Subject: Re: User-facing applications
From: "Stephen J. Turnbull" <stephen@xemacs.org>
Date: 29 Mar 2002 14:51:18 +0900

There was a subtext I was missing, and maybe it's useful to make it
explicit.  FSBs are "ethical businesses."  That is, they constrain
their business practice by principles other than (1) law, and (2) such
straight dealing as is essential to the long-term relationships they
need.

One might think that treating customers in a "civilized" manner
"should" be one of those principles.  However, I think it's more
robust to derive it from "straight dealing."

>>>>> "Ben" == Ben Tilly <Ben_Tilly@trepp.com> writes:

    Ben> Stephen Turnbull wrote:

    >> But you're kinda missing the point.  It's not your words that
    >> matter, it's the more or less willful misunderstanding of them
    >> by "magic box" users.

    Ben> Actually your characterizing the user's responses as "more or
    Ben> less willful misunderstanding" is treating them as being as
    Ben> much of "magic boxes" as their treatment of computers.

I don't think so.  Some users are willful.  Others are not.  The
appropriate ways to deal with them differ.  However, in both cases
_for the FSB_ the misunderstanding is not the customer's problem, it
is the FSB's.  At least if the FSB wants the relationship to continue.
This is false in comp.lang.* or *.help (Eric's topic), where matters
are reversed, no?

    Ben> The most common such misunderstandings are quite predictable if
    Ben> you try.

Sure.  I didn't deny that.  My point about "willful" is that the magic
box user comes in to the channel expecting to be given advice the way
her favorite uncle would have given it.  And she is not about to be
disabused of the idea that it is her right.  The profitable behavior
for an FSB faced with this attitude is called "pandering".  However,
in a non-business context, it's up to the regulars in the channel to
decide what proper behavior is---not the newbie.

    Ben> In this case I don't think that Lynn's characterization would
    Ben> have caused problems given her overall message.

Oh?  Then why did David, who surely would not feel himself to be a
"magic box user", react as he did?  I think that the word "magic"
would be enough to trigger most inferiority complexes.  Ie, use of the
word "magic" in this context is enough to indicate that "[the hacker]
think[s] that non-experts are morons who are beneath [his] notice."

    Ben> read the article [smart-questions.html] linked to at
    Ben> http://www.chiark.greenend.org.uk/~sgtatham/bugs.html.
    Ben> Unlike ESR, his message is delivered in a way that results in
    Ben> postive behaviour modification.

Sure.  That was my (implicit) point to Lynn.  I appreciate your making
it explicit.

The following is edited for emphasis:

    >> and here's what Eli Zaretskii had to say about _that_:

    Eli> Its implied assumption that hackers cannot talk like
    Eli> civilized human beings, and shouldn't be asked to, is
    Eli> especially disturbing.

    Ben> I *absolutely* agree with Eli.

You're kidding.  Do you really think there is an _implied assumption_
that hackers _cannot_ talk like civilized human beings?  AFAICS, the
implied assumption is that "lusers really exist (and the reader might
be one of them)."  The difference is important.

IMO, Eli's phrasing is highly disrespectful.  "smart-questions.html"
is quite explicit about why the style on high-traffic comp.lang.*
channels is what it is, and that, regrettable as it may be, the result
is that it spills over to the way experts often talk on help channels.
You may disagree with Eric's anthropology of hackers, but there is no
"cannot" and no "implicit" about it AFAICS.

I think the anthropology is quite accurate.  Take Eli himself.  Eli
wants everybody to be "nice".  And if people aren't nice on channels
where he has influence, he calls them on it.  This is _not_ costless
to him: he takes a lot of flak from the "smart-questions" camp for it.

But ... he is subject to the same incentives that Eric describes, and
quite often responds to them in classic fashion.  Eli is a sincere and
strong advocate of the "be nice" position.  He has a lot invested in
it.  If he can't pull it off successfully, why should people who
sincerely advocate a different set of values even try?

The only answers I can see are (a) because it directly serves some of
their goals (eg, bugs.html is an attempt to get better bug reports
which leads to better software) or (b) because they're paid for it.  I
don't think it's a good thing for fsb to conflate "being `civilized'
for its own sake" with "what's good for FSBs."

If you do, and try to "select nice people", I strongly suspect you
will find that (1) they're not as "objectively" competent as the not-
so-nice ones on average, and (2) they're not all that nice, either.
So this is a really difficult human resources management problem: how
to get people to combine technical excellence with a service attitude.
"Absolutely agreeing" with Eli leads to sweeping this under the rug, I
fear.

    Ben> The point of basic courtesy is not to be all nice and
    Ben> touchy-feely.  It is to help get communication to happen.  If
    Ben> you want to communicate, then don't start by pre-emptively
    Ben> shutting off lines of communication.

Well, sure.  But that's where "willful" comes in.  Eli didn't like the
implications of the analysis in "smart-questions.html," so he attacked
its style while claiming that the style was the substance.  Attacking
the "lusers exist" assumption would have forced Eli to concede quite a
bit on the original issue (and would have been OT for the thread).
The calculated courtesy of "bugs.html" forstalls this kind of attack.
(I'm projecting, of course.  Simon Tatham very likely "just wrote"
that page.  But I too can and do write like that when I need to---but
at great cost.)

Customers will conflate style with substance too, of course.  "He's
such a geek he can't possibly know what we the users need."

So let's at least be conscious of when we're being McLuhanesque.
Confusing the style of "smart-questions.html" with its specific
content is dangerous to the conduct of an FSB, if it leads to
rejecting the analysis without understanding it.  Conversely (and more
generally), to provide our customers with the best possible service,
we should avoid giving them an excuse to reject unpleasant advice
because of the style.

    Ben> This is as true for FSBs as for individuals, with the
    Ben> significant difference being that FSBs will want individuals
    Ben> who work for them to be aware of how they are causing the FSB
    Ben> to be perceived.

Well said.  But beware the problem of going beyond education to
selecting on the basis of "bedside manner", and possibly losing
"surgical skill".

    >> Does this have something to do with the (stipulated) suckiness
    >> of OSS GUIs?  I dunno ... but I find it plausible.

    Ben> Well, developers who take ESR's attitude will have a hard
    Ben> time finding people who are willing to give them useful UI
    Ben> feedback...

There's more than just getting feedback here.  I wonder whether the
hacker "just the facts, ma'am" attitude _directly_ interferes with
creating a UI that "kinder gentler" users will find transparently
enables their work.  Ie, are there "elliptical idioms" in user
interface design?  Terms like "desktop metaphor" would suggest so.


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
              Don't ask how you can "do" free software business;
              ask what your business can "do for" free software.