Well, since you refuse to follow instructions...
Evgeny.Makarov said:
Find the intercepts, i.e., $x$ when $y=0$ and $y$ when $x=0$.
\[
x=1-8/m,\quad y=8-m\qquad(*)
\]
This part you have done.
Evgeny.Makarov said:
Express $x^2+y^2$ through $m$ where $x$ and $y$ are obtained in the previous step.
\begin{align*}
x^2+y^2&=\left(1-\frac{8}{m}\right)^2+(8-m)^2\\
&=1-\frac{16}{m}+\frac{64}{m^2}+64-16m+m^2\qquad(**)
\end{align*}
Evgeny.Makarov said:
prove that $m=-2$ turns its derivative to zero.
The derivative of (**):
\[
\frac{16}{m^2}-\frac{128}{m^3}-16+2m\qquad(***)
\]
Now substitute $m=-2$ and check if (***) turns into 0.
Speaking about your previous response:
Evgeny.Makarov said:
Solving
\[
\frac{d}{dm}\Big((x(m))^2+(y(m))^2\Big)=0
\]
may be a little complicated because it's an equation of the 4th degree, but it is easy to check that $m=-2$ is a root.
leprofece said:
it is that I got tl derive that
2x+2y
The intercepts $x$ and $y$ that you found are functions of $m$. Each slope $m$ determines a new line through (1, 8) with its own intercepts. That's why I wrote $x(m)$ and $y(m)$. Next, you need to know with respect to which variable you take derivative. It should be the same variable that you vary to find a minimum (or maximum) of something. In this case, you vary the slope $m$ and among all resulting lines you choose the one that cuts the shortest segment against the axes. Therefore, you should take derivative with respect to $m$.
Now, $\frac{d}{dm}(x(m))^2\ne 2x(m)$. This is because you have a composition of $x$ and the square function:
\[
m\mapsto x(m)\mapsto (x(m))^2
\]
According to the chain rule,
\[
\frac{d}{dm}(x(m))^2=2x(m)\frac{d}{dm}x(m)
\]
and similarly for $\frac{d}{dm}(y(m))^2$. So another way to find the derivative (***) is to substitute (*) into
\[
2x(m)x'(m)+2y(m)y'(m)
\]
Again, to find an extremum, you need take derivative with respect to the independent variable. So far I've seen problems you posted with a single independent variable ($m$ in this problem). It is possible to consider situation where you can vary two parameters, say $u$ and $v$, and you need to find the minimum of some $f(u,v)$. Then you need to take partial derivatives of $f$ with respect to $u$ and $v$. My guess is that you have not covered that so far. But in a single-variable case, it makes no no sense to say that $(x^2+y^2)'=2x+2y$. Derivative with respect to what: $x$, $y$ or something else?