Greetings!
On 6/20/2011 3:31 PM, Stavros Macrakis wrote:
> It's unfortunate that solve doesn't solve this case.
>
> Solve does succeed on very similar cases, e.g. solve( factor(df), p )
> or even solve( df, [p, xxx]) (where xxx doesn't even appear in df!).
> And both linsolve and algsys solve it correctly.
>
> Why is this? Well, solve and linsolve each have a certain "bag of
> tricks", and it happens that solve's doesn't cover this case. Solve
> does call linsolve or algsys when there are multiple equations or
> multiple variables (e.g. the [p,xxx] case above), but not when there's
> just one.
>
> Looks like solve's bag of tricks should be extended... (or repaired
> ... for all I know, this could be a regression).
>
> Sorry for the inconvenience.
>
Not an inconvenience -- I can see now (based on your explanation) that I
*might* need to try multiple forms of 'solve' to get a right answer to a
particular problem. If Maxima were the only application I use where the
first most obvious approach isn't always the right one, I'd ask for my
money back. ;-)