Le 24 oct. 07, à 21:12, Richmond Mathewson a écrit :
This is an interesting discussion; and mainly, for me,
because it reads as, only, a discussion about how
users of Runtime Revolution are going to access the
proposed documentation.
What has not been addressed are:
1. Any pedagogical questions at all -
2. Heirarchy of difficulty.
my comments on these two. The first obvious thing is that quite a
number of persons are dissatisfied with the documentation that
Revolution provides. Until now, most (if not all) the answers implied a
user community effort to provide a better doc. Following Richmond's
message, I would like to change perspective:
revolution's documentation is bad mainly because the syntax of
Transcript has departed from the natural english approach of Bill
Atkinson to more a conventional one like function(arg1,arg2) etc...
Frankly, I appreciate to be able to "lock objects" instead of "set the
lockWhetever to true".
I recently had a look at the video library, and I find it awfully
complicated to remember all of these RevBit1Bit2 keywords: in some
commands, "Video" comes just after "Rev", in others "Video" comes in
third or fourth position: this smells ill assimilated externals. The
rev team should start thinking about updating the language and the
parser. For instance revVideoGrabDialog could be synomym with "get
Video Dialog" or "answer Video" like "answer file" etc...
Revolution advertises its language as being "natural english"; they
should be careful it remains true. This means more reserved words, but
not many. On the other hand, if one consider that "grab" for instance,
should be a keyword, it should be usable with as many kinds of objects
as possible. This is how the language can be merciful (and hence
"natural"): by allowing several formulations for the same meaning.
On the documentation: by looking at the excellent BvG stacks at
http://bjoernke.com/runrev/stacks.php , I discovered that the doc was
some kind of compressed XML; I mean, why this XML stuff? Why not plain
Revolution stacks? Some fifteen years ago, I bought "FreDOS'stack", by
Frederic Rinaldi, which was a collection and database (understand: HC
stack) of externals. Over the years, I added some buttons, some
maintainance scripts etc... This was possible because the intellectuals
means required to customize the stack did not exceed basic HC
programming, and, most important, improving the documentation stack was
an exercise in the HC programming and, in some cases, learning to use
the most useful externals in the stack.
To make it short, when I look at the XML files and their output, I
think all of this can be translated into stack format. This is what the
Hypercard team did: their docs were HC stacks. The obvious advantage is
the ability to do custom searches in the way we are supposed to work:
using revolution objects. When I think of this XML stuff, I have the
feeling that the people who did the doc did not believe in their own
product.
3. How one can design a set of documentation which,
while addressing extremely basic concepts (e.g. the
conceptual leap that is needed to understand that
sometimes A+1=A) also will satisfy rather more
sophisticated users.
Considerations such as:
1. Target age group . . .
2. Type of end users to be targetted ( e.g. educators,
professional computer programmers, pensioners who want
to try programming, and so on, ad nauseam) . . .
I sat my 11 year-old (well, now he's a 12 year-old)
down in front of my Mac mini and DC 2.6.1 and told him
to get on with things and read the documentation: he,
predictably, and understandably, said something
unrepeatedly rude after 15 minutes. I then had to sit
down with him and 'nurse' him through everything - I
don't mind that, I'm a Primary teacher.
My 75 year-old father wouldn't have a clue either;
even though he has an excellent B.Sc in Chemistry from
the days when Universities did not hand out degrees
like sweeties, and he's a Fellow of the Royal Society
for Chemistry. He, memorably, described the RR
documentation (1.1) to me as "unusable" - and he
taught kids for 35 years.
You are all right. The language, the objects, the documentation have to
be as simple as possible. People can improve the doc (and write custom
scripts) afterwards, PROVIDED they can understand the primary doc (and
language description) first.
I learnt RR by trial and error, old HC knowledge
(founded on Danny goodman's EXCELLENT book - excellent
that is for people who have already acquired some
computer knowledge) and learning how to understand the
slightly odd short-hand that had been used in the
current RR documentation (RR 1.1).
Somewhere, under all the other piles of nonsense, I
have a slim volume from an old version of something
(err, its referenced in my M.Sc thesis) which talks
about 'Cognitive Apprenticeships' - this term makes a
lot of sense; but only if the apprentice has somebody
standing by as the 'Master Builder' to guide them.
Really good documentation should be an effective
substitute for that Master Builder.
Now my Master's thesis (plug, plug) burbled on at
quite some length (err, quite boringly, come to think
of it - never mind, fooled the examining committee -
Ooops) about how 20 odd years ago there had been a lot
of hype about computers and how experts in
non-computer subjects would be able to just sit down
in front of a computer and press a few buttons and
"abracadabra" a super, whizz-bang 'thingy' for content
delivery and/or data processing would just
automagically drop out of the sky . . .
This has turned out to be a "load of fairies at the
bottom of the garden"; for reasons I have gone into in
that M.Sc thesis - it could happen; but it probably
won't [OK, all you companies with tons of ready money
ring me now for me to work on a practically endless,
open-ended project that will break all the rules of
SDLs].
Now, not many people are interested in fairies at the
bottom of the garden. But a hell of a lot of people
are interested in what computers can do, but feel
unempowered.
If RR is to go from, frankly, a niche RAD (tactful,
very tactful, Richmond) into the killer App it should
be (or, to be extremely corny, "the killer App
HyperCard once was" - Steve Jobs needs a smack on the
hand) it needs documentation that will allow people to
follow a cognitive apprenticeship fast and
efficiently.
completely agree with you. Revolution should be the IDE" for the rest
of us".
And that was written after a day's teaching 5-12 year
olds (12 contact hours) - be glad; it would have been
longer, and more pompous, had I not taught those kids
- and what is even more worrying is that I mean
everything I say :)
sincerely, Richmond Mathewson
bye
Francois
____________________________________________________________
A Thorn in the flesh is better than a failed Systems Development Life
Cycle.
____________________________________________________________
___________________________________________________________
Want ideas for reducing your carbon footprint? Visit Yahoo! For Good
http://uk.promotions.yahoo.com/forgood/environment.html
_______________________________________________
use-revolution mailing list
[email protected]
Please visit this url to subscribe, unsubscribe and manage your
subscription preferences:
http://lists.runrev.com/mailman/listinfo/use-revolution
Francois Chaplais
http://cas.ensmp.fr/~chaplais/
http://cas.ensmp.fr/~chaplais/index-e.html
_______________________________________________
use-revolution mailing list
[email protected]
Please visit this url to subscribe, unsubscribe and manage your subscription
preferences:
http://lists.runrev.com/mailman/listinfo/use-revolution
