On Wed, Mar 19, at 11:52 J. Greenlees wrote:
[...]
> Both sections need to be made as simple and clear as possible, with the
> absolute minimum required for a functional system to be base system
> standard.
While I understand your points and I can't but I agree with you, I am
not sure if we can change the rules.
Oh I am sure; we can't!
And we can't, because we don't even care or try to be LSB compliant.
How then, can we influence the development or how can we push it in
our direction?
No, I don't permit to myself to criticize something, when I don't
participate in that project.
Unless the project:
- doesn't give me the right to participate
- it makes hard to participate
- doesn't care about my participation or it doesn't respect my
contributions
Then I can criticize it.
In my opinion, LFS has to care for LSB/FHS compliance, if not for other
reason just for politics :), although conforming with standards is my
personal priority. As I already stated before, if we don't like the
standards, we are trying to change them, to fit the needs of today but
thinking also for tomorrow. And in this last regard politics make sense,
because I don't think is a good idea to leave the development in
the hands of various vendors or to inconsistent behaviors,
see (in comment #1) [1] for example.
And yes, he is the same Ulrich Drepper who is in Austin Group, yet his
software is not build-able by a POSIX compliant shell.
Speaking about inconsistencies, we have our share too, and in personal
level or in the development of our books and I am speaking most for BLFS.
Example? The common usage of /usr. Convenient but fundamental broken.
>From all the BLFS packages the half or even more, (they) really bellongs
to /usr/local hierarchy.
I remember in my early (unix) days in 2004, I spend some time with one
of the BSD's (free), where I learn to love the excellent and lean system
structure and for a while I followed their way to organize the installed
software in Linux, but then I noticed that I couldn't contribute a single
bit back in BLFS if I had it that way, where the proffered way is to put
all (even crappy software) in the same hierarchy with important
executables or crucial/sensitive libraries.
I know that is a bit controversial matter (where a package really belongs)
and I don't want to open that issue now, I am just saying that we don't
really care to follow FHS (for instance), so we can't really have an
opinion. And please don't say me that we do (care), because if we do/did,
this is an ongoing matter, but in the years I am in this project I never
noticed such a discussion.
In fact "everyone" in Linux land can do it in her/his way and some call
it a feature. Well, while I still care about "everyone", yet in LFS we
have to follow standards. And once more, If the standard is broken, we can
document it and we can choose to do it in our way or their way.
But at least we can then raise a voice to the LSB group and with expectations
to change it.
This is our share in my humble opinion, so don't shoot the LSB, even if
you got it right and I am sharing your sentiments completely, we are in
the same boat.
But as Bruce said and I agree, we are a part of the Linux world, how
important? (part) is debatable, but it's not dependable by others but
ourselves.
1. http://sourceware.org/bugzilla/show_bug.cgi?id=3266
--
http://wiki.linuxfromscratch.org/blfs/wiki/Hacking
--
http://linuxfromscratch.org/mailman/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page
