Cross LFS and UTF-8

Joe Ciccone jciccone at gmail.com
Tue Jan 17 15:59:39 PST 2006


Jim Gifford wrote:

>
> The question also has come up, how we should implement UTF-8, make it
> a standard part of the build, or add notes into the packages that
> require build changes. An example of what I propose is similar to the
> Shadow page with cracklib.
> http://www.linuxfromscratch.org/clfs/view/cross-lfs/x86/final-system/shadow.html.
> As the CLFS community we ask your opinion on how we should implement
> these changes.


    The notes idea makes the most sense to me. The builder can choose
weather they want to build a UTF-8 machine. Some PPC and some older x86
machines can have very low memory. You don't want to waste that memory
if you don't need to. Even on a high memory machine, you shouldn't use
more then you need. Personal preference is another issue, I don't want
anything todo with UTF-8. My builds are UTF-8less because I don't
personaly need or want it. Notes on the pages give the users the option
to include UTF-8 support in their builds. So, if someone that needs
UTF-8 to use their system in their native language, they can follow the
notes to get UTF-8 support but, if someone like me doesn't need UTF-8
support, I am not forced to include it. Sure, if someone that knows what
the build process is they can exclude all of the UTF-8 stuff, But what
about a new person, that knows nothing about LFS yet, They are forced
down one path, the UTF-8 path.

Another option would be to maintain a Non-UTF8 book and a UTF-8 book
but, who wants to do that?

I'm not trying to put down all of Alex's work, or throw it away, I just
think that there is probably a better way to handle UTF-8 support then
forcing every user to have it.



More information about the cross-lfs mailing list