The ultimate origins of feathers probably lay in the LCA of Ornithodira, the clade containing the last common ancestor of both dinosauromorphs and pterosaurs and all its descendants. The first proto-feathers were probably tubular structures not too dissimilar to fur, and we find evidence of this hair-like insulation in both pterosaurs and ornithscians such as Psittacosaurus which may explain why Psittacosaurus had all of those spikes at the end of its tail to begin with. Triceratops may have ultimately been covered in spikes.
Only in theropods do we find that the original tubular, hair-like structure splays out, forming "tufts" of feathers, as can be seen in Sinosauropteryx. A little later down the line of Theropoda we discover that the first contour feathers appear, what can be called the first "true" feathers. Theropods such as Velociraptor and Guanlong possessed these types of feathers. The first flight remiges appear in Archaeopteryx and its kin, but there are grounds to call Archaeopteryx a flying dinosaur, being only a stem-bird, rather than a true bird of the clade Avialae.
For example, Archaeopteryx only possessed the beginnings of a keel, something that another, more derived proto-bird, Confuciusornis, shares. Other supposedly avian features of Archie, such as the furcula, or "wish bone", have been seen to occur in other coelosaurs as well, even occurring in Tyrannosaurus. Archie also possessed a pelvis transitional between the one occurring in, say, Utahraptor and Carnosaurus and the pelvis of modern birds. This is probably an obvious anatomical feature of Archie, but Archie also lacked a beak, instead having a maw full of teeth.
Archaeopteryx also possessed feathers on its hind legs, therefore with all of these anatomical characteristics of Archie taken into account, Archie was probably an incredibly clumsy flier, only barely capable of powered flight.