That the speed of light is constant for all inertial frames is the second postulate of special relativity but this does not means that nothing can travel faster than light.
There are various ways to formulate special relativity. The different approaches illustrate various different aspects of the theory so one of the tricks is to choose the formulation best suited to the question you're asking. My own favourite approach is based on the invarience of the proper time, and in fact this answers your question rather neatly.
If you think back to learning about Pythagorus' theorem, this tells you that the distance from the origin to the point in space (x, y, z) is:
$$d^2 = x^2 + y^2 + z^2$$
Special Relativity extends this idea and defines a quantity called proper time, $\tau$, defined by:
$$c^2\tau^2 = c^2t^2 - x^2 - y^2 - z^2$$
where $c$ is a constant that will turn out to be the speed of light.
The key thing about Special Relativity is that it states that the proper time is an invariant, that is all observers will calculate it has the same value. All the weird effects in SR like length contraction and time dilation come from the fact that $\tau$ is a constant.
So what about that constant $c$? Well the quantity $\tau^2$ can't be negative otherwise you can't take the square root - well, you can, but it would give you an imaginary number and this is unphysical. So suppose we let $\tau^2$ get as low as it can i.e. zero, then:
$$0 = c^2t^2 - x^2 - y^2 - z^2$$
and rearranging this gives:
$$c^2 = \frac {x^2 + y^2 + z^2}{t^2}$$
but $x^2 + y^2 + z^2$ is just the distance (squared) as calculated by Pythagorus so the right hand side is distance divided by time (squared) so it's a velocity, $v^2$, that is:
$$c^2 = v^2$$
or obviously
$$c = v$$
So that constant $c$ is actually a velocity, and what's more it's the fastest velocity that anything can travel because if $v > c$ the proper time becomes imaginary. That's why in special relativity there is a maximum velocity for anything to move. Although it's customary to call this the speed of light, in fact it's the speed that any massless particle will move at. It just so happens that light is massless.
The fact that the speed of light is a maximum speed is a derived conclusion from the postulates of special relativity. it is not one of the axioms themselves.
You can Demonstrate this in a large variety
of ways, the most convincing one is the fact that the energy required to create a particle and accelerate it to a speed $v$ is given by
$$E=\frac{mc^{2}}{\sqrt{1-\left(\frac{v}{c}\right)^{2}}}$$
which approaches infinity as $v\rightarrow c$.
It says, "The speed of light in vacuum is constant in all inertial frames of reference (i.e. for all inertial observers)". To explain in a simple manner, "Light cannot be measured relative to any objects and is always constant in all inertial frames. In other words, Light is the maximum velocity allowed by nature. If something approaches near $c$ - Time, Length and even Mass changes.
No... Of course there's no possibility for the second principle of Special relativity to be wrong. And for this reason, it was accepted and has been in existence for nearly a century. This restriction increased the interest for Physicists to concentrate on Tachyons
So is it possible the point that nothing can travel faster than light was wrong?
No. The "nothing can travel faster than light" restriction logically follows from the two postulates of special relativity.
I'll try to briefly show you how to get to the conclusion.
First you have to convince yourself that the two postulates imply the phenomenon called the relativity of simultaneity. That is the first thing discussed in every textbook on special relativity, so I'm not getting into it.
Now we use a following claim from p.1: "If one would be able to get from event A to event B only if he could move with faster-than-light speed (spacelike events). Then we can change the time order of the events A and B just by changing our reference frame." We can make A and B simultaneous, make A precede B or make B precede A -- all that just by moving to different reference frame.
Now we can start a proof by contradiction. Suppose that we have some way to transmit faster-than-light signals. It then immediately follows follows from p.2 that we can transmit instantaneous signals (by making emission and reception events simultaneous) and even signals that are received before they are transmitted (by swapping the order of emission and reception events).
Imagine that we have two guys $\alpha$ and $\beta$, equipped with such a spectacular communication channel. Then $\alpha$ could send a signal to $\beta$ "back in time", and then $\beta$ will return the signal to $\alpha$ instantaneously. Which means that $\alpha$ will receive his own signal from the future. Such ability instantly leads one to lots of self-contradictory situations. Hence our assumption was false.
From a purely theoretical point of view, the Special Relativity (SR) is based on a space-time metric $$\eta=\begin{bmatrix}+&0&0&0\\0&−&0&0\\0&0&-&0\\0&0&0&-\end{bmatrix}$$ The most general transformation to preserve metric $\eta$ is global PoincarĂ© group which is the limit of the de sitter group with sphere radius $R\rightarrow \infty$. There is an other type of de Sitter transformation with $R \rightarrow$ finite which also leads to a special relativity theory. Basically one plays with cphoton and c.
But, keep in mind that if it is possible that SR be a finite large $R$ de Sitter transformation, it has not been experimentally confirmed, and as far as we know we can use Einstein special relativity.
