particular state is negligible compared to the pain of choosing a particular monad transformer library, and especially the pain inflicted on the users who have to deal with many a conflicts of monad transformer libraries. > The problem was that I wished Zippee. It means that external enumerator > must be "suspended" at some points so Zippee can process elements from > both left and right streams in desired order. It makes any other > approach I considered impossible to use. The file IterateeN.hs demonstrates zipping two streams together (in lock-step and not in-lockstep). It turns out, the existing Iteratee interface and type suffices. This is described in more detail in: Parallel composition of iteratees: one source to several sinks http://okmij.org/ftp/Streams.html#1enum2iter Parallel composition of streams: several sources to one sink http://okmij.org/ftp/Streams.html#2enum1iter

2 1

gtk2hs and hmake websites
by Joachim Breitner 12 Dec '10

12 Dec '10

Hi, I get regular error message from some Debian tools that check for new upstream versions that it cannot find http://haskell.org/hmake, and http://haskell.org/gtk2hs seems to be gone as well since the recent move of haskell.org. Is that temporary or will these projects have to find new homes? Greetings, Joachim -- Joachim "nomeata" Breitner mail: mail(a)joachim-breitner.de | ICQ# 74513189 | GPG-Key: 4743206C JID: nomeata(a)joachim-breitner.de | http://www.joachim-breitner.de/ Debian Developer: nomeata(a)debian.org

4 3

The Monad Reader links are broken
by Jason Dagit 12 Dec '10

12 Dec '10

Hello, I noticed today that the links in this article point to Haskell.org and they are broken: http://themonadreader.wordpress.com/previous-issues/ Maybe someone can fix this? Thanks! Jason

2 1

Avoiding the Y combinator for self-referencing types
by Florian Weimer 12 Dec '10

12 Dec '10

Suppose I've got some named objects which reference other objects by name: > data NodeS = NodeS {nameS :: String, refsS :: [String]} Through name resolution, the strings are translated to the actual nodes they denote: > data Node = Node {name :: String, refs :: [Node]} > resolve :: [NodeS] -> Map String Node NodeS and Node are quite similar, so they should probably be the same parametrized type. However, if I turn the type of the references into a type parameter, I need the type-level Y combinator to avoid an infinite type, and manipulating nodes turns a bit more tedious because of increased syntactic overhead. Is there a third choice, beyond manually expanding the type definition or using Y?

2 2

ANN: improve-0.2.2
by Tom Hawkins 12 Dec '10

12 Dec '10

ImProve [1] is an imperative DSL for high assurance, embedded applications. This release includes a new compositional proof framework where users can leverage previously proved theorems to aid the proof of new theorems. This new addition was inspired from discussions with Lee Pike [2]. Lee also recommended the strategy of building disjunctive invariants to reduce, if not eliminate the need for multi step induction, thus dramatically reducing proof time. Using these new strategies with ImProve, we where able to verify several realtime safety properties of an Eaton Hybrid Hydraulic program. -Tom [1] http://hackage.haskell.org/package/improve [2] http://groups.google.com/group/fp-embedded/browse_thread/thread/63cd023e8f1…

1 0

Contexts for type family instances
by Stephen Tetley 12 Dec '10

12 Dec '10

Hello all I've a type family DUnit that I use to allow the unit type, usually a Double, of geometric things (points, vectors...) to be parametric: > type family DUnit a :: * E.g for Points > data Point u = P2 u u > type instance (Point u) = u I can make an type instance for Maybe like this, using DUnit on the right hand side: > type instance DUnit (Maybe a) = DUnit a But pair is a problem. Is there a way to assert both parts of a pair should have the same DUnit? The code below won't compile but it should illustrate what I'm trying to do. > type instance (DUnit a ~ DUnit b) => DUnit (a,b) = DUnit a I don't want to pick an arbitrary side, e.g: > type instance DUnit (a,b) = DUnit a or > type instance DUnit (a,b) = DUnit b Thanks Best wishes Stephen

2 2

How can i insert graphic into a haddock document?
by Poprádi Árpád 12 Dec '10

12 Dec '10

Hi, I have found nothing about this topic in the haddock documentation. Is there a light way to do that? Best regards, Árpád

2 2

haskell2010 and state question.
by Magicloud Magiclouds 11 Dec '10

11 Dec '10

Hi, Formerly, I had IORef and some state monad to do the task of keeping states. Now in haskell 2010, I cannot find anything about it. Do I have to use ghc base package for this function? -- 竹密岂妨流水过山高哪阻野云飞

6 6