current testing state of glibc 2.3.4 and 2.4 headers

[Terry Chan]

Upon examination of the glibc-2.3.4 lunar module and the 2.3.4 source code,
I'm not sure that commenting out the glibc-2.3.2-ctype-compat.patch is such a 
great idea.  The patch probably needs some help in being applied, but the glibc-2.3.4
code does NOT appear to provide the patch's functionality.  You will find out very
shortly when you try to get your lunar box to compile a static binary, like tar-static,
bash_static, or sash.  If glibc-2.3.4 does allow you to correctly compile a static
binary and run older statically linked binaries, THEN the ctype-compat.patch is really
no longer necessary.

Terry Chan
--------------------------------------------------------------
On Sat, Jan 29, 2005 at 10:30:43AM +0000, Jaime Buffery wrote:
> Hello fellow devs,
> 
> After some nice testing with 2.6.X headers, and a dead sandbox that Ratler on 
> his own without any of my help b0rked (:þ), I have done some testing with 2.4.X 
> headers myself in chroot. What I have done is this:
>   * cp a 1.4.0 iso 
>   * chroot to it
>   * update theedge and moonbase a couple of times to make sure I don't have evil 
> problems :P
>   * put the glibc-2.3.4 module I handed Ratler in zlocal and set 
> ZLOCAL_OVERRIDES ( http://www.lunar-linux.org/~nestu/glibc-2.3.4-module.tar.bz2, 
> comment out the --enable-kernel=2.6.0 part in BUILD, don't forget, if you try).
>   * lin glibc succesfully
>   * lin gcc, tar segfaulting so, lin tar, lin gcc 
>   * lin binutils
>   * lin ocreutils
>   * lin installwatch
> Except for tar, all went fine. Tonight, when I arrive back, I will install the 
> latest headers, and start from glibc again. This is necessary, since sometimes 
> with this change installwatch doesn't work properly, and also we need to double 
> check everything is fine.
> 
>   So, basically, right now, without full test I must admit, this is thumbs up, 
> from my point of view at least. But more on that later, I'm off to work!
> 
> Ciao!
> Jaime ;)))