Tribler. Almost so cool! What I would do:

So I finally made it through to the Tribler website and I think it's *almost* really cool.  Don't get me wrong, I installed the app and it seems to do interesting stuff.  But I think they're biting off more than they can chew.  Basically, I'd layer the content-acquisition experience as follows:

Social: The experience of sharing and learning about new content
Index: Determining who has a given piece of content
Transport: Getting content from somebody else
Tribler tries to vertically integrate the entire stack, and they do a decent job, but it's just too much for one application.  It's like coming out with Prodigy today and trying to complete with "The Internet".  If Tribler came out before ThePirateBay then it'd be something.  But then it'd be called Kazaa.

No, instead I think they should do something less.  Indeed, the whole brilliance of BitTorrent was that it *didn't* do it all.  BitTorrent does nothing but standardize transport, which enabled a huge diversity on the upper layers.  Doing *less* than Kazaa is what made BitTorrent succeed.

So if BitTorrent is on one side, and Kazaa on the other, what should Tribler do?  I'd say stick with the old saying "only innovate one thing at a time" and just integrate the index into the transport layer, and then call it day.  Leave ThePirateBay and everyone else to figure out the social layer, recommendation engine, and just focus on getting rid of the tracker.  How?

I think they should create a tool where you can type in the SHA1 sum of any piece of content,* and it'll download it.  That's it.  Nothing more.  Think of what kind of interesting applications could come about if the recommendation engines and such didn't have to host torrent files, and didn't need to take on the risk of trackers!
* In practice it'd probably be a SHA1 sum of a list of SHA1 sums.
Because really, search engines like TPB only make sense when "what we have" is a small subset of "what you want", and when there is no general consensus on what a "good" copy of each thing is.  In this environment, you literally need to "search" for something you want.

But in practice, TPB has pretty much everything, and for each thing there's usually one version that pretty much everybody uses (ie, the version with all the seeds).  So there's usually little "searching" involved.  More often than not, just type in the name of what you want, pick the one with the highest number of seeds, and you're done.

Given this reality, it probably makes more sense to ditch the unstructured search interface and go to a structured "table of contents" indexing "all music" and "all movies", with the "best" version of each one given front and center.  Then you just click a "p2p://<sha1sum>" link and your client connects to the cloud and pulls it down.  (And all the recommendation engines would just layer atop that.)

Accordingly, TPB shouldn't copy Google: they should copy IMDB.  Create a comprehensive library of all content, and provide one recommended copy of each (or maybe a selection of encodings: iPod, HD, etc).

As for how to manage spam and such, again, embrace reality.  In theory, anybody could post a good copy of anything, and nobody is more trustworthy than any other.  In practice, there's usually one guy who is the uber-fan of a particular type of content, and that person posts all the good stuff.  Why not just explicitly recognize that uber-fan by making him moderator of the corresponding ThePirateIMDB wiki-like page for that band/TV-show/movie/etc.  Then you stop moderating content on a piece-by-piece basis, and start moderating on a curator-by-curator basis.  That uber-fan needn't be the one to post all good content (though in practice he probably would); others could recommend content to him and he'd sift through and find the good stuff.

Anyway, that's all just dreamy "what I'd do if I were a megapirate" talk.  Somebody's going to do it, and it probably won't be me.  I highlight it to make it clear that this is inevitable.  If this scenario frightens you, then your instinct is right: be afraid.  The future is coming, and if it's not to your liking, then now's the time for some deep introspection because there ain't nothing you can do to stop it.


Adam Fisk said...

Interesting post, David. One of the things I've integrated into LittleShoot is similar to what you're suggesting, and similar to good old RedSwoosh for that matter. Basically, you can send someone a link like:


Take out the extra cruft, and you're just sending a sha-1. The site detects if you've got the app and redirects to localhost if you do. LittleShoot of course looks up all the sources and grabs it from all of 'em.

Anyhow, Tribler's neat, but I agree simpler is the way to go.

David Barrett said...

Oooh, awesome. I didn't know Littleshoot could do that. Very cool!

Now you just need to find someone to start up the next generation of "tracker" site that is like a Wikipedia/IMDB, and have them Littleshoot links to the content. There must be IMDB competitors out there willing to partner up.

Llew said...

Could SoundUnwound be used for that?

David Barrett said...

Yes, that's a great idea. Granted, because it's run by Amazon they'd have an interest in stripping out any pirate potential because it's a competitive threat.

I think it might be easier to start from "scratch" by just going through all the TV shows listed in ThePirateBay and making an awesome site focused on that. (Use copy/paste from IMDB liberally.) Don't try to be everything all at once, just be *the* best place to download TV shows. Gradually expand your coverage of the most common TV series and completely lock up that end of the market.

Once that's refined, open up the floodgates to movies by just having a "whatever" section. It'll initially start out pretty rough with a lot of spam, but eventually figure out who the good moderators are and gradually sift through the muck an find the good stuff.

- Jan 2014 (1) - Mar 2012 (1) - Nov 2011 (1) - Oct 2011 (1) - Apr 2011 (1) - Mar 2011 (3) - Feb 2011 (2) - Jan 2011 (9) - Nov 2010 (1) - May 2010 (1) - Mar 2010 (1) - Feb 2010 (1) - Jan 2010 (1) - Dec 2009 (1) - Nov 2009 (1) - Oct 2009 (1) - Sep 2009 (1) - Aug 2009 (2) - Jul 2009 (1) - Jun 2009 (4) - May 2009 (3) - Apr 2009 (3) - Mar 2009 (10) - Feb 2009 (5) - Jan 2009 (3) - Dec 2008 (5) - Nov 2008 (5) - Oct 2008 (5) - Sep 2008 (4) - Aug 2008 (5) - Jul 2008 (11) - Jun 2008 (8) - Feb 2008 (1) - Aug 2007 (1) -