SCO port update - what now?
daniel at fooishbar.org
Fri Jun 10 10:15:58 PDT 2005
On Fri, Jun 10, 2005 at 10:03:06AM -0700, Kean Johnston wrote:
> > * why was the AssembleObject change made?
> That one is becuase older make's (in fact, so far only GNU make seems
> to support it) do not allow $* in normal rules, only in inference rules.
> So you can use $* in a .c.o: rule but not in a normal rule.
Ah, right, I remember seeing this one before -- sorry.
> > * why was the zlib change made?
> Becuase not all (in fact many) compilers support #error, but not
> #warning. Thats a GCC-ism. Unless you are refering to the change in
> lbx_zlib_io.c, in which case, OSR5 has writev() and doesn't need lbx
> to provide it.
Hm, agh. Do you know which compilers these are?
> > * the glxinfo change is wrong.
> Which one? There was one to an Imakefile, to not build glxinfo if
> GLU wasn't built, and one to use SimpleCplusplusProgramTarget instead
> of SimpleProgramTarget. I dont see how either of those is wrong, so
> can you tell me why they are?
glxinfo is a C program, not a C++ program. The only problem arises with
broken linkers that do not properly link in libstdc++ to dependent
libraries, but this is not glxinfo or imake's problem.
> > * why was the xdm/auth.c change made?
> Becuase it was completely duplicating code found in Xlibint, and doing
> a bad job of it. _XGetHostname() had all the same smarts, but xdm/auth.c
> and _XGetHostname were out of sync. So I reeuced it to call that other
> function instead of maintaining the same bits in two places.
Just curious as to what the rationale was behind most of these -- it's
really hard to try to make sense of a massive patch covering lots and
lots of areas with only 'SCO port update' covering the entire thing.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 189 bytes
Desc: Digital signature
More information about the xorg