[Tfug] Alan Cox: "I've had enough"--what else is new?

[Bexley Hall]

Hi Marco,

> I don't know you personally, I
> came to a couple of meeting (that seemed to me a meeting of
> friends more than a Linux user group), but now I don't
> associate a name with a person. I've been monitoring
> this mailing list for a while, and I can say that the posts
> that catch more attention are the most useless, windows
> against mac against linux against BSD,  who cares if this
> guy prefer hackintosh or think linux won't go anywhere,
> this discussion won't change anything, seem more a
> discussion in a pub about soccer teams than a discussion to
> make a system and an idea grow and improve.

<grin>  And why do folks spend countless hours arguing
about which soccer/football/baseball/etc. team is "better"?
Especially when they change players routinely?

[Hint: that's why these same sort of posts see just as
much activity!  It's pretty hard to argue about whether
2+3=5 or not (though some can try :> )]
 
As to "[making] a system and an idea grow and improve"...
I suspect no one here has commit privileges and few (if any?)
regularly contribute bugfixes back to the development teams.
So, the sports analogy still holds:  you can argue about
whether Player X should be traded to Team Y for Player Z
and 2 dozen tacos...  but all the wind you expel in that argument
will do exactly NOTHING towards changing the composition of
the team(s) *or* their relative performance(s)!  :-/

The saddest thing about the 386BSD/FreeBSD/Linux/NetBSD/OpenBSD
issue is all of the wasted manpower that has gone into reinventing
a 35 year old OS.  I fault Linux especially for an obsession
with "performance".  It would be hilarious to try backporting all
of this code to a PDP-8 and seeing just how well it runs against
some *ancient* version of "UNIX" (Tm Reg).  I.e., take away all the
hardware advances (to which none of the developers can claim credit)
and you're probably no better off than The Way Things Were.

Why are folks commenting about "system hangs" (freezes)?  You
mean after all this time and experience, folks haven't learned
how to make a *reliable* system?  ("Oh, its just a bug in _____")
Why can't the system be resilient enough to disallow these
sorts of problems?  Why do they *keep* cropping up despite
patches and re-patches, etc.?  (it's like MS and their stewpit
"security patches" -- always the same sort of bugs that hackers
keep exploiting!  Why don't they figure out how to protect the
system from their own ineptitude as developers??)

I started out in the FOSS world just after the 386BSD release
(ca. 1992?) and quickly shifted focus to NetBSD (1993-ish?).
Progress in providing a "stable" UN*X platform (for the tools
I needed to run) was too slow in the NetBSD camp -- I think
hindered by their philosophy of being *really* portable in
not just their *codebase* but, also, the actual structure of
the system itself -- so I jumped ship and moved to FreeBSD in
~1994 or 1995 (?).

FreeBSD eventually seemed to decide they were Linux-wannabes
and shifted gears towards quantity instead of quality (this
was especially true of the packages system -- the *BSD's deal
with many of the applications separately than the OS and its
core functionality).  So, I switched back to NetBSD (which had
made big advances in support and stability by then) and that's
where I stay, today (if it works, why screw with it?).

None of these camps ever came out with truly unique "positionings".
Linux wanted to replace MS; FreeBSD wanted to be Linux; NetBSD
wanted to run on *everything*; etc.  Only OpenBSD made an
attempt to position itself somewhat "nobly" -- claiming to
tout "security" as paramount.

But, none of them truly "Just Work".

It's like Athena... "wow, look at this nice new feature..."
"Yeah, but the system is broke, *again*..."

Imagine a salesman trying to pitch a new hammer to a carpenter:
"It has a built in spirit level, phillips *and* R&P screwdrivers,
nail file, iPod, roadside flare *and* can run off solar power"

"Yeah, but does it DRIVE NAILS???"

("Of course!  Well... *most* of the time.  *But*, we're working
on that problem and plan to have a fix released any day now...")

C'mon... even *plumbing* has made advances in the past 35 years!