help-hurd
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Hurd FS hierarchy (was Re: LD_LIBRARY_PATH troubles)


From: Guy Bormann
Subject: Re: Hurd FS hierarchy (was Re: LD_LIBRARY_PATH troubles)
Date: Tue, 19 Mar 2002 01:18:46 +0100

> On Tue, Mar 19, 2002 at 12:05:02AM +0100, Marcus Brinkmann wrote:
> > On Mon, Mar 18, 2002 at 11:35:36PM +0100, Jeroen Dekkers wrote:
> > > I think the whole /sbin directory is old unix-craft like /usr.
> >
> > We will not link /sbin to /bin.
> 
> It's nice that you give a good argumentation. I'm really convinced.
Go to bed Jeroen! Your responses usually get consistently nastier or
more
defensive the later it gets.

I know there is a time lag with our American friends so it's not a fair
discussion
:-) but this should not turn into a flame war over what after all is
just a draft
version of an annex(!) to a standard. And the standard is mainly meant
for
system packagers to promote a uniform directory structure for Unix,
Unix-like and
maybe even "just" POSIX-compliant systems (BTW, what does the POSIX
standard say
about this?).

So, it is my humble opinion that in order for the Hurd to make its drive
into the
"market", it should do it on the basis of its POSIX emulation layer
(including
file system appearance). Since it is NOT Unix(like), any attempt to make
it fit
with "common practice" in the Unix world is only downplaying its
capabilities.
Conversely, trying to force its characteristics on the *nixen through
important
standard changes will fail (because they are mostly irrelevant to those
"others") and
making GNU/Hurd just an irrelevant annex.
  The FHS is not the appropriate forum to promote the "superior"
capabilities/characteristics of the Hurd. It is very unlikely that
ordinary users
will read it, anyway and distributing packagers don't have to be
convinced anymore.
But playing it sneakely it could speedup adoption by other packagers
(than Debian)
and increase diffusion. The evangilism and education about its unusual
(seen from the
point of view of a Unix(like) user) features can then happen gradually
through system
documentation and dedicated websites.
  Therefore it is better to spend your energies in finding the best
means/ways to map
the FHS to the POSIX layer and maybe request some changes in case this
requires
ugly hacks or leads to confusing inconsistencies. People who don't like
the FHS
layout after an install can hack around as much as they like, exploiting
extremely
nonUnix behaviour/capabilities. Hey, you could even distribute the
necessary 
scripts/docs/rationales/total-Hurd-without-any-reference-to-POSIX-distribution
separately. It is just not realistic to force it on packagers
specialized in f.i.
distributing Linux, especially not if you are targeting soon-to-be
former Linux
users (or any other *nix user, for that matter). You just don't want to
punish them
for being converted to the Hurd or even for just trying it out. We're
not the Borg!
(In fact, I'm certainly not since I didn't find the time to try it out
yet but I'm
already convinced...:-)

Just my 0.02 eurocents,

Guy Bormann

--------------------------------------------+---------------------------------------
Dr. Guy Bormann                             | email : Guy dot Bormann at
ua dot ac.be
Research Assistant                          | smail : Building N, Room
1.06
Theoretical Physics of Condensed Matter     o         Universiteitsplein
1
Dept. Natuurkunde                           |         2610 Wilrijk
Universiteit Antwerpen                      |         Belgium



reply via email to

[Prev in Thread] Current Thread [Next in Thread]