hlfs nALFS xml using package users hint?

Declan Moriarty junk_mail at iol.ie
Sun Dec 18 11:30:38 PST 2005


Recently, Somebody Somewhere wrote these words
> Darren McGrandle wrote:
> 
> >Hi all,
> >
> >I'm curious if anyone out there has created the nALFS scripts 
> >for hlfs, and especially if you've incorporated the package 
> >user hint.  I'd rather not re-invent the wheel.
> >
> >Regards,
> >-Darren
> >  
> >
> There is a HLFS profile in svn.  I am having problems with it though. 
> If you would be intrested in testing it Id appricate it. 
> svn://svn.linuxfromscratch.org/ALFS/profiles/HLFS/trunk
> 
I did get this and have a try. 

It's promising, and many packages work well. The package.ent was a
bit skewed, and the guys at lfs-matrix are not up with your
versions. I ended up unzipping tar.gz files and running bzip2 on
them. Here's what I noticed (midway through chapter 5)

Updating to my package.ent, glibc is a version behind, and the
wrong gcc tarball is used. 

Setup errors: get /mnt/lfs and /mnt/hlfs out of the thing, and let
the user be set up. Have $HLFS as a configurable item (One distro
here has already fathered another lfs, which lives on /mnt/lfs).

Rewrite the gcc pages entirely - they are wildly different. 

The hlfs user is skewed. Group lfs. Logging in, you get this
root:/mnt/hlfs/tools/build#su - hlfs
: invalid option
set: usage: set [--abefhkmnptuvxBCHP] [-o option] [arg ...]
: octal number out of range
': not a valid identifier

and nothing works. I ran it down to the .bashrc. I attach mine
from the book, and yours (minus the $LFS variable line). To the
eye they are similar; diff shows the nine lines of each, and
nothing is parsed correctly. Diff -w shows only one line, which I
actually modified. Try a hex editor.

Replacing the .bashrc and running usermod -g hlfs hlfs squared
things. It compiled away (I skipped gcc-pass 1 and did it from the
book by hand). That got me to gcc-pass2, which again I did from
the book, and it failed thusly on the make -k check


FAIL: g++.old-deja/g++.jason/tempinst1.C (test for excess errors)
FAIL: g++.old-deja/g++.jason/template1.C (test for excess errors)
FAIL: g++.old-deja/g++.jason/template10.C (test for excess errors)
FAIL: g++.old-deja/g++.jason/template11.C (test for excess errors)
WARNING: g++.old-deja/g++.jason/template11.C compilation failed to
produce executable
FAIL: g++.old-deja/g++.jason/template12.C (test for excess errors)
FAIL: g++.old-deja/g++.jason/template14.C (test for excess errors)
WARNING: g++.old-deja/g++.jason/template14.C compilation failed to
produce executable
FAIL: g++.old-deja/g++.jason/template15.C (test for excess errors)
WARNING: g++.old-deja/g++.jason/template15.C compilation failed to
produce executable


That stuff was rolling off the page. I feel something small is
wrong. But I only get a few minutes at this here and there, so I
haven't debugged it fully.


-- 

	With best Regards,


	Declan Moriarty.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: files.tar.bz2
Type: application/octet-stream
Size: 4076 bytes
Desc: not available
URL: <http://lists.linuxfromscratch.org/pipermail/hlfs-dev/attachments/20051218/dd93f7fa/attachment.obj>


More information about the hlfs-dev mailing list