Re: Schedule for GNU make 3.81

[Alessandro Vesely]

Hi!

address@hidden wrote:
> [...]
> So, please try this beta with your favorite platform and makefiles.  If
> you have been sitting on bugs, now is the time to report them.  If you
> have patches, now is the time to get them confirmed and sent in.

4 points:


1. I know this source doesn't belong to make, but if at all possible,
I would include the fix for case insensitive $(wildcard). I repeat
the patch below:

--- glob\glob.original.c        Sat Jan 22 06:43:04 2000
+++ glob\glob.c Wed Feb 16 18:56:40 2005
@@ -1303,7 +1303,7 @@
             {
               int fnm_flags = ((!(flags & GLOB_PERIOD) ? FNM_PERIOD : 0)
                                | ((flags & GLOB_NOESCAPE) ? FNM_NOESCAPE : 0)
-#if defined _AMIGA || defined VMS
+#if defined _AMIGA || defined VMS || defined HAVE_CASE_INSENSITIVE_FS
                                    | FNM_CASEFOLD
 #endif
                                    );


2. The Copyright notice is outdated. I'd patch it as

--- main.original.c     Wed Feb 16 06:03:42 2005
+++ main.c      Wed Feb 16 19:20:48 2005
@@ -2803,8 +2803,8 @@
      year, and none of the rest of it should be translated (including the
      word "Copyright", so it hardly seems worth it.  */
 
-  printf ("%sGNU Make %s\n\
-%sCopyright (C) 2003  Free Software Foundation, Inc.\n",
+  printf ("%sGNU Make %s\n"
+          "%sCopyright (C) 1995-2005  Free Software Foundation, Inc.\n",
           precede, version_string, precede);
 
   printf (_("%sThis is free software; see the source for copying conditions.\n\

BTW, is that correct? I mean, is 3.81 coming on GNU make's decennial?


3. $(shell) doesn't work in win98 under command.com. I investigated it a bit and
found out it is because of the `Path' rather than `PATH' environment variable.
This doesn't happen for build commands, but at that point I ran out of spare 
time
and stopped. Anybody has some pointers on the `Path' vs. `PATH' question?

As a side effect of that investigation I posted
http://savannah.gnu.org/patch/?func=detailitem&item_id=3747
I don't think it's worth the hassle of testing it, unless someone
has unknown problems with CreateProcess or CreatePipe in W32.


4. Also, I ran the gmsl-tests (see <http://gmsl.sf.net/>.) After a few 
successful
tests gmake catches an Interrupt/Exception. It looks like resource consumption,
and happens with both command and mingw. I haven't had time to investigate that
either :-(