Re: Problems configuring libtclobjc

[glen e. p. ropella]

Hey Karl!

Sorry we didn't get a libtclobjc-1.2 out with v1.0.  But, we were
just swamped.  Also, we're trying to let swarm-dev die peacefully.
Could you stick with swarm-support?

Now, on with important things:
> I'm using gcc 2.7.2.1 with the objc-960906 patches.  When I try configuring
> libtclobjc-1.1b6 I get this message:
> 
> mycroft$ ./configure
> [...]
> checking for NeXT compiler
> checking for -lobjc... (cached) yes
> checking for objc_next_class... (cached) no
> You must apply the patch ./gcc.patch to the gcc source dir
> then recompile and install libobjc.a,
> then rerun this configure script.
> 
> So I have a patched gcc, but configure says I don't.  Suggestions?  I'm
> not having problems configuring, compiling, and running other Obj-C
> programs.
> 
> -- 
> Karl Sackett                                           address@hidden

I went ahead and parsed out the little test program that the
configure script uses.  It looks like:
-----------------
/* System header to define __stub macros and hopefully few prototypes,
    which can conflict with char objc_next_class(); below.  */
#include <assert.h>
/* Override any gcc2 internal prototype to avoid an error.  */
char objc_next_class();

int main() { return 0; }
int t() {

/* The GNU C library defines this for functions which it implements
    to always fail with ENOSYS.  Some functions are actually named
    something starting with __ and the normal name is an alias.  */
#if defined (__stub_objc_next_class) || defined (__stub___objc_next_class)
choke me
#else
objc_next_class();
#endif

; return 0; }
---------------------

This is simply testing for cpp DEFINEs.  So, it's possible that
you've applied the patch and recompiled successfully, but are
still getting some mix of header files from an unpatched version.

Or, the patch failed in some respect???? maybe?

glen