[Axiom-developer] foreign concepts

[Bill Page]

On Saturday, October 02, 2004 4:11 PM Bob McElrath wrote: 
> ... 
> And now I've gotten beyond 3 sentences and am just rambling.  
> Bill and I seem to be good at this.

Rambling on is much easier than programming ...  :} :} :}

> So let's see how many foreign concepts we've piled on:
> 
>     Object-Oriented
>     Wiki
>     Literate Programming
> 
> Not to mention LaTeX, Computer Algebra, Open Source, Zope, 
> ZWiki, Plone. I just hope the set of concepts most people
> will have to figure out isn't so large that this project
> will have difficulty attracting contributors.

Bob, I think you are absolutely right to be concerned about
this. It probably is too complex and it probably is already
having difficulty attracting contributors. I really don't
know what to expect.

The problem is, even the simplest version of what I (we?)
would like to do with all this seems to be too complex -
just Zope + LatexWiki + Axiom. In fact, surely some might
claim that LaTeX is too complex, doing algebra by computer
is too complex, wiki-based collaboration is too complex ...
even individually.

> Your early confusion is a bad sign...  I think I've got
> them all mostly figured out except "literate programming".

No doubt there could be a mix of similar admissions by
people reading this: "I think I know enough about 'x' and
'z' but what is 'y'? Actually this is not such a bad state
to be in and maybe a good reason to be hanging out here.

In my case I certainly might say: I think I've got them
all mostly figured out except "types in Axiom". :)

> 
> Wiki's generally have a good number of "curious" edits...
> people that come along and think "can I really edit this?"
> "What happens if I do?" and this is to be encouraged because
> it then leads to more interest...

Yes, again I agree completely. So far MathAction has only had
what I think is a rather small number of "curiosity edits" in
spite of being fully open. To really know for sure I guess
would might have to take a look at the web logs and set how
the ratio of 'page views' to 'page edits' is compared to other
wiki.

But I have made only limited attempts, mostly indirect, to
advertise MathAction. I am open to suggestions as to how
best to approach this, if at all.

More broadly speaking, I think the Axiom project as a whole
could use more PR. I think there are some quickly evolving
ways to do this on the Internet ranging from being mentioned
in popular blogs to an article in a part-paper / part-electronic
trade magazine. But someone has to step up to the task of
preparing the equivalent of "press notices" etc. We might
worry that "we are not ready yet" for such publication, but
on the other hand one of the reasons we are not ready yet
could well be *because* of the lack of publication.

One of the reasons for my reacting quickly to the possibility
of adding Reduce to MathAction was the thought that sharing
the available audience might be beneficial to both. Of course,
as we say on the FrontPage of MathAction this potentially
extends to other computer algebra systems as well. I have
not seen many comments pro or con about this strategy.

> Pages must also be simple enough that newbies can figure
> out what it is when they hit "edit".

You are right to imply that MathAction does not meet the
usual wiki design goals very well in this respect. But I
think it is first necessary to define what we consider
the target audience for this system to be. For example,
if the result of clicking "edit" and seeing a LaTeX
format-like input file would be interpreted as too complex,
then I am inclined to think that perhaps we have set the
bar a little too low to start with. But I am quite open
to other opinions on this.

> The previous discussion of literate programming/++
> comments and document sections... results in a very
> complex, non-intuitive document to a newcomer...

That is true, however even the fill-in-the-box-and-click-
save" comment box at the bottom of each page seems (so far)
to be little used. Therefore I do think that the main
problem so far is that the audience is just too small. I am
quite reluctant to believe that the underlying complexity
is so obvious to first come page viewers that it prevents
them from even commenting. :(

> 
> Since this Axiom wiki is destined to be relatively complex,
> a good start would be to add a simple set of instructions
> to the editform.

This a good suggestion. We already have a "For editing help,
see HelpPage. But I presume you mean something a little more
"in your face"?

BTW, has anyone thought of providing a settable user
preference that might allow one to select the amount of
help and other verbosity that one might want to see, e.g.
"I'm an expert, don't bother me with that stuff", "I
still need a little help", "I am a novice user who needs
some serious hand holding", etc.

--------

Ok, back to work. Where did I put that Python code that
I was just working on ... 

Regards,
Bill Page.