The idea is quite old:

Shortly after the 1859 publication of Charles Darwin’s On the Origin of Species, the British biologist Thomas Henry Huxley proposed that birds were descendants of dinosaurs. He compared the skeletal structure of Compsognathus, a small theropod dinosaur, and the “first bird” Archaeopteryx lithographica (both of which were found in the Upper Jurassic Bavarian limestone of Solnhofen). He showed that, apart from its hands and feathers, Archaeopteryx was quite similar to Compsognathus.

But having fossil evidence is quite young:

One of the earliest discoveries of possible feather impressions by non-avian dinosaurs is a trace fossil (Fulicopus lyellii) of the 195–199 million year old Portland Formation in the northeastern United States. Gierlinski (1996, 1997, 1998) and Kundrát (2004) have interpreted traces between two footprints in this fossil as feather impressions from the belly of a squatting dilophosaurid.

https://en.wikipedia.org/wiki/Feathered_dinosaur