Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 12 Apr 1997 12:05:56 -0700 (MST)
From:      Terry Lambert <terry@lambert.org>
To:        dk+@ua.net
Cc:        rssh@cki.ipri.kiev.ua, freebsd-hackers@freebsd.org
Subject:   Re: on the subject of changes to -RELEASEs...
Message-ID:  <199704121905.MAA15540@phaeton.artisoft.com>
In-Reply-To: <199704120736.AAA11209@dog.farm.org> from "Dmitry Kohmanyuk" at Apr 12, 97 00:36:13 am

next in thread | previous in thread | raw e-mail | index | archive | help
> > I think, exists some standarts.
> >  SIUC, POSIX
> 
> s/SIUC/SVID/ ? That's System V Interface Definition...  I don't 
> think it defines and names within file system...

SIUC is the X/Open single UNIX standard...


> I don't know about POSIX much, but judging from my experience,
> I can tell you:
> 
> There _cannot_ be a standard under which Solaris, SunOS, AIX, HP-UX, 
> SCO, Digital Unix, FreeBSD and Linux all comply.   The set intersection
> is _empty_.

There is POSIX, and the only uncertified member of that set is currently
FreeBSD.

Now as to whether POSIX is a useful standard, that's another argument;
I'd say that as long as you can't turn off non-POSIX features in the
OS build environment to make sure the software doesn't depend on some
non-standardized extension, it's pretty damn useless.



> The idea of having read-only root filesystem and separate
> configuration information has its merit, but some files have to
> be there or you loose look-and-feel compatibility with whatever
> Unices we still have it.

Actually, we don't call that "look-and-feel" anymore; we now call
it "learning curve".


> How about absolutely basic /etc which re-mounts /etc partion on top
> of itself and re-executes boot scripts?   I don't think its a good
> thing for a standard configuration, though.

The most useful overlay mount has an fs with the following structure:

	/
	|
	`--- /dev

Getting it's '/' part overlay mounted, leaving it's '/dev' part exposed
by the lack of a 'dev' in the root of the overlay.


This gets us all the way to a single user shell during an OS port
without implementing any machine dependent device code other than
the bottom end (if it's abstracted right) of the console driver.

Whee!


					Regards,
					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199704121905.MAA15540>