May 2005 - Libraries - Haskell.org

sequences (was Re: Specify array or list size?)
by Ross Paterson 10 May '05

10 May '05

On Sat, May 07, 2005 at 10:56:11PM +0200, Benjamin Franksen wrote: > It would be a great improvement if Haskell's standard library (including the > Prelude) would provide a uniform framework for collections, and in particular > sequences (linear collections, indexed by natural numbers), and there should > be standard implementations that don't favour operations on the left end over > those on the right end. It is a shame that a language like Haskell doesn't > even have a Deque (double ended queue) in the its standard library. > > Robert Will has proposed a new unified standard for collections (as well as > some default implementations) with above properties (and a lot more). IMHO, > his work is (at least) a very good starting point and deserves more > attention. There was some discussion of this on the libraries list last year: http://www.haskell.org//pipermail/libraries/2004-April/thread.html http://www.haskell.org//pipermail/libraries/2004-June/thread.html (look for "sequences"). The background was that most people (Robert was an exception) thought that a general collection framework wasn't the way to go, partly because the proposals needed non-H98 features but mainly because there was no clear best choice, so a few concrete implementations (Data.Set, Data.Map, Data.IntSet and Data.IntMap) were added. However sequences can be described by a H98 class, and are also the area where a class with several instances is most useful.

1 0

RE: Should inet_ntoa Be Pure?
by Simon Marlow 10 May '05

10 May '05

On 07 May 2005 11:59, Dominic Steinitz wrote: > Does anyone know why these are in the IO monad? Aren't they pure > functions converting between dotted-decimal strings and a 32-bit > network byte ordered binary value? > > Dominic. > > http://www.haskell.org/ghc/docs/latest/html/libraries/network/Network.So cket.html#v%3Ainet_addr > http://www.haskell.org/ghc/docs/latest/html/libraries/network/Network.So cket.html#v%3Ainet_ntoa Clearly it should be pure. To avoid breaking the interface, we could add: showHostAddress :: HostAddress -> String and deprecate inet_ntoa? Cheers, Simon

2 1

RE: Bug in Cabal creating user package.conf
by Simon Marlow 09 May '05

09 May '05

On 04 May 2005 10:18, Ashley Yakeley wrote: > In article > <3429668D0E777A499EE74A7952C382D1038EBEE9(a)EUR-MSG-01.europe.corp.microso > ft.com>, > "Simon Marlow" <simonmar(a)microsoft.com> wrote: > >> It's not a newline issue, something else strange is going on here. > > You're right, if I run "runghc Setup.hs configure" a second time, it > works. Also, I have the same issue on Linux. Now fixed, thanks. Simon

1 0

problem with xlib binding
by Frederik Eaton 09 May '05

09 May '05

Hi, Hmm, have the new xlib bindings been tested? For instance, when I run the following program: ---------------------------------------------------------------- module Main where import Graphics.X11.Xlib import Graphics.X11.Xlib.Display main :: IO () main = do display <- openDisplay "" root <- rootWindow display $ defaultScreen display getGeometry display root closeDisplay display return () ---------------------------------------------------------------- It prints out: xtest: user error (Error bad status 1 raised in function getGeometry) >From the code it looks like the haskell getGeometry is calling XGetGeometry - which has type Status XGetGeometry(...) - and interpreting the 1 return value as an error (via the internal wrapper 'throwUnlessSuccess' which takes 0 for success). But the Xlib Programming Manual says "Some errors, such as failure to open a font, are indicated by return values of type Status on the appropriate routine ... The returned values are *zero* on failure and *nonzero* on success." (emphasis added) So unless I'm really misinterpreting things, the binding to each and every one of the 13 functions that uses 'throwUnlessSuccess' is wrong. Well, before I realized this I'd gone ahead and written a reasonably significant program against the binding assuming that at least the basics would work, but it seems there is some distance to go before the library is usable. One would think that with a "Stability: provisional" library every function has been run successfully at least once, but I can't imagine that this is the case here. Is there even a working Haskell equivalent of 'basicwin' from the Xlib Programming Manual to test things out? This would also be a helpful example for users. Regards, Frederik -- http://ofb.net/~frederik/

3 3

Should inet_ntoa Be Pure?
by Dominic Steinitz 07 May '05

07 May '05

Does anyone know why these are in the IO monad? Aren't they pure functions converting between dotted-decimal strings and a 32-bit network byte ordered binary value? Dominic. http://www.haskell.org/ghc/docs/latest/html/libraries/network/Network.Socke… http://www.haskell.org/ghc/docs/latest/html/libraries/network/Network.Socke…

2 1

Cabal and c2hs
by Dominic Steinitz 06 May '05

06 May '05

I am trying to migrate some old code to cabal. It uses c2hs. Now I wasn't sure I had c2hs installed so I ran cabal anyway. 1. Why isn't configure telling me where c2hs is? It tells me where hsc2hs is! 2. Why is cabal telling me it can't find a source when the source is present? Thanks, Dominic. Here's ping.cabal: Name: Ping Version: 0.0 License: BSD3 Author: Dominic Steinitz Synopsis: The Haskell equivalent of ping Executable: ping Main-Is: test.hs Other-modules: IP_ICMP.chs When I run ./Setup configure, I didn't see any message saying whether c2hs was installed and when I run ./Setup build I get the following: [dom@tility ping]$ ./Setup build Preprocessing executables for Ping-0.0... can't find source for IP_ICMP.chs in ["."] But the source is there: [dom@tility ping]$ ls codes.hs dist/ ip_icmp.h ping.cabal Setup* Setup.hs test.hs CVS/ IP_ICMP.chs Makefile ping.cabal~ Setup.hi Setup.o

4 10

RE: Request: Open Module Resolution
by Simon Marlow 06 May '05

06 May '05

On 03 May 2005 23:01, S. Alexander Jacobson wrote: > Proposal: > > 1. Define a standard resolver module > > module Resolve where > > type Resolver = Args -> ModuleName -> IO HaskellSource > resolve::Resolver > > 2. Require that Haskell compilers allow users to supply resolvers > e.g. > > ghci -r ../../Resolve.hs That requires compile-time code generation and execution (i.e. template haskell), so I can't see all compilers implementing it. Instead, let me point you at the place in GHC to insert your resolver. In the module Finder, we have: findModule :: HscEnv -> Module -> Bool -> IO FindResult data FindResult = Found ModLocation PackageIdH -- the module was found | PackageHidden PackageId -- for an explicit source import: the package containing the module is -- not exposed. | ModuleHidden PackageId -- for an explicit source import: the package containing the module is -- exposed, but the module itself is hidden. | NotFound [FilePath] -- the module was not found, the specified places were searched. You can almost replace findModule with your own resolver. However, the question is what do you do for modules like "Prelude": it sounds like you'll need to have pre-compiled code for all the standard libraries around (I assume you want to avoid using packages). Cheers, Simon

2 1

NewBinary Library
by Peter Simons 06 May '05

06 May '05

Hi, the NewBinary library used to be available via Darcs at this place: http://www.n-heptane.com/nhlabca/repos/NewBinary It appears, though, that this repository has vanished. Does anyone know what happened? Is anyone aware of the new home of the library? Peter

2 1

RE: Bug in Cabal creating user package.conf
by Simon Marlow 04 May '05

04 May '05

On 30 April 2005 09:06, Ashley Yakeley wrote: > In article <87ekcsr9u0.fsf(a)syntaxpolice.org>, > Isaac Jones <ijones(a)syntaxpolice.org> wrote: > >> Hmm. How sure are you that it was cabal that created the >> package.conf file? This is the code that does so, and it definitely >> seems to add a newline: > > This is on Mac OS X, and it's definitely Setup.hs that's creating the > file, and it definitely has just "[]". > > Transcript (prompt reduced to "$ " for brevity): > > $ cat Setup.hs > import Distribution.Simple > main = defaultMain > $ rm -rf ~/.ghc > $ runghc Setup.hs configure > Warning: Package is copyright All Rights Reserved > Warning: No license-file field. > Configuring TimeLib-0.1... > configure: searching for ghc in path. > configure: found ghc at /usr/local/bin/ghc > configure: looking for package tool: ghc-pkg near compiler in > /usr/local/bin/ghc > configure: found package tool in /usr/local/bin/ghc-pkg > configure: Using install prefix: /usr/local > configure: Using compiler: /usr/local/bin/ghc > configure: Compiler flavor: GHC > configure: Compiler version: 6.4 > configure: Using package tool: /usr/local/bin/ghc-pkg > configure: Using haddock: /usr/local/bin/haddock > configure: Using happy: /usr/local/bin/happy > configure: Using alex: /usr/local/bin/alex > configure: Using hsc2hs: /usr/local/bin/hsc2hs > configure: No cpphs found > configure: Reading installed packages... > cannot parse package list > $ cat ~/.ghc/powerpc-darwin-6.4/package.conf > []$ uname -a > Darwin Avebury.local 7.8.0 Darwin Kernel Version 7.8.0: Wed Dec 22 > 14:26:17 PST 2004; root:xnu/xnu-517.11.1.obj~1/RELEASE_PPC Power > Macintosh powerpc > $ ghc --version > The Glorious Glasgow Haskell Compilation System, version 6.4 > $ ls -Al ~/.ghc/powerpc-darwin-6.4/package.conf > -rw-r--r-- 1 ashley ashley 2 Apr 30 00:57 > /Users/ashley/.ghc/powerpc-darwin-6.4/package.conf It's not a newline issue, something else strange is going on here. Your local package.conf is being created by ghc-pkg (which is a little eager to create it in 6.4; 6.4.1 will be better). The parse error is coming from Cabal (Distribution.Simple.Configure) which is having trouble parsing the list of packages from ghc-pkg list. Could you post your output from "ghc-pkg list" after running the configure step above? Cheers, Simon

2 1

Request: Open Module Resolution
by S. Alexander Jacobson 04 May '05

04 May '05

Problem: Haskell has both no standard way to resolve module names to implementations and no easy way to experiment with different solutions. For example, I would like to experiment with a resolver that uses a lookup table to map module names to URLs of module implementations. Proposal: 1. Define a standard resolver module module Resolve where type Resolver = Args -> ModuleName -> IO HaskellSource resolve::Resolver 2. Require that Haskell compilers allow users to supply resolvers e.g. ghci -r ../../Resolve.hs -Alex- PS This request is the output of discussion I was having with Dimitry Golubovsky (and others). ______________________________________________________________ S. Alexander Jacobson tel:917-770-6565 http://alexjacobson.com On Fri, 29 Apr 2005, Dimitry Golubovsky wrote: > Alex, > > Quoting your message in Libraries: > > ================= > > We should think of module names as being like domain names, and module > implementations as being like IP addresses. Just as the owner of a > domain name determines valid IP addresses for it, the owner of a > module name should be able to determine valid implemementations. > Just as you can't provide domain name resolution for domain names you > don't own, you should not be able to provide module name resolution > for module names you don't own. > > We need a Module Name System analogous to the Domain Name System. > Haskell compilers should be able to query a root server to find which > MNS server is authoritative for a particular module name. A query to > that MNS server should result in URLs for implementations of that > module. > > =============== > > Some time ago I wrote up on Haskell-cafe: > > http://www.haskell.org//pipermail/haskell-cafe/2005-March/009515.html > > about proposed mechanism of module name resolution. What do you think? > This can be done flexible enough with minimal changes to the compiler > (just make it open a pipe instead of a file) and letting an external > program do the business. > > Nobody answered though. > > I am writing off-list, but if you find my ideas useful feel free to > move the discussion to Libraries or Cafe. > > -- > Dimitry Golubovsky > > Anywhere on the Web >

1 0