Re: [Chicken-hackers] Re: [Chicken-users] Release candidate for 2.6 available

[Kon Lovett]

On Feb 26, 2007, at 10:29 AM, Brandon J. Van Every wrote:

John Cowan wrote:

felix winkelmann scripsit:
  

Hi!

A release canditate for CHICKEN 2.6 is now available at
the web-site. If you find the time, please give it a try.
    

Works fine under Cygwin with CMake.

  

Builds and installs fine under MSVC.  Haven't really tested the functionality though.  It always pings my lazy-o-meter, the lack of tests.  Are there some simple tests we could agree on to validate the build?  If someone would cough those up, I could probably be arsed to incorporate them into the CMake build, ala the CTest / Dashboard mechanism.

I do the following (excerpted from build_chicken shell script):

DEV_ROOT="$HOME/Developer/Source/Scheme/Chicken"

DEVDIR="$DEV_ROOT/chicken"

cd $DEVDIR

case "$MODE" in

(bench | benchmark)

make bench "$@"

cd benchmarks

exec rm tmpfile.out tmpfile tmpfile.c

;;

(check)

pushd tests

export DYLD_LIBRARY_PATH=`pwd`/../.libs

export LD_LIBRARY_PATH=`pwd`/../.libs

../csi -script srfi-18-tests.scm

../csi -R numbers -script library-tests.scm

../csi -R numbers -case-insensitive -script r4rstest.scm

../csi -R numbers -R syntax-case -script r5rs_pitfall.scm

../csc -compiler ../chicken locative-stress-test.scm -o a.out && ./a.out; rm a.out

rm tmp*

popd

exec build_chicken bench "$@"

;;

...

esac

Then, after a "successful" build I run the 'check' as above. Note that my version of 'runtests.sh' is a little different in that I use numbers & ci.

The problem is no comparison against a "golden result", I must review the output manually.

Best Wishes,

Kon

Cheers,
Brandon Van Every

_______________________________________________

Chicken-hackers mailing list

address@hidden

http://lists.nongnu.org/mailman/listinfo/chicken-hackers