Re: tlug: PJE

["Stephen J. Turnbull" <turnbull@example.com>]

>>>>> "jb" == Jonathan Byrne <- 3Web <jq@example.com>> writes:

    jb> On Fri, 21 Aug 1998, Christopher Sekiya wrote:
    >> On Fri, 21 Aug 1998, Jonathan Byrne - 3Web wrote:
    >> 
    >>> The really, over-archingly great thing is that a Linux system
    >>> can be built (and someday will be) where installing a package
    >>> is as simple as dragging it from point A to point B.
    >> I strongly disagree.  This is EXACTLY what I was railing
    >> AGAINST.
    >> 
    >> Such a thing removes the clue factor.

Well, looking at the quotes above (all those stupid ">"), I can
certainly agree that clued-in users with flexible software is a
GoodThang[tm].

    jb> I know what you're railing against; I just completely
    jb> disagree.  If software and hardware systems were adequately
    jb> capapble, they would need essentially no intervention from the
    jb> user to function properly.  The fact that you need so much of
    jb> a clue to do so much stuff points up the fact that systems are
    jb> still extremely immature.

Ah, the myth of the Dave Hayes Artificially Intelligent Spam-Fighting
'Bot-Baiting Auto-Flaming Newsreader raises its beautiful visage once
again.  This is just not true for the systems that Chris and I are
interested in; your focus is somewhat different, much more OA-
oriented, but today OA is more and more about communication, and that
simply requires more sophistication than a machine can manage.

Did you see the report on the Paris RoboCup for soccer-playing robots
last night?  The CMU team won with a single multithreaded program
running multiple robots in the mini category.  It was exciting to
watch.  Handai won the large communicating CPUs category.  It was
boring.  Handai won because it managed to have a robot actually in
motion 25% of the time; the Utsunomiya team got hammered because it's
robots were basically stationary 100% of the time.  What's the
difference?  'bots can't do communication yet.

Yet I don't agree with Chris that StarOffice is the start of a
slippery slope that leads to Linux-based spamming by clueless scum
with a Linux port of "The Stealth Mailer", with an intermediate stop
at double-CDROM-drive workstations so that when one auto-extruding
cupholder breaks, you don't drop your coffee mug.

But it _is_ true that giving Linux to the clueless is not only going
to cause them problems, but also going to cause the 'Net problems.
Viz MS Office Excrement, only raise it to the nth power.  I don't
think you're clueless, Jonathan, but focusing on serving the clueless
does have an inherent danger that interactions with other members of
society will be ignored in favor of a slick UI.  That was your basic
attitude at first, in case you don't remember.

    >> We're moving away from Open Source, believe it or not.  Sure,
    >> the base system will still have source available.

    jb> I don't think so.  The things that are open will remain open,
    jb> and new Open Source material will not cease to be produced.
    jb> Regular commercial software is an augment, and a good one, to
    jb> Open Source.  It is rarely in competition with it.  After all,
    jb> imagine how good a proprietary product would have to be to
    jb> step in and take the market away from an Open Source product
    jb> that is not only Open but free; it would have to be very, very
    jb> good indeed.  If a proprietary product that good comes along,
    jb> then the Open Source authors, too, must innovate if they want
    jb> to maintain market share.  That's competition.  That's good.
    jb> That's the market system.  That's good.

Yeah, what he said.  Chris?

    >>> That should be trivial.  The fact that it's not is an
    >>> indictment of the current state of the system.

    >> It's an indictment of the user base that they (not WE, per
    >> Stephen's definition) are targetting.

    jb> No, it's not.  It's an indictment of the system (mostly).

It's both. It should be fixed; my time is MORE valuable than a luser's
(sorry, but it is true) although the luser will undoubtedly spend more
time fixing and working around.  However, the fact that things are
being targetted at people who can't and won't understand is dangerous.

You don't just hand out licenses to drive tractor-trailer rigs, and a
single Linux box is a potential instrument of mass destruction.  (I am
not equating lives with spam or the Internet Worm, although today the
Worm could conceivably kill.  However, once you factor out the value
of the entity at risk, Linux or any computer system multiplies the
danger in ways that a mechanical device cannot.)

    >> I maintain that users should UNDERSTAND how the system works.
    >> Maybe that's too much to ask.

    jb> My experience providing user support indicates that it is
    jb> indeed too much to ask :-)

I think you're underrating your users.  "Lusers" are a mostly mythical
race, despite my own frequent references to them.  There are far too
many of the willfully ignorant, though.

    jb> I'm curious about something, though.  I wonder what it is
    jb> about computers that inpires people to say "you should really
    jb> know how it works if you're going to use it."  (Yes, I'm
    jb> guilty of that, too; maybe not as guilty as you, but pretty
    jb> guilty :-) .) We do not usually make similar demands of the
    jb> users of other machines, even fairly complex ones.  When was
    jb> the last time you told someone they should thoroughly
    jb> understand how their microwave oven works, or they shouldn't
    jb> heat food in it?  Ditto for storing food in their
    jb> refrigerator.  How about cars?  How much should a person know
    jb> about a car?

It's hard to kill more than about 200 people using one car.  It's
trivial to spam several million using one PC.

    jb> A computer may never be quite as simple to use as a microwave
    jb> oven, but the microwave makes a nice ease-of-use model.

Shiver.  The microwave makes a nice ease-of-use model for nuclear
weapons from the point of view of the NSA and Omar Bin Raddin.  I'm
glad it's only nice, not feasible.

-- 
University of Tsukuba                Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences        Tel/fax: +1 (298) 53-5091
--------------------------------------------------------------
Next Nomikai: 18 September, 19:30 Tengu TokyoEkiMae 03-3275-3691
Next Meeting: 10 October, Tokyo Station Yaesu central gate 12:30
--------------------------------------------------------------
Sponsor: PHT, makers of TurboLinux http://www.pht.co.jp