On Wed, 15 Mar 2000, der Mouse wrote:

# Recently (= approximately this past year), it seems, someone has
# decided that a larger user base is a sufficiently good thing to be
# worth dropping the niche people for, and this has resulted in things
# like the package system and /etc/rc.d, things designed to make life
# easy for the "I don't know computers and don't want to have to" crowd,
# the people whose idea of complicated system administration is having to
# actually type pkg_add at a shell prompt instead of clicking in a GUI.

You know, actually, I consider myself to be something of an OS hacker/
hardcore admin type.  (You should know this -- I sent you my version
of an interactive archiver, inefficient as it was!)

But I *like* the pkg system.  It saves me *time*.  Yes, I know how to
configure things, and yes I know how to just do format/boot/untar/
"edit /etc/*".  In most cases it's faster.

But after my system is skrogged to the point that all I can do is nod,
smile, wave and upgrade, the ability to use the pkg system to restore my
system to a stable state (for prog in $list; do (cd $prog; make && make
install); done) and then just walk away from it to come back after
work, well, that's just cool.

I'm not entirely sure of the rc.d business, just yet, actually.
If it looks too much like sysV or Linux, well, yes, I'm going to
be somewhat bitter about it.

On the other hand, as much as I am used to the thought of 'ps, kill,
/sbin/whatever -opts args', and I still do that, I have written more
than one instantiation of a script which could be used to stop/start
individual processes.  namedc doesn't always cut it, for example.
If the scripts are there, that actually can make life a bit easier
(I have to rewrite it now because the one copy I had was zorched when
 I had to reinstall...)

I'm one of those OS/admin types who is actually looking forward
to the system getting pkgized.  If I had the time, I'd blow it out
that way myself.

# 
# Yes, there are hell of a lot more of them than there are of the
# hardcore hacker types.
# 
# But they're already very well served by Windows, Linux, etc.  Perhaps
# FreeBSD as well; they seem to have gone farther towards catering
# towards that crowd than we have, from what little I know of them.

But NetBSD will run on more platforms IN SYNC than will Windows, Linux,
FreeBSD.

# I don't want to see NetBSD trying to compete with Linux and Windows on
# their own ground, both because I believe it will lose and because it
# will mean that NetBSD will then no longer suit the niche I am part of.
# Yet that seems to be where it's headed.

Don't count it out yet -- if there's options to be had, the niche will
probably invent them (/usr/src/pkg/niche?).

# > 	rc.conf is no more 'the BSD way', than splitting up rc into
# > 	individual subsystem controlling files, but they both intended
# > 	to ease administration.
# 
# And they do...for some people.  The trouble is, *which* people?

For people who _don't_ have that much time on their hands, for example.
I think I'm getting this.  I don't think it's an intentional slight toward
the hackers as much as it's making it easier to start the system up.

In defense, though:  rc.conf strikes me as more of what _would have been_
the BSD approach.

# Just as "BSD way" != "CSRG products", "SysV way" != "SVR4 spec".  The
# rc.d scheme, as I understand it, has the same basic problems that real
# SysV startup scripts do: they make human understanding and manipulation
# of the startup sequence significantly harder, for the sake of making
# mechanical ditto easier.

Uh, not all that much, come to think of it, save that the init levels
(please don't tell me we're going _there_!) really take one through
some twisty passages, depending on how init is set up to run (i.e.
level 2 runs level 2 and then level 3 on some machines; on others
it runs levels S, 2 and 3, and on yet others it just runs level 3).

But on a single level, it's not that hard to follow...

				--*greywolf;
--
BSD, stupid.