June 2006 - Glasgow-haskell-users

recursive import
by Serge D. Mechveliani 06 Jun '06

06 Jun '06

Concering the current state of recursive import, Simon Peyton-Jones <simonpj(a)microsoft.com> writes on 5 Jun 2006 > [..] > > GHC's module-at-a-time compilation model means that GHC needs an hs-boot > file to "get started". The manual tries to explain what you can and > can't do. If you follow the rules it works pretty well. We use > recursive modules extensively for compiling GHC itself, and it's just > fine with 'make'. (ghc -M does the right thing.) > > [..] Thank you for the explanation. And how does this feature interact with packages, with making via Cabal ? Will Cabal work with recursive imports and with ghc-options: -fglasgow-exts -fallow-undecidable-instances -fallow-overlapping-instances -fno-warn-overlapping-patterns -fwarn-unused-binds -fwarn-unused-matches -fwarn-unused-imports -O -prof -auto-all ? Generally, I tend to use recursive modules because they simplify the global algorithm presentation. But fear of technical difficulties in the project support. Maybe, I would try this recursion. ----------------- Serge Mechveliani mechvel(a)botik.ru

3 2

Bug? Or at least a better error message?
by Geoffrey Alan Washburn 05 Jun '06

05 Jun '06

While trying out the following example, in an attempt to learn something about the fiddly case where a type class instance tries to use an instance that is more specific than itself module Testing where class Foo a where { bar :: a -> Int } instance Foo Int where bar i = i instance Foo a => Foo [a] where bar [] = bar [1] bar ([x]) = 1 bar (x:xs) = (bar x) + (bar xs) It is kind of like polymorphic recursion, I suppose. I get the following exception [1 of 1] Compiling Testing ( strange.hs, interpreted ) *** Exception: typecheck/TcEnv.lhs:(365,0)-(392,32): Non-exhaustive patterns in function find_thing Is the example supposed to work? I'm trying to determine the source of a problem with type classes and GADTs and I figured this example using "normal" data types would be a good place to start in understanding what was going wrong. I was using ghci version 6.5.20060503.

3 6

Data.Sequence?
by Michael Marte 02 Jun '06

02 Jun '06

Hello *, when importing Data.Queue in 6.4.2, I am told that it is deprecated and that I should use Data.Sequence instead. However, Data.Sequence is not part of the base package - it is there but it is missing from the package.conf file. Moreover, there is no documentation available apart from the source code itself. Regarding Data.Sequence itself: - I am missing a function for converting a sequence to a list directly; I have to create a view and convert it to a list. - Isn't the name too fuzzy? After all, a standard Haskell list is a sequence as well. Why not stay with Queue? Michael

5 8

Understanding directory structure of ghc installation
by Brian Hulley 01 Jun '06

01 Jun '06

Hi - I've just downloaded and installed Edison using Cabal. However I now see that Edison seems to have been already installed with the ghc 6.4.2 installation, even though there's no mention of Edison in /doc/html/libraries/index.html, because there are old Edison .hi files in ghc-6.4.2/hslibs-imports/data whereas the Cabal install process copied newly built Edison files to C:\Program Files\Haskell\EdisonAPI-1.2\ghc-6.4.2\Data and C:\Program Files\Haskell\EdisonCore-1.2\ghc-6.4.2\Data I was about to just delete the old .hi files from the hslibs-imports/data directory, but then I noticed there are a few .hi files there for Set and FiniteMap, even though there are also .hi files (though of different sizes) for Set and FiniteMap stored in the imports/data directory as well. So my question is: how to understand all this? Why are there different copies of Set.hi which have different sizes, but in the same installation, and which one does ghc use when I use import Data.Set in a program? Also, what does .p_hi mean? (The new build of Edison only creates .hi files not .p_hi files but all the ghc-6.4.2 modules have both .hi and .p_hi files) (It would also be interesting to know where Cabal gets the information from to configure a package, and where the package database is stored...) Thanks, 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