2012/8/8 Jeff Hobbs <jeffh@activestate.com>
Is there any concern with having the 'tk-cocoa-8-5-backport' merged into
'core-8-5-branch'?  This will only effect OS X.  Both Apple and
ActiveState build from the Cocoa port, and aside from some tricky edge
cases with event handling, it has been considered the better variant for
years now.

I would be carefull with that. Comparing the core-8-5-branch with
the "tk-cocoa-8-5-backport" branch there are some things I note:

- In core-8-5-branch, all RCS ID's were removed, in tk-cocoa-8-5-backport
  they are all back (e.g. in tk.decls, tkInt.decls, )
Differences like (macosx/README) :
-  (making relinking unnecessary was added in 8.4.2)
+  (making relinking unnecessary was added with 8.4.2)
or (unix/tcl.m4)
-        AC_MSG_ERROR([Can't find Tcl configuration definitions. Use --with-tcl to specify a directory containing tclConfig.sh])
+        AC_MSG_WARN([Can't find Tcl configuration definitions])
+        exit 0
or (unix/configure.in):
-    AC_CHECK_TOOL(AR, ar)
+dnl FIXME: Replace AC_CHECK_PROG with AC_CHECK_TOOL once cross compiling is fixed.
+dnl AC_CHECK_TOOL(AR, ar)
+    AC_CHECK_PROG(AR, ar, ar)
+    AS_IF([test "${AR}" = ""], [
+    AC_MSG_ERROR([Required archive tool 'ar' not found on PATH.])
+    ])

Those show that some fixes done in core-8-5-branch
were not done in tk-cocoa-8-5-backport'. What more
would we lose? I guess that the backport branch
was maintained separate for a while (CVS branch?)
and some files where simply copied over....

So, please be carefull with that.

          Jan Nijtmans