2 years, 27 days

## Elegant solution...

Thanks for demonstrating how to use solve/identity.

## Thank you....

This was my original method, but your code is simpler/more elegant than mine.

## @vv  It is obvious that this probl...

It is obvious that this problem does not always have a solution. I don't know if that is the point of your answer.

But I'm interested in finding f if it does exist. (There are interesting situations where this is the case.)

## Thank you. Can the inner polynomial be f...

I didn't know compoly.

Is there a way to force compoly to compute f (if possible) in the decomposition f@g=h when g and h are given? That is what I'm looking for.

## Works fine...

Thank you (either way is fine).

Looking at your code, I can't see that it is different from what I have tried, but now it works! :-) And that is nice. Thanks.

## @Kitonum  Thank you very much for ...

Thank you very much for your solution.

## @Carl Love  Thanks a lot for shari...

Thanks a lot for sharing your work. It looks very useful.

## @vv  I think I understand what is ...

I think I understand what is going on in your code. Thank you very much.

## @Carl Love  Do you know why this d...

Do you know why this doesn't give any solutions:

select(type, eval~(''[x, y]'', {isolve}({169 = 10*x + 3*y})), list(posint));

I just replaced the equation (and not the variables).

## @Carl Love I'm preparing some teach...

@Carl Love

I'm preparing some teaching material about diophantine equations (both the sum of squares problem for two or more squares (I know the theory) and other equations). So I'm playing around with concrete examples for exercises and experiments.

I would be happy to see your module :-)

## @Carl Love Thank you very much for ...

Thank you very much for your variation.

My question was more general than sum of squares, but thank you for pointing out that specific command.

## @vv Thank you very much! That is a ...

Thank you very much! That is a nice way of doing it.