Hello
I'm trying to build the module under Windows and I get several "initializer element is not constant" error. My guess is that the error is caused by the the use of &NyImmBitSet_Type, &NyCplBitSet_Type,&NyNodeSet_Type when initializing the struct array nysets_heapdefs. The declarations on sets_internal.h (lines 8,9,10) might be conflictive with the win32 gcc.
The only helpful info I've found on the net were the following posts:
http://mail.python.org/pipermail/distutils-sig/2001-May/002390.html
http://www.cygwin.com/ml/cygwin/2003-01/msg00040.html
Sadly I were not able to implement the changes myself.
This error has been reproduced on
* Cygwin - Gcc 3.4.4 (cygming special, gdc 0.12, using dmd 0.125
* Migwin - Gcc Idem Cygwin
* Using guppy-0.1.8.tar.gz
I hope this helps
Thanks and regards
---
src/sets/sets.c:77: error: initializer element is not constant
src/sets/sets.c:77: error: (near initialization for `nysets_heapdefs[0].type')
src/sets/sets.c:77: error: initializer element is not constant
src/sets/sets.c:77: error: (near initialization for `nysets_heapdefs[0]')
src/sets/sets.c:78: error: initializer element is not constant
src/sets/sets.c:78: error: (near initialization for `nysets_heapdefs[1].type')
src/sets/sets.c:78: error: initializer element is not constant
src/sets/sets.c:78: error: (near initialization for `nysets_heapdefs[1]')
src/sets/sets.c:79: error: initializer element is not constant
src/sets/sets.c:79: error: (near initialization for `nysets_heapdefs[2].type')
src/sets/sets.c:79: error: initializer element is not constant
src/sets/sets.c:79: error: (near initialization for `nysets_heapdefs[2]')
src/sets/sets.c:80: error: initializer element is not constant
src/sets/sets.c:80: error: (near initialization for `nysets_heapdefs[3]')
error: command 'gcc' failed with exit status 1
---
Logged In: NO
by the way, my username is facuman. Regards
The patch by Chad Austin for MSWINDOWS, being discussed on the mailing list recently,
may fix this problem though I haven't tested it myself other than in Linux (yet...).
I have applied the patch to the trunk, and after some minor fixes it compiles cleanly in Linux.
It is SVN rev 52 now. If you want to try it out in cygwin, migwin...looking forward to hear
how it turned out :-)
Sverker
I tried the patch on Windows Vista + Python 2.5 + MinGW and unfortunately did not work.
I delete the /Z compiler option from setup (because it's a MSVS option) and the error reported here popped up again :-(
Ric.