Le Sun, 26 Nov 2006 17:02:15 +0200 Tapio Kelloniemi <[EMAIL PROTECTED]> a
écrit :
> What about this (just one example among others). This command can be found
> from the SVN LFS book in one of the GCC sections:
> patch -Np1 -i ../gcc-4.1.1-specs-1.patch
>
> This command is to be executed in the gcc-4.1.1 directory and the command
> works if, and only if, the patch file is in the directory just above
> this gcc-4.1.1 directory. I never have built an LFS system, where I could
> have used this command as it is in the book. IMHO all commands that often
> need modification, should be marked as such, eg.
> patch -Np1 -i [your source directory]/gcc-4.1.1-specs-1.patch
Exactly what I had in mind, except that it's better explained.
> The LFS layout, where all patches are in ../ is not very practical, since
> in directory listing source directories get buried under the names of package
> and patch files. It also makes it impossible to delete the gcc tree like this:
> # rm -rf gcc-*
Again this is what I ment. Thank you for precising my thoughts.
So, is this such a bad idea ? This can be quite awfull in the BLFS book,
and it is not such a great change. Even if the change is unnoticed by
users, they would be wakened the first time a command doesn't work.
Another advantage of this method : it could permit the book to contain
even the "unpackaging" instructions, which could permit more "automatic"
building by extracting the instructions from the book. And avoid some
troubles about where to unpack extra stuff as doc, extra libs (like in
glibc a while ago) and other things like that.
\bye
--
Nicolas FRANCOIS
http://nicolas.francois.free.fr
A TRUE Klingon programmer does NOT comment his code
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
