Was it always that way? At some point someone must have tried to measure the speed of light...
Oh, absolutely. Usually by bouncing light back and forth between mirrors far apart. And we still do - we've just have realized that's it's not really the speed of light we're really measuring, but the distance/time relation.
Because time itself is variable. A second here is not the same as a second at a GPS satellite. But the speed of light in vacuum is the same.
Anyhow, that's our current point of view, because it makes it easiest to do calculations and make observations. We could use a point of view where time was a constant, and have a very variable light speed instead. It's just as mathematically legal as a point of view, but it would complicate how we have to perceive things. Planck's constant would be variable, and atoms smaller or larger depending on location. We'd have to shift our view of distant galaxies to being very small, close, and low energy, but having a very slow speed of light compared to our speed of light. A headache, even if mathematically valid.
So Einstein took the simpler point of view, and let distance and time be variables with a fixed relation.