It now looks like feathered dinosaurs were a lot more common than that though. Virtually all of the raptors (velociraptor, utahraptor, etc) are assumed to be feathered based on fossil evidence today. When I was in grade school the real breakthroughs were in the area of evidence of warm-blooded dinosaurs. More recently though, there have been a lot of breakthroughs suggesting feathers were much earlier than we had previously thought.
Velociraptor almost definitely had feathers, as well as wings. In fact, it pretty much was a flightless proto-bird (though scientists believe it may have descended from a flying predecessor). If it were alive today, we'd basically see it as a curiously shaped, medium-to-large land bird with teeth.
Check out some of the post-2007 renderings of Velociraptor. My mind was certainly blown, especially after having grown up with the image as portrayed by Jurassic Park.
(2007 was when Velociraptor bones were discovered with quill knobs, almost indisputably indicating that the animal had wings and feathers).
I wouldn't say that all dinosaurs had features, but a fair number did (and fairly large taxonomical groups!).
Where the breakthrough is here is that we have evidence of the sorts of colors. In the past all of our speculation has been that of "artists impressions" but now we have something more tangible to work with.
So not knowing anything about genomics / evolution - what advantage might proto-feathers (and 'pre-proto-feathers' before them - if there is such a thing) have conveyed that caused them to become full feathers? How does any evolutionary advantageous feature get started?
For one thing, evolution can sometimes work by going through several stages, of which the intermediate are not particularly useful but not harmful either. By mere chance, they happen to last long enough for the actually useful changes to build on top of them. For example, see Lenski's long term evolution experiment with e. coli. He's spent 25 years breeding several populations of e. coli, and freezing samples every 500 generations or so (every 75 days). After about 20 years, he discovered that one of the strains had evolved the ability to metabolize citrate in an aerobic environment, which e. coli normally can't do. Going back to the historical record, unfreezing some of the past samples and repeating the experiment, he found that sample from after generation 20,000 could re-evolve this trait, but clones from before that time could not, indicating that there was a potentiating mutation that that particular strain had had shortly before generation 20,000, which the other samples did not have. Further evidence has shown that it may have actually been two separate potentiation mutations.
Beyond that, however, proto feathers would have probably been useful for insulation. They seem fairly similar to down feathers, which are great insulators. Given the the feathers were also colorful, they may also have been decorative, and used for catching the eye of potential mates.
But we don't know that those potentiating mutations didn't have some kind of benefit on their own, so that doesn't do much for the question of how proto-feathers provide enough benefit to stick around and spread long enough for the next several million mutations to arrive and turn them into feathers. Is insulation adequate for that?
If you take a look at the Lenski paper, it says:
> The potentiating mutations are not known to confer any phenotype amenable to screening, so there is no simple way to distinguish between potentiated and non-potentiated clones.
Now, that isn't conclusive evidence that there was no benefit on its own, but it doesn't sound like there's any evidence that it did provide some kind of greater fitness. It sounds like all evidence shows that the original mutation didn't provide any particular phenotypic change on its own, it was only after the later mutation that it had any measurable effect.
Now, with dinosaurs it's much harder to tell. We obviously can't do any kind of controlled experiments, or observe them directly. But we can offer up a few potential explanations. One is that some mutations can occur without either helping or harming, but later on another mutation can lead to something useful.
Or there may be benefits conferred by proto-feathers. Remember, in addition to their similarity to down, proto-feathers are somewhat similar to hair, and may provide many of the benefits that hair provides: warmth, colors which allow either catching the attention of mates or blending in to the background, better sense of touch (you can feel something before your skin brushes against it), better sensation of the wind, and protection from things that may scrape or cut you.
According to Wikipedia, the insulation and display (colorful feathers for attracting mates) hypotheses seem to be the more probable. Obviously, we're not going to solve this by speculating here on HN, and it may be that the reason can never be conclusively determined because you can't go back in time to run controlled experiments, but based on the evidence we have those theories sound reasonably plausible.
The same reasons why many flightless birds continue to have feathers. First off they are excellent insulation. That could be critical for a not fully thermoregulated organism. Dinosaurs are thought to be warm-ish blooded but it's possible they didn't have the kind of thermoregulation that modern birds or mammals do.
Add to that waterproofing, which can aid in thermoregulation as well. The phrase "like water off a duck's back" is quite apt, feathers can shed water better than a modern high tech coat. Keeping dry means keeping warm. Also, dinosaurs that were occasionally aquatic could have benefited from feathers.
And very fast moving dinosaurs could have used feathers to aid in aerodynamics, which would allow them to run faster and maintain greater control over their movements.
I think features like this start off as mutations that don't cause an evolutionary disadvantage to the creature that bears them. After a few generations where they change a bit every time, they might actually become useful to the animal's survival and start evolving even faster. If they became disadvantageous at any point they'd probably become excluded because they'd make it harder for the animal to reproduce.
EDIT: Reading on I see that if stored at cold temperatures it can last a lot longer.
Yes, though it seems they need to be stable, constant temperatures too. The Denisovans (human related species, ancestors of some of us) are known entirely from 41,000 year old DNA from a cave in Siberia, and the sequence recovered is basically the full genome - as good as for living humans and a lot better than any found for Neanderthals so far.
Honestly, if you're worried about your image the citation to slate isn't the one I would have disclaimed.
That's quite some editorial comment at the end of the crev.info piece:
These claims should be remembered if ancient DNA older than that is confirmed in future finds[...] if intact DNA is found in a dinosaur or other fossil older than the upper limit they just stated, it could have the effect of falsifying the evolutionary timescale. Since evolutionists are such staunch believers, though, most likely the reaction will be, “Well, what do you know; DNA can survive for 65 million years.” [...]
The rest of us should remember what they said beforehand about DNA’s upper limit age, and not let them get away with it.
Now we know -- if you've had result A for hundreds of years, and result B for singles of years, then evidence suggesting that the two conflict clearly supports result B.
EDIT: link to http://crev.info/2013/07/longevity-of-dna-estimated/ (and vacillation about the propriety of citing to slate) was removed even before I posted this reply. It seems to be a nice, on-topic summary of current knowledge on the longevity of DNA, but wow, the editorializing. :/
I really should have taken a second look at what I was citing! I removed that before I saw your comment. For some reason I thought I'd cited some interesting genetics blog. I read the paragraph. Yep. Took a quick look around — looks science-y.
Oops. Bah. My shame. I spat my coffee out when I read the whole article.
On a related note, I really wish HN had a preview feature. I often end up posting & then editing because 1) no preview 2) sessions timing out.