The page says I have to login using a facebook account without any information on what it's supposed to do.

Either this is spam, either you'll have to put a page explaining what it does.

It uses facebook graph api calls to retrieve statuses and dynamically load the content from those statuses into the page. I didn't put a whole lot on the original page because my goal was to show how much of a site could be held in Facebook statuses. If you want to see what statuses will be loaded you can check them out at https://www.facebook.com/pages/SocialCDN/263306957124999

Ok, this looks like a nice hack :)

Here is a screenshot after you've logged in for the ones who wonder what it does : http://i.imgur.com/3ZvNa.png

I'd be much more interested in an explanation of the method rather than a demo of the concept. It sounds interesting, but I don't really see what I'd gain by logging in.

The demo is the explanation. I think it's clever.

I had a similar idea about storing data in slashdot comments as a "free S3" service (probably inspired by seeing other comments with random-looking garbage) but never got around to implement it.

(Partly inspired by an idea about using SMTP bounce mails as a temporary storage system, which I first saw in the book Silence On The Wire)

And a link to that particular SotW chapter, for the curious: http://lcamtuf.coredump.cx/juggling_with_packets.txt

it's a cute hack, but there is something uncomfortably parasitic about this whole class of "exploit" (building a fuse-based filesystem atop gmail was probably the most prominent example). like or hate these services, they have assumed the status of commons where lots of people are living, and abusing their resources merely raises the spectre of degrading everyone's service if it catches on.

(i agree that this particular hack is more cute and creative than actively damaging to facebook; i just dislike the general culture of appreciation for what is little more than overusing a free, shared resource just because you can)

Facebook is not a free shared resource. We pay for the service with our privacy. It's not money, but it is a tradable good.

that's just sophistry. as far as the people using facebook to socially network with their friends are concerned, it's a free resource that benefits everyone. trying to layer a cdn atop it benefits no one except the defectors getting free hosting[0], and hurts everyone else by degrading their qos[0].

[0] again, this is all in theory; as far as this particular hack goes i agree it's fairly harmless insofar as no one is really going to try hosting a website off fb updates.

The author has this page hosted from a micro EC2 instance, but they could just as easily serve it from S3, avoiding needing to run any servers / pay for anything except for the bandwidth fees (12c/gb).

Either way, it's a really neat hack.

Yeah I'm pretty new to the AWS stuff and am still running on my free usage tier. Originally I was running some php on the page, hence the server. Now it's all javascript based so the micro instance isn't needed but I just left it up as it's still free.

Super interesting hack!

I wonder if its be more flexible to try integration into something like Wordpress instead with Cloudflare running on top. This would get the flexibility of Wordpress for the data storage and utilizing Cloudflare for their CDN.

Cool idea. Is it allowed by Facebook's Terms of Service?

Why wouldn't it be? Maybe OP is a base64 junkie. I'm sure it'll get a chuckle from the graph team. It's not a if its gonna spark a revolution of sites hosted off fb. It has obvious limitations in spite of it's creativeness.

Pretty cool! You should think about show the explanation and what it actually does and have a demo section where the user optionally logs in with FB.

I changed the site a little to attempt to explain the method before login is required.

That's an awesome hack! Have you considered using other platforms to do it? like WordPress or something

its down. probably shouldn't have posted it to HN if you want it to live more than a few hours ;)

It's still up for me.

Clever. Does it still get cached?

Excellent hack!