[lfs-support] LFS - 5.29 - perl-5.14.2 - possible errata
jasonpsage at jegas.com
jasonpsage at jegas.com
Mon Jan 9 03:19:19 PST 2012
>>On Sun, 2012-01-08 at 18:03 -0700, jasonpsage at jegas.com wrote:
>> I'm doing a few things there - but I do set up that sim link - but I
>> don't quite understand it
>> because for what you describe, my small mind thinks it should be written
>> ln -s $LFS/tools /tools
>> But I'm sure its me just not understanding. Though I did test that
>> syntax and it seems to
>> work how you describe the symlink in 4.2 is supposed to.
>Yes, the syntax for the 'ln' command is kind of weird. But see the man
>page, which lists four different forms that can be used - you're
>thinking of the first one, where the last parameter is the name of the
You're absolutely right - how flexible indeed. Hmm.. I Wasn't aware.
>However, the book is using the third one - when the parameter is the
>name of an existing directory, the link is created in that directory,
>with the same name as whatever it's pointing to. Thus, the command in
>the book does exactly the same as the one you give above.
So I'm having an issue surely - but my sanity "ln" seems somewhat
validated.. Thank you.
>Addressing your original question about the commands in chapter 5
>referring to /tools rather than $LFS/tools, the reason is that we don't
>want anything to end up with $LFS/tools hardcoded in it, as that will
>stop working in chapter 6. Obviously that doesn't matter for a simple
>copy command like the one you asked about, but it does for others
This makes a lot of sense. Thanks for pointing it out for clarity.
>As a result, there should be very few places in the book where commands
>actually refer to $LFS/tools. The only one I know of is the point where
>we set the owner in 4.3, and that's because we want to set the owner of
>the directory, not of the symlink.
This my friend is like one of those WARNINGS Bruce put in the book with
bright colorful warnings! I will watch out for this.
Thank you Simon.
Now, When Bruce pointed out my fstab error <xxx> "How can we know that?"
(he's right) but I didn't know what to set them to.. because my /dev/
empty. So, this "ln" issue, my permissions issue copying the perl files
(even though I did the previous chown to lfs stuff) and my having gone
the book like 6 times now me think of some thing Bruce or Kevlin (forgot
They said its probably being logged in as the wrong person.... somewhere
along the line that is messing things up. And I thought about what the
book says about extracting source code...
It says you should always be lfs user. ALSO, you are to start "FRESH"
each time you extract source code as to not have left over "stuff" from
a previous compile of the same software.
Now in my Scripts - you start as ROOT, make then become LFS, you then
chroot awhile as LFS... then you finish up as root again... Now I'm
wondering if I need to chown to LFS any tar files
I open up while logged in as root for those portions.
But my REAL puzzle is the PERL thing I'm addressing here happens pretty
early in the big picture, and AFTER a very global chown -R lfs /tools
and chown -R lfs /sources
I've COMBED through my script - line at a time (it has references to
each spot in the LFS 7.0 book - so it's like a step by step MATCH lock
in step sort of process)
And that is what troubles me. Human Error is FIRST on my LIST of
probabilities, but I'm starting to wonder if Slackware might have a
couple "quirks" or behaviors of its own (SILLY I know)
that might have JUST enough of a difference that like a house of cards -
it can cause an assumption to bring it all crashing down.
I even did the EFSProgs Check to make sure my partition wasn't made with
a overly enhanced set of file tools and such as the books describes so
I appreciate everyone's responses - but I won't give up or feel "relaxed
a little" until I see the /dev/ folder populate with devices, and I can
get a static or dynamic network connection going...
from there I'm hoping it starts to be much more fun.
and scriptted the entire thing -
More information about the lfs-support