Re: [AD] Release Candidate 2

["Eric Botcazou" <ebotcazou@xxxxxxxxxx>]

> Just these files this time round:
>
>   http://alleg.sf.net/future/rc2/all400.zip             (2752357 bytes)

Extensively tested under DJGPP, MinGW32 and Borland C++.

> I like alleg40.dll better, but I don't feel like mucking about with so
> many makefiles. If someone sends a patch (completely tested) then okay.

Attached (use VERSION instead of SHORT_VERSION). The tests under Windows
were conducted after the patch had been applied. But I can't test with MSVC.

And the two others dlls were named ald40.dll and alp40.dll, now they will
really be named alld40.dll and allp40.dll .

It seems that you didn't commit a change in fixver.sh, I think you should at
least for the 4.0 branch. You will have to rerun fixver.sh to update the DLL
info sheet.

> I removed the line, and put makedoc.c back into the allegro-tools RPM.

Here's what BC++ says about makedoc.c:
Warning W8004 docs\makedoc.c 2040: 'title' is assigned a value that is never
used in function write_rtf
Warning W8004 docs\makedoc.c 2019: 'pardebt' is assigned a value that is
never used in function write_rtf

> I committed a different fix.

I'm still not very comfortable with the name of the compiler: for example,
you may want to use kgcc under Mandrake 8.0 to compile Allegro. According to
the autoconf info manual, you can set $CC before running configure and, in
this case, your test is cheated.

The robust way (still according to the same manual) of testing for the
presence of GCC is:
    if test $GCC = yes

--
Eric Botcazou
ebotcazou@xxxxxxxxxx

Attachment: dll_name.diff
Description: Binary data

Attachment: aclocal.diff
Description: Binary data