2.5.1

KURIYAMA Kazunobu kazunobu.kuriyama at nifty.com
Wed Sep 20 13:30:36 PDT 2000


Serguei Ostrovskii wrote:

> KURIYAMA Kazunobu wrote:
>
> > Gerard Beekmans wrote:
> >
> > > On Wed, 20 Sep 2000, you wrote:
> > > > > Couldn't you include nano in LFS? I personally gets the shivers as soon
> > > > > as
> > > >
> > > > I
> > > >
> > > > > see a VI clone.
> > >
> > > No, I chose for vim and I keep it there. I'm not adding 3 different text
> > > editors, 2 different compilers, 3 different this, etc. That's not what the
> > > book is about. I just give a blueprint and you don't have to follow it.
> > > Nothing is stopping you to pretend you read 'nano' when I write vim. if I now
> > > replace vim with nano people will ask me to replace nano with emacs or pico
> > > or knotepad or ms notepad. I'm not willing to start doing those things. Sorry.
> > >
> > > --
> > > Gerard Beekmans
> > > www.linuxfromscratch.org
> > >
> > > -*- If Linux doesn't have the solution, you have the wrong problem -*-
> >
> > Even if so, I would be happy if our guru takes into consideration adding the
> > following note, or something
> > like that, to a corner of his new book, presumably recommending use of gcc-2.95.2
> > or later.
> > I don't need two different compilers on my LFS2.4, too.
> >
> >       "If your PC uses i810 and you need X401, it would be better to choose
> > egcs-2.91-66 from
> > the beginning because gcc-2.95.2 is able to compile X401 all but an X server
> > driver module for i810.
> > (This failure does not quit the compilation, so you are likely to overlook it
> > until the compilation has finished.)
> > This is due to a bug which prevents the compiler from parsing the kernel header
> > file agpgart.h, and thus
> > the driver module fails to be compiled."
> >
> > In case the latest gcc is free from this bug, please forget this comment.
> >
> > Regards,
> > Kazu
>
> Correct, but for better or worse i810 isn't in every PC. In general there is no need
> in compiling
> any drivers but those for your computer(s).
>
> Sergey Ostrovsky.

My points are as follows:

1) In LFS2.4 Book, gcc is 'highly' recommended. But I couldn't see why it is better
than the egcs. It seems
      that there is no strong reason to choose it other than it was tested for building
LFSs.

      It is better for us to recognize BY EVIDENCE that there is a source which cannot
be complied with
      the gcc. No one knows when he or she comes accross such a stuff. At lease,
'highly' is redundant for me
      and  not necessarilly useful for all.

2)  When seeing the menuconfig of the latest kernel-2.2.4-test8, you will find there
are 5 or 6 items for
       i810 proper. I don't know whether i810-type motherboard becoming major or not.
But kernel developers
        are paying much attention to it. This shows that they think the number of PCs
having i810-like-
          motherboard is increasing.  Even the kernel is forced to have compliance with
i810, why does LFS
      ignore it? I wish the coming version of LFS should pay the same attetion to it to
enhance its potential.
     Also, considering the ALFS project,  I hope this kind of note is useful.

3) After all, I think it is not good for LFS to have restriction on the choise of
hardware stronger than the 'normal
      system''s distribution has.

Do you think this is a local matter?

Regards,
Kazu






More information about the lfs-dev mailing list