Suppose $f(x)$ suffices to the following expression: $(\forall a\in D)(\exists\delta\gt0)(\forall\epsilon\gt0)(\forall x\in D)(|x-a|\lt\delta\Rightarrow|f(x)-f(a)|\lt\epsilon)$
Proof that the above expression isn't the same as $\epsilon$-$\delta$-definition for continuity.
On one hand one can start by looking for a counterexample which is a continuous function that doesn't suffice the condition and on the other hand you can look for a discontinuous function for which the expression holds.
For the $2^{nd}$ case I would use the stepfunction at a certain point a for which it intuitively holds. But I can't find a counterexample that contradicts the above expression.
Any hints?
Recall the definition of continuity for $f:D\to\mathbb R,D\subseteq\mathbb R$
The condition you gave is:
But now we must have a $\delta$ that is universal for every $\varepsilon$ opposed to finding a $\delta = \delta (\varepsilon)$. Switching the order of the quantifiers changes the expression.
For a counter-example you may consider $f(x) = x$. Fix $a\in\mathbb R$. Does there exist a $\delta$ such that the difference between any $f(x)$ and $f(a)$, where $x\in (a-\delta, a+\delta)$, is arbitrarily small?
Suppose such a $\delta >0 $ exists. Let $\varepsilon = \frac{\delta}{2}$. Then for $x = a+\delta /2\in (a-\delta,a+\delta)$ and $$|f(x)-f(a)|\geq \varepsilon\ $$