Chapter 3. After LFS ... Issues - Creating a custom bootdisk

Bill's LFS Login lfsbill at
Sat Jun 14 14:52:47 PDT 2003

On Sat, 14 Jun 2003, Larry wrote:

> On Sat, Jun 14, 2003 at 01:39:01PM -0400, Bill's LFS Login wrote:
> >
> > Like I said at the start of the thread "... doomed at the outset".
> > 'Nuff said.

> I don't see it that way. I never think in terms of what should be where.
> BLFS will always fill in any gaps to producing  a well rounded system,
> dependant on the users very own definition of well rounded.  While we
> may never get that fulfilled, the least we can do is give the users
> things we know will be important to them whether they know it or not.

In case you missed the start of the thread, I had asked if anybody
thought it was worth suggesting that a basic recovery floppy be included
and built in the LFS book because it seemed to me to be a *basic* part
of any end-of-installation process. I also said I didn't want to start a
thread that was doomed from the outset. It was asking if anubody thought
it was worth bringing it up on the LFS dev list, not BLFS. The responses
were one thought it was a good idea and two others thought that LFS
editors wouldn't or shouldn't have it there. From there there was some
wandering, as is fairly common. :)

After some thread wandering, it became apparent that we had addressed
that question enough. So that is what the "... doomed at the outset.
'Nuff said." was about. It was not a ref to anything BLFS.

> In this instance, I have not been able to produce a boot disk from our
> current instructions.  The "/dev" directory is too vague and the cop out

Let me know, public or private, and I will be happy to help.

> is to install all (which does not fit), and if it really, really needs
> devfs, than we need a 'kernel building section' and outside reading
> material for devfs, because this is the users first exposure to it.

Devfs is *not* needed. I made mine without it w/o problems. But I have
been exposed to this before, so the pitfalls and solutions were
apparent to me before I did any scripting. Also, I interpreted the
instructions as "other things you need that will fit" instead of "all".
So my effort was easier than yours. I immediately recognized that we had
no way to do *basic* recovery ops w/o a couple FS related make/repair
things. That lead to suggestions to augment the current book contents.

In that same thread, I asked if we might want to include a few lines, or
a small section, for non-devfs because LFS does not include devfs. And
if the editors thought it worthwhile, I volunteered to cobble together
some verbiage and make a patch to the XML files for a first review.
Offer still stands. But I don't do those sorts of things w/o a go-ahead
from someone of responsibility, for the obvious reasons.

> Then we need to be assertive that having a boot disk is important enough
> to learn these additional areas.

One outcome of the wandering thread was that Csaba (I think ?) suggested
that we might suggest the LFS folks put a line or two in the ending
section mentioning that they should get to BLFS and make a recovery
diskette. That seems a reasonable thing to me.

> Bill, I haven't had time to read you hint yet, nor do I have time to fix
> the current write-up, but it will be addressed.  If there are any more
> suggestions, attatch to this thread, then if someone take them and rolls
> them in, fine, otherwise, I will get to it as soon as I can.

My hint has nothing to do with this issue. Well, I guess it is slightly
related in that both have the potential of offering recovery, but that's
about as close as they get to each other.

As to suggestions, only my originals related to BLFS specifically:
possibility of some non-devfs verbiage (or section) and augmenting the
"copy anything else..." phrasing with a couple of additions, like adding
a mk*fs and *fsck for your installation (we would need to make a dynam
e2fsck to get it to fit).

And, as mentioned, I am willing to contribute some effort towards this.

> Larry

Hope that clarifies things. I got the distinct impression that my
thoughts were misinterpreted or I was bitching about BLFS. That wasn't
happening. If you review from the start of the thread, I think you'll
see the "flavor" of it.

Bill Maltby
lfsbill at
Unsubscribe: send email to listar at
and put 'unsubscribe blfs-dev' in the subject header of the message

More information about the blfs-dev mailing list