More detailed profile

Rod Roark rod at
Thu Aug 31 13:01:00 PDT 2000

Gerard, your approach and profile samples are certainly well
considered.  What I'd like to suggest is first having some discussion
at a higher level.

What we're basically talking about is the automation of a large,
complex build.  This is precisely what a Makefile is for, and
accordingly I have been undertaking the creation of a Makefile to
execute the procedures in the LFS book (actually it's already done and
working, but I'm doing some reorganizing of it and adding the capability
to build a few other things).

So anyway aren't you really talking about making a better Make?  There's
certainly value in that far beyond the scope of just LFS.

There's also a need to precisely define the objectives of ALFS.  For
me, it's all about having precise control over what's in my system and
not being dependent on commercial interests, nor the choices and
schedules associated with the major distributions.  Also important is
to *understand* what's installed in my system, and why.

So I think it's vital to do this in a way that is open and
understandable.  If a new language must be invented, it should be a
very quick study and/or have other compelling uses.  I don't personally
care about a GUI front-end, though perhaps others do.

By the way I'll make my Makefile available for everyone's scrutiny in a
couple of days.

-- Rod
Sunset Systems                           Preconfigured Linux Computers                      and Custom Software

More information about the alfs-discuss mailing list