Well, this takes me back. I’ve probably spent hours staring at the download queue in eDonkey then Overlord and eMule. I might spin this up just for old time’s sake.
This is the direction the web could have taken, but the venture funded ad giants grew platforms that engulfed all of publishing. The wrong focus for users, but great for making money for a select few.
A small p2p client like emule might provide a good alternative to RSS and Reddit if the only things shared are articles and comments. I'd love to see copy pasted HN articles with no paywall, distributed over P2P. Archive.is style. No Javascript nonsense. Text with minimal, clean markup. Fast and performant.
Cryptographically sign comments and share your interest graph - instant trust network and high signal social network.
Agreed, I wish we'd have gone down that road. How is it more efficient for my signals to go a thousand miles away and back just to talk to my neighbor across the street?
I've often thought, the web isn't designed for efficiency, it's designed for scale. Scale around big tech giants, to concentrate wealth, exploit gray areas, and magnify harm.
I used to love JavaScript, but when I look at what it's become, I'm glad I left it. We weren't doing the world any favors by contributing to a massive body of casual execution of untrusted code.
The web was a mistake, but I think the internet, good old TCP/IP, still has a few tricks up its sleeve for us yet.
> How is it more efficient for my signals to go a thousand miles away and back just to talk to my neighbor across the street?
For that, your neighbor would have to run their own infrastructure, keep it updated, moderated against spam, hackers and whatnotelse. And most people don't have the resources to do that - time, money or a DSL uplink with at least something halfway decent upstream bandwidth.
> I've often thought, the web isn't designed for efficiency, it's designed for scale.
Not intentionally designed. But the economies of scale always win in the end - agriculture, manufacturing, even education. As a conglomerate, the Internet as a sum of actors has always converged to see inefficiencies as damage - and route and work around them.
Superobscure stuff. The thing with the ed2k network is: people tend to share whole directories of their content.
If you want to find a documentary that ran once on TV, is not available commercially anywhere, you might have a chance that amule will find it. You may have to run your client for weeks because it goes offline from times and will only give you a few kb/s download rate, but you may get stuff you just cannot get otherwise.
Fundamentally ? Not too much. AMule and other similar clients use both the eDonkey and Kademlia protocols. Kad is a DHT too, only taking different approaches to it than torrents do. Part of it includes it being searchable. Otherwise, similar P2P with buckets and downloading files in verifiable chunks.
The only proof you get from BitTorrent being more popular than eMule is that BitTorrent is more popular than eMule. Doesn't mean eMule doesn't work for what it's supposed to do. "Crappy results" is at least possible with eMule while BitTorrent needs to rely on centralized tracking. That eMule still exists, is being used and no one has been able to stop it is proof to me that it does work.
Every single file sharing client other than bittorrent had search. BitTorrent easily beat them all. You can argue if you want that this is a pure coincidence, but it's not going to be very convincing.
Why do you think decentralized technologies still help people pirate software to this day while centralized services eventually gets shut down? The amount of trackers that has disappeared over the years is huge, and some of them quite sad as well. What.CD would be a famous example where we basically had a enormous backup of long lost music in the world. What.CD disappeared when French authorities managed to figure out where the servers were and got permission to raid them.
Same thing happened with TPB a couple of times, but amazingly it's still standing. We can only hope it remains so, but in the future we might not be as lucky. Not to say that plenty countries block access to TPB via having the ISPs block DNS/IP to known addresses.
All of these things wouldn't have been a problem if the search index was also decentralized.
You had a gigantic list of servers and didn't have to care if law enforcement managed to ban a few dozen? You could also search for anything shared by your peers. Major downside was that it wasn't curated and files tended to be shared under a dozen different names, so just because you think you were downloading Debian.iso didn't mean that it wasn't xXx_AnalCompilation5_xXx.avi, you had to check which name was the most frequent to be sure.
Popularity has absolutely nothing to do with centralized search sites. Nothing technical ever prevented "ed2k-based Pirate Bays" to take off the same way, it's just that eMule included search so sites were never required
15 years ago, when I was using eMule, I made the switch to Torrent because Torrent was way faster for me. That's all.
ShareReactor was a user submitted index of content on the eDonkey network and its layout and functionality were clearly an influence on the not-yet-created Pirate Bay: https://en.wikipedia.org/wiki/ShareReactor
When SR was at its peak, that was also ed2k's peak. In my teen and possibly post-teen years, I managed to upload some hard to find (pirated) pieces of software on the network and indexed on SR.
A quick search some months back showed there were still some of them available (via a tag I used) which was a pretty fun discovery considering they're almost 20 years old.
There used to be a direct link to the download as well as a link that opens a mirror selection dialog on the countdown page. Is that visible somewhere?
Edit: ah, these are now behind the "Problems downloading?" button.
Maybe it's blocked because one of your browser extensions has sourceforge on their blacklist. At least that is what happened to me in the past. I decided to leave it blocked and downloaded somewhere else.
