Re: [Axiom-mail] Defining piece-wise functions and drawing, integrating, ...

On 06/02/2007 08:46 AM, Martin Rubey wrote:
> Dear Bill,
> 
> "Bill Page" <[email protected]> writes:
> 
>> http://wiki.axiom-developer.org/358VariableIsApparentlyAlwaysAssumedToBePosi
>> tive
>>
>>> The latter filing is more complete and correctly categorised.
>> Good. Thanks for submitting the report!
>>
>> Normally I would add my comments directly to the report, but I think this one
>> deserves a little discussion first because it is important to understand what
>> Axiom is doing with your commands.
> 
> Thanks for your nice explanation.  I temporarily forgot that axiom evaluates
> all the arguments before passing them to the function. :-) A feature I *really*
> love about axiom, by the way.

Bill, that is a really nice explanation of what happens and it should 
*not* be forgotten inside the mailing lists.

> Thus it first finds a type for f(x), then for x=-1..1, and then tries to find a
> fitting signature for draw.  However, doesn't axiom sometimes try several
> signatures, until it succeeds?  Here, it sticks to the very first type it can
> find for f(x).  It seems that Axiom assumes that compiling a function with two
> different signatures yields the same result.  Well, that's probably reasonable.
> 
> I guess we can close the issue.  Maybe we should rename it though?

I would say close it, when someone has started yet another file 
"tutorial.pamphlet" (maybe even mathaction is OK) which contains exactly 
what Bill has written. There should be several such tutorial sections. 
It is good if somebody tells you what actually happens in particular if 
you don't expect Axiom's output.

Just recently I gave up on TeXmacs, since I ran into too many problems 
that I could not solve and that were not properly explained in a tutorial.

Yes, I think it is even good to show "error messages" and explain why 
they appear. If you deal with some software and often find that it 
behaves differently from what you expect, you get frustrated.

Ralf