bug in radicalSolve caused by 50f391bc

[Qian Yun]

I found this by accident, when comparing trunk book against
1.3.10 book.

In section 8.5.3 (page 316), there were 6 solutions.

(1) -> radicalSolve([3*x^3+y+1,y^2-4],[x,y])

(1) [[x = - 1, y = 2], [x = - 1, y = 2]]

Git bisect points to 50f391bc.

====

This raises the question that how can we make sure the
book content does not change unexpectedly like this?

- Qian

[Ralf Hemmecke]

First, thank you for taking care and finding this issue.

When I recreated the book generation I also thought of this problem, but
was too lazy to actually tackle it.

There could be two solutions. I would prefer the second one.

(a) For each command, add a testcase into the src/input directory.
(b) the xtc like environments should also have yet another parameter,
\xtc{
TEXT
}{
COMMANDS
}{
TEST_CODE
}
then the testcase would be generated and automatically executed.

Both are a lot of work, but the only way to make sure that everything is
fine.

The book is anyway not yet perfect. There are still a number of places
where there are overly long formulas. I have no idea how to make them
"automatically" disappear. Formatting of formulas is a craft and hardly
automizable. Ideas welcome.

Ralf

[Waldek Hebisch]

Well, this change shows in generated HyperDoc pages and should
have been found by normal checks before commit. How it passed
the checks? I do not know.

--
Waldek Hebisch