February 2002 - Glasgow-haskell-users

RE: CVS Build problem
by Simon Marlow 14 Feb '02

14 Feb '02

If you're building the CVS version of GHC, it usually helps to keep an eye on cvs-ghc(a)haskell.org where these kind of things are generally announced (in this case, I mentioned it in the commit message). The building guide tends to lag a little, but it will get updated. Cheers, Simon > -----Original Message----- > From: Hal Daume III [mailto:hdaume@ISI.EDU] > Sent: 13 February 2002 22:41 > To: Nicholas Nethercote > Cc: GHC users list > Subject: RE: CVS Build problem > > > Ah, thanks. I was having the same problem but was too shy to > ask about > it, figuring it must be my bad. This should probably be > mentioend in the > build documentation (hint, hint). > > - Hal > > -- > Hal Daume III > > "Computer science is no more about computers | hdaume(a)isi.edu > than astronomy is about telescopes." -Dijkstra | www.isi.edu/~hdaume > > On Wed, 13 Feb 2002, Nicholas Nethercote wrote: > > > On Wed, 13 Feb 2002, Nicholas Nethercote wrote: > > > > > That's what I did. It made ghc ok, but then when it > tries the 'make > > > boot' in hslibs it falls over. The same happens if I change into > > > $(BUILD)/hslibs and do 'make boot'... > > > > Ok, I've worked out the problem: I just checked out the > modules "ghc" and > > "hslibs". It seems that the standard library has just been > moved into a > > new separate module "libraries". So my build was missing > the standard > > library which was causing the problem. > > > > -- > > Nick Nethercote > > njn25(a)cam.ac.uk > > > > _______________________________________________ > > Glasgow-haskell-users mailing list > > Glasgow-haskell-users(a)haskell.org > > http://www.haskell.org/mailman/listinfo/glasgow-haskell-users > > > > _______________________________________________ > Glasgow-haskell-users mailing list > Glasgow-haskell-users(a)haskell.org > http://www.haskell.org/mailman/listinfo/glasgow-haskell-users >

1 0

RE: i386-pc-os2-emx port
by Simon Marlow 14 Feb '02

14 Feb '02

> I'm interested in doing a port of GHC to OS/2. > > Any particular documentation I should be reading before I > even think about > embarking on such a crazy endevour? GHC's Building Guide is a good place to start: http://www.haskell.org/ghc/docs/latest/building/building-guide.html Although note that bootstrapping on a new platform using the 5.xx series is a little difficult due to the way we preprocess some of the library sources using hsc2hs (although it has been done by at least one brave soul). You might want to start from 4.08.2 using .hc files and bootstrap up to 5.02 from there. 4.08.2 is the oldest version we support building the current sources with. We'll be here to provide guidance and moral support, as usual... Cheers, Simon

1 0

RE: CVS Build problem
by Simon Marlow 14 Feb '02

14 Feb '02

> I just checked a fresh copy of ghc + hslibs out of CVS, and > tried doing a > vanilla build. It fell over doing "make boot" in hslibs. > Here are the > last few make messages: ... > ../../ghc/utils/ghc-pkg/ghc-pkg-inplace --update-package > <lang.conf.inplace > Reading package info from stdin... done. > Expanding embedded variables...done. > dependency `base' doesn't exist > make[2]: *** [boot] Error 1 > make[1]: *** [boot] Error 1 > make[1]: Leaving directory `/local/scratch-4/njn25/ws2/build2/hslibs' > make: *** [all] Error 1 You need to 'make all' in fptools/ghc before 'make boot' in hslibs. The top-level fptools Makefile will arrange this if you just say 'make' at the top level. Cheers, Simon

4 5

RE: A question about GHC-5.03
by Simon Peyton-Jones 14 Feb '02

14 Feb '02

First thing to do is to read the manual (comes with the snapshot, I hope), where I tried to document it. Mark and I are writing the paper! You could try "Putting type annotations to work" by Odersky and Laufer for some useful background Simon -----Original Message----- From: Sean Seefried [mailto:seefried@itee.uq.edu.au] Sent: 14 February 2002 00:52 To: glasgow-haskell-users(a)haskell.org Subject: A question about GHC-5.03 One of the new features of the latest snapshot of GHC is that : "The type system now supports arbitrary rank polymorphism, given appropriate type annotations." Could you point me to some literature on this feature and how it is used? I'm familiar with rank-2 polymorphic types. I am not however familiar with what an "appropriate type annotation" is. Sean Seefried

1 0

A question about GHC-5.03
by Sean Seefried 13 Feb '02

13 Feb '02

One of the new features of the latest snapshot of GHC is that : "The type system now supports arbitrary rank polymorphism, given appropriate type annotations." Could you point me to some literature on this feature and how it is used? I'm familiar with rank-2 polymorphic types. I am not however familiar with what an "appropriate type annotation" is. Sean Seefried

1 0

Weird symbol prevents GHCi from loading my package.
by Axel Simon 13 Feb '02

13 Feb '02

Hi, I built quite a big library and turned it into a package. It works fine if use ghc -package gtk2 Blah.hs to compile things. But when I convert all the libraries to object files as described in the documentation, GHCi refuses to load the whole machinery: ~/source/mogul/fpDemo:$ ghci -package gtk2 Arabic.hs ___ ___ _ / _ \ /\ /\/ __(_) / /_\// /_/ / / | | GHC Interactive, version 5.02.2, for Haskell 98. / /_\\/ __ / /___| | http://www.haskell.org/ghc/ \____/\/ /_/\____/|_| Type :? for help. Loading package std ... linking ... done. Loading package lang ... linking ... done. Loading package gtk2 ... GHCi runtime linker: fatal error: I found a duplicate definition for symbol gcc2_compiled. whilst processing object file /home/as49/local/lib/gtk2hs/gtk2hs.o This could be caused by: * Loading two different object files which export the same symbol * Specifying the same object file twice on the GHCi command line * An incorrect `package.conf' entry, causing some object to be loaded twice. GHCi cannot safely continue in this situation. Exiting now. Sorry. The symbol obviously has nothing to do with GHCi, but I don't have a clue how it got there and it probably does not matter as long as I don't use GHCi. Does anybody know where it comes from? My library consists of a huge bunch of modules, most of them having C calls and callbacks. Thanks for hints, Axel. P.S.: Output from ~/source/mogul/fpDemo:$ ghc-pkg -s gtk2 Package {name = "gtk2", import_dirs = ["/home/as49/local/lib/gtk2hs/hi"], source_dirs = [], library_dirs = ["/home/as49/local/lib", "/home/as49/local/lib/gtk2hs", "/usr/X11R6/lib", "/usr/local/lib"], hs_libraries = ["gtk2hs"], extra_libraries = ["X11", "Xext", "atk", "gdk-x11-1.3", "gdk_pixbuf-1.3", "glib-1.3", "gmodule-1.3", "gobject-1.3", "gtk-x11-1.3", "iconv", "intl", "m", "pango", "pangox"], include_dirs = ["/home/as49/local/include/atk-1.0", "/home/as49/local/include/glib-2.0", "/home/as49/local/include/gtk-2.0", "/home/as49/local/include/pango-1.0", "/home/as49/local/lib/glib-2.0/include", "/home/as49/local/lib/gtk-2.0/include", "/home/as49/local/lib/gtk2hs/hi", "/usr/X11R6/include"], c_includes = ["Container_stub.h", "General_stub.h", "Signal_stub.h", "TreeSelection_stub.h", "gtk/gtk.h"], package_deps = ["lang"], extra_ghc_opts = ["-fglasgow-exts"], extra_cc_opts = [], extra_ld_opts = ["-u TreeSelection_dwzw", "-u TreeSelection_dwzx"]}

1 0

i386-pc-os2-emx port
by Russell O'Connor 13 Feb '02

13 Feb '02

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I'm interested in doing a port of GHC to OS/2. Any particular documentation I should be reading before I even think about embarking on such a crazy endevour? - -- Russell O'Connor roconnor(a)alumni.uwaterloo.ca <http://www.math.berkeley.edu/~roconnor/> ``This is not a time, as it is never a time, to seek vengeance, but a time to seek the courage to forgive'' -- George W. Bush -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.6 (SunOS) Comment: For info see http://www.gnupg.org iD8DBQE8as3QuZUa0PWVyWQRAj58AJ0bZg8qm7GZVx9eJH0uA8p+qvhB8ACfe8ki iaHWvxyrcR7IenWjplFVMy0= =95xI -----END PGP SIGNATURE-----

1 0

problem with source 5.03
by Hal Daume III 13 Feb '02

13 Feb '02

There seems to be a problem with the source tar.bz2 distribution of the 5.03 snapshot. The web page claims it's 5.3M, but the actualy file is only 4117517 bytes and doesn't untar properly: x ghc-5.03.20020204/ghc/InstallShield/Setup Files/Compressed Files, 0 bytes, 0 tape blocks x ghc-5.03.20020204/ghc/InstallShield/Setup Files/Compressed Files/Language Independent, 0 bytes, 0 tape blocks tar: directory checksum error I was able to grab a CVS image, but someone should probably fix this. - Hal -- Hal Daume III "Computer science is no more about computers | hdaume(a)isi.edu than astronomy is about telescopes." -Dijkstra | www.isi.edu/~hdaume

1 0

CVS Build problem
by Nicholas Nethercote 13 Feb '02

13 Feb '02

Hi, I just checked a fresh copy of ghc + hslibs out of CVS, and tried doing a vanilla build. It fell over doing "make boot" in hslibs. Here are the last few make messages: ------------------------------------------------------------------------ ===fptools== Finished making `all' in includes utils driver docs compiler rts ... PWD = /local/scratch-4/njn25/ws2/build2/ghc ------------------------------------------------------------------------ make[1]: Leaving directory `/local/scratch-4/njn25/ws2/build2/ghc' make[1]: Entering directory `/local/scratch-4/njn25/ws2/build2/hslibs' ------------------------------------------------------------------------ ===fptools== Recursively making `boot' in lang concurrent posix util data text net hssource tools ... PWD = /local/scratch-4/njn25/ws2/build2/hslibs ------------------------------------------------------------------------ ------------------------------------------------------------------------ ==fptools== make boot -wr; in /local/scratch-4/njn25/ws2/build2/hslibs/lang ------------------------------------------------------------------------ ../../ghc/utils/ghc-pkg/ghc-pkg-inplace --update-package <lang.conf.inplace Reading package info from stdin... done. Expanding embedded variables...done. dependency `base' doesn't exist make[2]: *** [boot] Error 1 make[1]: *** [boot] Error 1 make[1]: Leaving directory `/local/scratch-4/njn25/ws2/build2/hslibs' make: *** [all] Error 1 Is this my fault, or a problem with the CVS sources? Thanks. -- Nick Nethercote njn25(a)cam.ac.uk

1 0

RE: Edison library
by David Feuer 12 Feb '02

12 Feb '02

I'm sorry if I was confusing, but by "efficient" I meant "something other than unbalanced trees". Since the library provides a fairly large variety of heaps, it is surprising that it only provides a toy implementation of sets. It is particularly surprising because I know that Okasaki has written about balanced tree implementations before (e.g. in Purely Functional Data Structures). On Tue, 12 Feb 2002, Peter White wrote: > > There is an unbalanced set implementation of the Set class in there.... > Cheers > Peter White > Galois Connections > > -----Original Message----- > From: glasgow-haskell-users-admin(a)haskell.org > [mailto:glasgow-haskell-users-admin@haskell.org]On Behalf Of David Feuer > Sent: Tuesday, February 12, 2002 3:16 PM > To: glasgow-haskell-users(a)haskell.org > Subject: Edison library > > > Anyone know if Okasaki plans to expand Edison to include efficient sets > and/or other non-heap collections? It looks like the library hasn't > changed for 2 years... > > David Feuer > > _______________________________________________ > Glasgow-haskell-users mailing list > Glasgow-haskell-users(a)haskell.org > http://www.haskell.org/mailman/listinfo/glasgow-haskell-users > > >

1 0