The Inform Designer's Manual


29 Dec 1995 08:53:13 GMT

In article <4bvdks$it4@grid.direct.ca>, Neil K. Guy <nkg@grid.direct.ca> wrote:
>Paul David Doherty (h0142kdd@joker.rz.hu-berlin.de) wrote:
>
>> Come on... First there's a long thread about how terribly unfair it
>> is to post bug reports, and now you can't even criticise the Inform
>> manual without getting lambasted. Lighten up, folks :-)
>
> There's criticising, and then there's being a discourteous jerk. The
>post that started this thread fell squarely in the latter category.

I don't think anybody would flame Mr. Anonymous for criticizing the
Designer's Manual. He's being flamed not for what he says, but for the
way he says it (the Usenet Equivalent of an abusive, anonymous phone
call, only it's being broadcast all over the world).

But while our ingrateful, loud-mouthed, obnoxious, cowardly, anonymous
"friend" is in bad need of some manners lessons, he does have some
points.

We can ignore his hysterical whining about exercises being "unfair",
since who said that exercises have to be fair (it's not as if Graham
were grading your answers and only licensing Inform to those in the
top ten percent :-)). But it is true that many of the "exercises" in
the Inform Designer's Manual are more like research assignments than
exercises.

I think this is partly a cultural phenomenon: most modern teaaching
material, especially of the American variety, abounds with simple
exercises that are there to verify that you've read the preceding few
pages. For example, a typical American calculus textbook has a chapter
on differentiating polynomials; then there are three pages of
polynomials to differentiate.

Graham has quite a different pedagogical style. To extend the calculus
analogy, it's more like following up the "differentiating polynomials"
chapter with a few theorems to prove; and the proof requires that you
read ahead a few chapters, do some supplementary reading on the
convergence properties of Taylor series, and then spend a week or so
working really hard.

Many of Graham's exercises simpky require you not only to read ahead
(in some cases, to read the entire DM) but also to do some research
by reading the library source code, experimenting a lot, and so on.

This is a method of teaching that's perhaps more suitable for Oxford
than for Americanville High, but there's nothing "unfair" about it.
And in the end, I think you learn a lot more from this kind of
"exercises" than from the simplistic "differentiate these 400
polynomials" kind.

What we can criticize Graham for is that he doesn't prepare the
unsuspecting, previously-spoonfed reader for what's in store. A
non-programmer may well give up in despair after trying the first few
exercises, *believing that you're supposed to solve them with no other
knowledge than what youve just read*.

The DM is a rare and wonderful beast: a computer manual with literary
value. I'd hate to see it replaced by a streamlined, Americanized,
step-by-step manual. And we must remember that nobody's payed Graham a
penny for his work. Anonymous flames are not only ingrateful, but
totally uncalled for.

Magnus

P.S. One thing that I think can be a greater obstacle than the
difficult exercises is that the DM presupposes some knowledge of C
programming (the same thing holds true of the TADS manual). This is a
turn-off for aspiring IF authors who don't know how to program. On the
other hand, I'm not sure what can be done about it, without adding
another few hundred pages to the manual.