[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Octave & Fortran continued
From: |
Michael Goffioul |
Subject: |
Re: Octave & Fortran continued |
Date: |
Sun, 11 Jan 2009 22:51:06 +0000 |
[hit "Send" too early...]
> :) If I found a non-libtool way to build, I used this one. It yields the
> same results much faster and with less overhead.
That's what I did at first. Then I found that the time I spent to tweak
every single build system to make it work with MSVC was higher
than just using libtool (hacked, but at least this hacked libtool can
be re-used for all packages). Moreover, libtool-based build are usually
much better maintained by the package developers than Win32 build.
> You mentioned that you had to re-libtoolize many packages. This way you
> need to touch the sources anyway, and I need to touch the libtool code again
> for shared libgcc suport, so one way or another it is not really
> an "out-of-the-box" build, is it?
Did I say "many"? Sorry then. IIRC, it was only 2 of them.
> But, hey, it's only naming, so what's the fuzz about it. The import library
> must use the lib prefix, otherwise the linker won't find it, and the dll can
> use the same prefix - here it does not matter which prefix you use.
Indeed, it's only naming. More important to me is whether shared
libstdc++ is required for octave or not.
Michael.
- Re: Octave & Fortran continue, (continued)
Re: Octave & Fortran continued, Benjamin Lindner, 2009/01/08
- Re: Octave & Fortran continued, Michael Goffioul, 2009/01/08
- Re: Octave & Fortran continued, John W. Eaton, 2009/01/12
- Re: Octave & Fortran continued, Benjamin Lindner, 2009/01/12
- Re: Octave & Fortran continued, Michael Goffioul, 2009/01/12
- Re: Octave & Fortran continued, Benjamin Lindner, 2009/01/13
- Re: Octave & Fortran continued, Michael Goffioul, 2009/01/13
- Re: Octave & Fortran continued, Christopher Hulbert, 2009/01/13