Dan.Osterrath at gmx.de
Wed Jul 14 12:24:27 PDT 2004
Am Mittwoch, 14. Juli 2004 20:36 schrieb Matthias B.:
> > That is the situation now. But why not merge or contribute?
> Because of the complexity. A distro is complex and even the best technical
> writer (and no, we don't have him here, unfortunately) can not write away
> that complexity. So the educational value would suffer because an
> LFS-distro-book would be too complex for most people to be understood in a
> realistic timeframe. Have you ever tried to GROK a distro, to know each
> and every line of each and every boot script, to know all the
> configuration files to all the packages,...?
Of course every project (=distro branch) should have its own maintainer. They
even have their own maintainer(s) at the moment.
It's not meant that some LFS guy should maintain a gentoo branch. The gentoo
guys should share their knowledge with us.
Of course it might be difficult to convince them to do so as this means more
more work for them. But I think if we all share our knowledge at some central
place we could benefit together.
I do not want to unite these multiple projects into one big book. All of them
stay independent and can make it's own decisions.
They just store their build mechanisms, patches, packages in a common
> I find it very disturbing that LFS is already drifting slowly in that
> direction. The boot scripts are the perfect example. From simple, easy to
> understand scripts printed in the book they have transformed into a
> sophisticated package of their own. They include treatment for special
> cases, different configurations and I don't know what else.
As I mentioned the base LFS book will stay as raw as possible but the distros
would become an own branch.
Sometime ago you discussed 'bout merging LFS and BLFS into one book. (Sorry,
didn't follow the thread excactly, so I don't know the result) Imagine we
will merge 'em. Then we got an own distro like book with instructions for
many packages up to a complete desktop or server system. An external distro
with its own branch can add other packages, skip some packages, update
versions, adjust instructions for fitting some special dependencies and so
With that we can merge modifications easily between the projects.
> A lot (most?) of the people currently working on "improving" LFS are
> slowly killing it. I guarantee you that unless this is stopped there will
> be a back-to-the-roots movement in a not-so-distant future. Somebody is
> going to write a new LFS book and it'll start with something like this:
> "Unlike the LFS book, this book focuses on giving you a detailed knowledge
> of a how a Linux system works behind the scenes. For instance, instead of
> giving you complex pre-packaged boot scripts, it will teach you how to
> write your own boot scripts."
Well, I really don't like the LFS bootscript system, too. That's why I wrote
my own. Of course LFS could teach you to write your own system but I think
this goes to far. LFS means "Linux from scratch" and not "how to learn bash
in 15 minutes". ;-)
> I'm really looking forward to reading this new LFS book.
I'm really looking forward to read your "back to the roots LFS" branch. ;-)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
More information about the lfs-dev