Seeking Help with Windows Maxima installer localization

I think the distinction is right -- however describing precisely what
commands do is hard to read and write (a large program can have many subtle
parts), and especially hard to understand for a novice  "integration
includes the pure exponential and pure logarithmic case of the Risch
Algorithm, but algebraic cases of genus 2 only unless the case can be
reduced to genus 2"  might be precise but quite useless.

The original Macsyma manual had a particularly annoying style in which a
command was described in general and then all or most of the examples were
boundary cases or exceptions in which the command did something else. This
failed to be adequate descriptively or prescriptively.  

I think that the current typesetting convention uses a painfully large
amount of space for examples (at least in the version I look at); the
on-line Mma manual which provides for "click here for more examples" makes
excellent use of space.

Also keywords and distinctions would be useful.  E.g. factor/ factorize
/roots should refer to the factor command but also  root finding commands
like realroots and Newton and polynomial approximate zeros etc.  The
commercial macsyma has a "natural language" interface "Mathtips" worth
looking at. It's not great, but it sometimes hits the nail on the head,
anticipating common kinds of requests. With enough FAQ and an automatic
lookup we can probably get 95% of all novice questions answered.  If we can
get the novice to use the system.  

RJF


> -----Original Message-----
> From: maxima-bounces at math.utexas.edu [mailto:maxima-
> bounces at math.utexas.edu] On Behalf Of Robert Dodier
> Sent: Wednesday, November 15, 2006 11:08 PM
> To: Richard Fateman
> Cc: Maxima at math.utexas.edu; van Nek
> Subject: Re: [Maxima] Seeking Help with Windows Maxima installer
> localization
> 
> On 11/15/06, Richard Fateman <fateman at cs.berkeley.edu> wrote:
> 
> >  so Robert's objection [ugh] to the use of "you" in my earlier example
> is
> > presumably a matter of some delicacy.   There is a philosophical notion,
> not
> > worth pursuing, that anthropomorphizing a program is a bad idea (e.g.
> > "Factor likes to return unsimplified answers such as 2*3^2".   Clearly
> > programs do not have desires.  But other wordings take more room and are
> not
> > as clear in some sense.)
> 
> The distinction we want to make, I believe, is between descriptive
> texts and goal-oriented texts. The reference manual falls in the
> former category and tutorials in the latter. The point of descriptive
> texts is to give the reader information sufficient to deduce a course
> of action to solve a particular problem. Goal-oriented texts show
> particular examples of that.
> 
> "You can do X with the expand function" is less informative,
> as a bit of a descriptive text, because it says nothing about
> other capabilities. Even attempting to cover the bases with
> "You can do X Y and Z with expand" doesn't solve the problem,
> because it leaves open the possibility that there also exists A
> which expand can do. "expand does X Y and Z" on the other hand
> closes the door.
> 
> FWIW
> Robert
> _______________________________________________
> Maxima mailing list
> Maxima at math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima