Live data from Hacker News

How we built a P2P file sharing network from scratch

xorro-p2p.github.io

11–20 of 22 posts

Re: How we built a P2P file sharing network from scratch

#11

Earlier quoted context omitted.

Why? It doesn't seem like time-sensitive content. Should any link to tutorials or documentation also have a date on it?

Perhaps yes. Unless you don't mind kids learning to code like it's 1995 again.

What's wrong with coding like it's 1995? Did I miss some major paradigm shift in programming?

Re: How we built a P2P file sharing network from scratch

#15

Earlier quoted context omitted.

Why? It doesn't seem like time-sensitive content. Should any link to tutorials or documentation also have a date on it?

Perhaps yes. Unless you don't mind kids learning to code like it's 1995 again.

An article titled "Teaching kids to code like it's 1995" would definitely spark my interest. Somebody please write that!

Re: How we built a P2P file sharing network from scratch

#16

Is it P2P but not decentralized because clients get metadata from supernode1.xorro-p2p.com?

They have a DHT; I think the supernode is just for bootstrapping. Similarly all Bittorrent clients start with a hardcoded list of peers.

Re: How we built a P2P file sharing network from scratch

#17
post #11

Earlier quoted context omitted.

Perhaps yes. Unless you don't mind kids learning to code like it's 1995 again.

What's wrong with coding like it's 1995? Did I miss some major paradigm shift in programming?

#1 Thanks to the internet, Async has because far more important to understand. #2 CPUs have gotten waaay faster and ram is now plentiful so there's little need for heavy optimization in most cases. #3 you're almost guaranteed to have at least 2 core available so multi-threading is a real option now. #4 monolithic code bases are the devil now #5 some 3rd gen and I think all 4th gen laungages include automatic garbage collections so memory allocation is not a concern for most people

Ok I guess none of those are huge paradigm shifts but in many ways we've gotten pretty spoiled over the past 25 years. A lot of the old struggles have been mitigated at the expense of a lot of cpu cycles. NPM is a great example of the unneeded waste that has accumulated over that time.

We went from think about every byte to just tossing everything in and letting a transpiler remove all the fluff. You can certainly make the argument that computers should do that work but I think theres a middle ground between neurotic 1995 and kitchen sink 2019 that we'll hopefully eventually find.

Re: How we built a P2P file sharing network from scratch

#20
post #11

Earlier quoted context omitted.

What's wrong with coding like it's 1995? Did I miss some major paradigm shift in programming?

#1 Thanks to the internet, Async has because far more important to understand. #2 CPUs have gotten waaay faster and ram is now plentiful so there's little need for heavy optimization in most cases. #3 you're almost guaranteed to have at least 2 core available so multi-threading is a real option now. #4 monolithic code bases are the devil now #5 some 3rd gen and I think all 4th gen laungages include automatic garbage…

Hey, nice points. Really, no sarcasm, I wholeheartedly agree. However you, like me, seem to disagree with our parent comment - that coding 1995 was not really that bad...
Post reply on HN