November 2002 - Glasgow-haskell-users

RFC: External library infrastructure
by Simon Marlow 12 Nov '02

12 Nov '02

Hi Folks, It has occurred to us that the GHC community really needs an infrastructure (i.e. like a build system, but more) that people can use for shipping their own libraries independently of GHC. Several people have asked recently if they can ship their libraries with GHC - for a while we were quite receptive to this, but GHC is getting rather "heavy" both in terms of build time and maintenance effort, and we feel it's time to cut back a bit. But we don't want to leave library developers out in the cold. Suppose we had an infrastructure which someone could plug their library source code into, and immediately get - a source tarball which will configure/build/install on a Unix/cygwin system with GHC installed, and provide a GHC package when installed. - skeleton binary packages for RPM, Windows installer, *BSD ports, Gentoo ebuild, etc. The binary packages will require some programmer intervention (such as setting the dependencies correctly), but much of the work can be done automatically. Furthermore, we'd provide links to compatible library packages from the GHC web pages. It doesn't particularly matter whether fptools is used as a starting point, or we pick another existing build system or start from scratch. The important thing is that someone who wants to ship a library can pick up this infrastructure and quickly provide all the above functionality with very little extra work. Comments? Cheers, Simon

4 3

RE: Floats and Doubles
by Simon Marlow 12 Nov '02

12 Nov '02

> I have been using some of the functions of the classes Real and > Fractional and I have observed that with the funcion > "toRational" we can > obtain the fraction that represents a given number. For instance: > *P2> toRational (5.2::Float) > 5452595 % 1048576 > Why we obtain this numbers instead of "52 % 10" or "26 % 5"? Because you asked for the number to be converted to a Float first, and 5.2 can't be represented exactly in a single-precision float. If you don't go via Float, you get the exact result: Prelude> 5.2 :: Rational 26 % 5 > I have also obtained the following results with the functions > "fromRational" and "toRational": > *P2> (fromRational ((toRational 4) - ( toRational > (5.2::Float) )))::Double > -1.1999998092651367 > *P2> (fromRational ((toRational 4) - ( toRational > (5.2::Double) )))::Double > -1.2000000000000002 > *P2> (fromRational ((toRational 4) - ( toRational 5.2 ))) > -1.2000000000000002 > *P2> (fromRational ((toRational 4) - ( toRational > (5.2::Float) )))::Float > -1.1999998 These are all symptoms of rounding. Cheers, Simon

1 0

Re: Generating a Windows executable
by Chris Saunders 11 Nov '02

11 Nov '02

I discovered what my problem was. I was not aware that the executable would be deposited into the current working directory of the command prompt. Since the intermediate files were generated in the same folder as the source file, I assumed the .exe would go there also. The executable was generated correctly with the command lines suggested by both Seth and Hal - and thanks for your responses. Regards Chris Saunders chris.saunders.(a)sympatico.ca

1 0

Building a Windows executable
by Chris Saunders 11 Nov '02

11 Nov '02

I am new to Haskell and GHC and am using a tutorial that I found on the Web to learn. In the tutorial it says that a command like this can be used to build an executable (Windows): ghc --make test.hs -o test.exe The compiler had no complaints and output 2 files: Main.hi and test.o This is the code I was attempting to compile: module Main where main = putStrLn "Hello World." I also tried using: ghc --make test.hs and got the same result. Sorry for such a lame question. Regards Chris Saunders chris.saunders.(a)sympatico.ca

1 0

functional dependency syntax?
by Malcolm Wallace 11 Nov '02

11 Nov '02

I'm trying out MPTC with functional dependencies for the first time in GHC. However, I'm not sure of the syntax. MPJ's ESOP 2000 paper gives the following example: class FiniteMap i e fm | fm -> (i,e) where emptyFM :: fm lookupFM :: i -> fm -> Maybe e extendFM :: i -> e -> fm -> fm but ghc (5.02.2 and 5.04.1) complain $ ghc-5.04.1 -fglasgow-exts -c -o FunDep.o FunDep.hs FunDep.hs:10: parse error on input `(' where the indicated line is that of the class declaration. What should the syntax be if it is not the above? Regards, Malcolm

2 1

RE: problems with the ffi
by Simon Marlow 11 Nov '02

11 Nov '02

> I am having problems with the ghc ffi, I have compiled the > Haskell source on the ffi example page to create the stub > file and the c file, the only problem now is that I can't > compile the sample c code. I assume that if you place the c > code and all the files in the ghc include directory that it > should compile. However I am getting errors, to begin I get > something like can't find stint.h when it is definitely > present. Then I get errors with the special Haskell ffi > procedures i.e. startupHaskell(argc, argv, __stginit_Foo). > The compiler says that this is an unresolved external symbol. > > I am attempting this on windows xp using visual c++ 5. Any > help would be greatly appreciated. It would help if you could post the exact procedure you're using, and the error messages you get (just cut 'n' paste it). Cheers, Simon

1 0

problems with the ffi
by David Abbs 10 Nov '02

10 Nov '02

I am having problems with the ghc ffi, I have compiled the Haskell source on the ffi example page to create the stub file and the c file, the only problem now is that I can't compile the sample c code. I assume that if you place the c code and all the files in the ghc include directory that it should compile. However I am getting errors, to begin I get something like can't find stint.h when it is definitely present. Then I get errors with the special Haskell ffi procedures i.e. startupHaskell(argc, argv, __stginit_Foo). The compiler says that this is an unresolved external symbol. I am attempting this on windows xp using visual c++ 5. Any help would be greatly appreciated. ____________________________________________________ David Abbs 3rd Year Computer Science University Of Sheffield u0da2(a)dcs.shef.ac.uk ____________________________________________________

1 0

Newslwtter
by Marta 09 Nov '02

09 Nov '02

http://strappone.arecool.net sono arrivate le strappone!!!! finalmente nel web ------> vieni a conoscerle!!! -----> http://strappone.arecool.net <----- per non ricevere altre mail: mbaxto(a)hotmail.com

1 0

RE: profiling mutator time per function vs. GC time
by Simon Marlow 08 Nov '02

08 Nov '02

> It seems I have two options: > > A) compile without profiling support and run the compiled > program with +RTS -sstderr > > B) compile with profiling support -prof -auto-all and run > the compiled program with +RTS -p -sstderr > > In case A, I get a good measure of GC vs. mutator time, but I > don't know > the amount of time used by single functions, so I can't > seperate between > mutator time spent in the functions that really interest me > and the time > spent for the test frame. > > In case B, I can seperate mutator time spent in single > functions, but I > have the impression that the GC time I then get includes the > GC for the > profiler, and hence is useless for me, because different GC times for > different algorithms might just mean that one of the > algorithms requires more profiling overhead. If you run a profiled program with +RTS -sstderr, the time breakdown includes an extra category, PROF, which counts the time spent in the heap profiler. The amount of GC time consumed by the profiled program will indeed be different from the unprofiled program, because of profiling overheads - there's no way around this, I'm afraid. But you may find that the ratio of mutator to GC time in the profiled program is similar to the unprofiled program (I'd be interested to know whether this is/is not the case). To get the most reliable measure of GC time in a profiled program, do not turn on heap profiling, because this will cause extra GCs to be performed and will inflate the GC time. Turing on time profiling should be ok. Cheers, Simon

2 1

RE: GHCi on Mac OS X
by Simon Marlow 08 Nov '02

08 Nov '02

> When I start the ghci with -package data, I get the following output: > > Loading package base ... linking ... done. > Loading package haskell98 ... linking ... done. > Loading package lang ... linking ... done. > Loading package concurrent ... linking ... done. > Loading package posix ... linking ... done. > Loading package util ... linking ... not found: _rl_funmap_names > Abort > > I know that I must install the readline library seperately; > can anybody tell me how to do this? It may be that you have the wrong version of the readline library installed. We're chasing a similar problem on SuSE Linux, but I haven't managed to trace the cause yet. Sorry I can't be any more help... Wolfgang - any ideas? Cheers, Simon

2 1