Try again with smaller size so it is not rejected by list server. -----Original Message----- From: Mike Thomas [mailto:miketh@brisbane.paradigmgeo.com] Sent: Monday, May 19, 2003 9:51 AM To: johago@web.de; glasgow-haskell-users@haskell.org Subject: RE: building ghc from source Hi Johannes. Notice that the build fails the first time that "ghc-inplace" gets called. Now look back at your configure output and see the line: checking whether we can open files in binary mode... no which should read: checking whether we can open files in binary mode... yes The problem is that when this step goes wrong on Windows, the compiler is built in such a way that it can't read binary files (eg "*.hi") correctly. I'm afraid I can't remember what caused this configuration test to go wrong for me in the past, but I solved it by looking at the file "configure.log", which I suggest you do. I seem to recall that, due to a compiler configuration problem, a header or library was not found, which meant that the test program was doomed to fail. I can't say that this is the only problem you will have, but it is certainly fatal on Windows so: Best of luck! Mike Thomas. -----Original Message----- From: glasgow-haskell-users-admin@haskell.org [mailto:glasgow-haskell-users-admin@haskell.org]On Behalf Of johago@web.de Sent: Sunday, May 18, 2003 2:15 AM To: glasgow-haskell-users@haskell.org Subject: building ghc from source Hi there! I'm trying to build ghc from scratch under Windows XP... which gives me problems, problems, problems... Steps taken: ....bits removed.... checking whether we can open files in binary mode... no .... more bits removed ... ../utils/ghc-pkg/ghc-pkg-inplace -f ../driver/package.conf --update-package