Controlling simplification of type a*a^k -> a^(k+1)

Dear Richard san, Robert san,

Thanks for your replies. Having opinion from you guys,
I have started to think of the issue in a different way.

Dealing with generating functions, it is important to observe
  each term's numerator when its denominator is in certain form.
My previous solution to control simplification is bad because it
is applicable for very specific types of generating functions.

Since my specific interest is in computing some Dirichlet series
and his L series, I am going to write a function which takes D or
L series and make them in appropriate form so that I can easily
observe their numerator series.

Yasuaki Honda

On 2007/08/05, at 9:45 AM, Robert Dodier wrote:

> On 8/4/07, Yasuaki Honda <yhonda at mac.com> wrote:
>
>> For this purpose, I want to introduce a control variable $exposimp.
>> If the value is nil the a*a^k->a^(k+1) will be suppressed.
>
> Yasuaki, I guess I'm not in favor of this. It seems like a pretty
> narrow problem to introduce another global flag, and it seems
> like there is potential to interact badly with other parts of the
> simplification code -- I would guess that the a*a^k->a^(k+1)
> simplification is coded in more than one place.
>
> best
> Robert Dodier
> _______________________________________________
> Maxima mailing list
> Maxima at math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima