On Apr 7, 10:52 am, Max Erickson maxerick...@gmail.com wrote:
Most of the remaining failing tests have unicode characters.
I guess the problem is with file input in the mingw libraries, but I
haven't figured out where to look yet. Any tips there?
The root of the issue is that std::isspace
Next step is to try to get the tests working (the simple commands I
have tried do work and at a glance do output correct values).
A few changes to RegressTests.py to account for different path
seperators, shell quoting and shell escapes and most of the baseline
and regress tests pass
So my attempt to get acprep to run aclocal doesn't work. That explains
libtool emitting the wrong commands. A little fiddling with the
autoconf stuff got it working (I haven't minimized and verified the
commands, but I think just aclocal -I m4 and ./config.status libtool
are enough.).
The last
On Mar 21, 10:42 am, Max Erickson maxerick...@gmail.com wrote:
http://paste.pocoo.org/show/569048/
I guess something with mingw is missing or set wrong, but I have no
idea what to look for (and zero experience with C++ and minimal
experience with mingw and msys...).
Any suggestions?
Max
Max Erickson maxerickson-re5jqeeqqe8avxtiumw...@public.gmane.org writes:
Apparently attempting to send something coherent helped me gain a little
clarity; the problem shown in my paste is probably because one of the mingw
system headers contains
#define VOID void
Would it be a problem to
On Mar 22, 4:21 pm, John Wiegley jwieg...@gmail.com wrote:
How about at the top of value.h you just add this:
#if defined(VOID)
#undef VOID
#endif
Does that work? If so, I can check that in.
John
Yeah, that does work. Then mingw doesn't have setenv:
make[2]: Entering directory
Max Erickson maxerickson-re5jqeeqqe8avxtiumw...@public.gmane.org writes:
More broadly, I wonder if this 'should' work? I guess in the sense, do you
expect that minimal changes should be enough to successfully compile with
mingw32?
I have no reason off-hand to believe that it shouldn't work.
Hello-
I made enough changes to acprep to get configure to complete in a msys
shell and then things bomb out during make.
The changes are shown in the pastebin, but summarizing, they are
'use / as path seperator', 'don't look for python-config' and some
monkeying around to run aclocal with a