Hacker News Reader

Comment by frisco | Hacker News Reader

BitTorrent, by Bram Cohen. Napster. Ruby on Rails (maybe; I'm uncertain how much it was just dhh in the early days).

3 clever ideas with very poor implementations, heavily refined by many others over years.

I can do go for that for RoR; but Napster was a much bigger leap in its time than we give it credit for now, and the same for Bram/BT. The BitTorrent protocol was well done, even once P2P was far from an obscure idea.

And all of them were "innovations", not inventions. None of them were original ideas, they were just rehashed ones that were right for the time in which they became popular and widely known.

To be fair to Bram, he had the technical insight to deliver a workable P2P protocol with the ability to distribute demand at a nice granularity. Because of that, his protocol was technically superior to the competitors. Also, he did not try to tackle search and do it badly.

Napster was a bad implementation at the right time.

RoR - credit should be given. Sometimes just getting off your duff and starting something is what the world needs. Motivating many others to refine code into something useful deserves credit.

(EDIT: I am not saying anything was/wasn't a work of genius. I'm just saying it's truly creditable, whatever it was.)

To be truly fair to Bram, he had the insight to take an existing marginally working p2p system that he had previously been employed to develop and strip it down to the essential components needed to shift bits quickly. Innovation, not invention.

I definitely agree with you on Napster. I am less sure about RoR. It was a a very nice improvement over existing systems, but how much of that was DHH genius and how much was simply due to Ruby providing a better platform upon which to build this particular framework I do not know. It was definitely better than existing alternatives in other scripting languages, but I was not a part of the Ruby world at the time and have no idea where the initial RoR work fit in to the whole Ruby ecosystem at the time.

Well, you can write fortran in Ruby, but it took DHH to show where it does shine. It is easy to underestimate the initial insight in hindsight "oh, that's obvious", but it only becomes obvious when someone does it. And does not definition of geniality has something on having that insight for the "obvious"?

Others had shown were it shines as well; there was Nitro, IOWA, a few other apps that were solid examples of Ruby qua Ruby at the same time as (or prior to) Rails.

Why do some people/code get attention and accolades, while others, equally deserving, do not? Timing? Personality? Bluster?

The ability to mobilize people into a large, cohesive community.

Sure, but how does that happen?

DHH said (more or less, I forget the exact words) that Rails had no new ideas, just implementations of existing good ones.

Nitro, another MVC-ish Ruby Web framework released at the same time as Rails, always stuck me as being a much better approach to the same problem, and better engineered (thread safe, less "magic" and munging of core Ruby classes, a form of pipeline transformations somewhat similar to Rack middleware today), but for whatever reasons Rails managed to capture far more attention.

The big win for Rails was not the initial code itself, but in capturing attention and motivating people to contribute, to fix it up, and make it better.

Reply on news.ycombinator.com