Couldn't you just use Facebook's "Fetch as FacebookBot" feature and just fetch the pages yourself? Example: https://developers.facebook.com/tools/debug/og/echo?q=https%... I was thinking about doing this since on my network Facebook access is free but Internet isn't. (2nd world country, but barely)
To me, the worst part of this is that there are places where Facebook is free but you need to pay for an internet connection. If the average user can only access the mindless corporate money-making machines and not the real internet, we are moving in the wrong direction.
They went from having no free internet to having Free facebook + other core sites like wikipedia through Facebook's internet.org, but it's moving in the wrong direction?
>If the average user can only access the mindless corporate money-making machines and not the real internet
They can, but it's not free, and a lot of people have no other option but to use what is free. This corporate money-making machine just provided pure value to a ton of people, there's no downside to having this additional option available.
there's no downside to having this additional option available.
Sure there's a lot of value in communicating with other Facebook users, but if you can't see the possible downside of having a large group of people's first and only connection being through a corporate-controlled view of the world, you're not using your imagination.
Wouldn't it prevent people interested in creating an ISP in any area that has "free" facebooknet? Especially if the services provided by facebooknet is most of which people would want with an ISP anyway?
So not only does it prevent competitor websites, but it stops competitive ISPs, too. It's especially bad if this coporate network is censoring information. No opportunity to even discover non-filtered information outside the info bubble. This triple whammy is the consequence for those involved or not.
Are those competitive ISPs going to give the audience we're talking about access to the net for free as well? Because for a lot of people in this world, they will never be connected to the net unless they don't have to pay; they simply can't afford it.
That's why we need government or non-profit programs, giving general access to the net. Because only then it is a level playing field, and the next Facebook can be from Brasil.
This is only an intentional effort to control and shut out competition (e.g. future Facebook competitors) and has the risk of killing local ISPs as a by-product.
tl;dr It only serves Facebook. Wikipedia et al. are just thrown into the mix to make it look like a humanitarian effort.
Why would they need to be "free?" Not everyone needs personal internet to benefit from a free (as in freedom) internet being available in their area. Unfiltered information spreading in a community is almost guaranteed if it exists.
How on earth are we going to get them innovative/affordable internet solutions if no one can compete to get there?
It would be somewhat contrived, but you could argue that the shown ability of Facebook to influence the moods of a population via filtering content could be used to oppress underprivileged people.
Yes, absolutely, and it's not a new idea, e.g. "sending shirts destroys local textile economies by flooding the market with free goods and undercutting local t-shirt producers" [1].
This allows those communities to dedicate resources that would have gone towards shirt manufacturing into something more valueable. Textile economies are a mostly waste of human ingenuity and time, and have been since the Jacquard Loom at least.
Free Facebook means the ability to send and receive messages with ~1 billion people at no cost. That's hugely valuable to a lot of people without the money for a general-purpose Internet connection.
Back in the day when Skype was a pretty cool peer-to-peer encrypted mesh network, that successfully evaded firewalls and traffic analysis, I wrote a tool to proxy Internet traffic ontop of the Skype mesh. I was quite young when I wrote it, but I managed to dust off an old demo screencast I made (and convert the swf to an mp4), which gave me a laugh. If you're curious:
I don't know the specifics about how the surveilance is implemented, but embedding encrypted payload in benign files is a neat idea to bypass internet censorship, at least in principle.
That's how the Tor network manages to bypass the Great Firewall. Google "Tor Bridges" if you want to get some fine examples. I think the newest tor bridge can smuggle the traffic over Google servers.
Humanity needs steganography-as-a-service. But think how difficult it is to implement in a regime where all networks are monitored and even Android phones are assumed to be backdoored. Most services will re-encode data during transmission, potentially munging whatever secret happens to be embedded. And more generally, its the metadata that often de-anonymizes, not the data itself. If all the dissidents one day begin communicating via Lutheran Insult memes, its simply becomes a matter of picking one up and applying "rubber-hose" cryptanalysis to put an end to the party.
This is really cool! In so many 2nd and 3rd world/poor countries internet is a luxury, as the price of bandwith is so high. Philippines for instance, a country that mainly provides connectivity through limited 3G and LTE, have a monthly price of about 30 beers from the local bar. In comparison, I pay around the price of 4-5 local beers, for 100/100 fiber.
Its sucks so hard that people cant afford to get knowledge from internet. Especially considering they would benefit greatly as schoolbooks are expensive and in some countries, censored.
Hack author here, in my country, Paraguay, you get 10 MBPS for ~140 U$ per month.
I discovered the Internet through my father, because he is a journalist and the press had really early access to this kind of tech, 20 years later I'm still crazy about it.
Thanks for the comment!
Yeah, I've played with that a bit..it was fun. It worked in airports, as DNS was available, even though you had to pay for actual surfing. Not surprisingly, it was quite slow.
I've heard this is also true for Twitter and a few other social networking sites, so you could potentially tunnel your traffic through them too.
However, these tunnels are going to be relatively low-bandwidth (especially if they start throttling it) and high-latency, so the "unlimited" would in reality be "how much can you transfer continuously in a month?"
Interesting, but wouldn't you need a server somewhere to act as the other end of the tunnel, for which you'd have to pay for hosting and/or bandwidth anyway?
But hosting and bandwith for servers is extremely cheap if you don't care about reliability and trendyness. You can get a vps with 1Tb bandwidth/month for $4 (probably even less if you look around more than I did). Try getting mobile bandwidth for anything near that price.
I am pretty much a beginner dev, so can anyone briefly explain how this works? It's not really clear to me from the code.
It's mentioned in the README that packets are being sent as base64. Let's say I would like to send a GET request to google.com. How does Facebook Chat serve me with google's page?
I was just reading about gumbo parser yesterday! h/t @nostrademons
Very nice hack... its just fun to do something like this.
Reminds me when I used to fiddle w/ "phreaking" when it was really easy. I didn't think about ethical implications - just got a really big thrill when I got a free call.
Thanks! I'm glad to see it used for things...when I wrote it (and decided to open-source it), it was very much a "Well, I have one immediate need for this, but it seems like people could do a lot with a robust HTML5 parser with a simple API."
What makes this interesting is the fact that Facebook chat traffic is subsidized by Facebook and free of charge in some countries as part of their anti-net-neutrality "Internet.org" initiative. This project allows you to hijack the Facebook subsidy for purposes they didn't intend. Of course, if it becomes popular then Facebook will have to block it, emphasizing the self-serving commercial nature of the subsidy.
Internet.org might not be net neutral, but that doesn't make it bad. Much of the world isn't the US, and differently structured markets need to be considered differently.
So if you are in an impoverished area your packets are different from first world packets, magically making it harder to send them to 207.241.224.2 (archive.org) than 173.252.120.6 (facebook.com)?
No, that is not the case. Instead, Facebook will pay for you to get Internet if that is Internet that only lets you access Facebook. And that is not an Internet at all. The infrastructure, the physical reality, means that if you have any connection, you can have all connections, and in these circumstances only profit hungry greedy monsters would consider effectively blacklisting every site but their own to guarantee them revenues to show to shareholders for providing network connectivity to you - or I guess it should be more appropriately called Facebook connectivity.
The only bad thing they do is call it "internet". It's not internet access. Free access to things like wikipedia is a good thing, as long as it's not taking money away from actual free internet initiatives.
Price gouging actual internet access would also be bad, but that would be someone else doing it, and I don't think it's happening here.
Did you know that yombinator.com o ly allows you to transmit comments on HackerNews posts, subject to heavy moderstion, not free and open internet speech?
Here's one telecom's implementation of this proprietary Internet.org app showing free tier fb/messenger the photos are still displayed just obscured until you pay for a data plan (notice it uses free.facebook.com which is complicit in enforcing this. https://youtu.be/sUGZdFXs22o?t=2m16s
Guess no embedding creative commons university pdfs in wall posts.
I did something very similar to this last summer, written in Python. No idea what state the code is in at this point because it's been months, but here's the code I had sitting around: https://gist.github.com/schlarpc/86e20aa3f3aaf3d78269
With a built-in encryption layer (on top of eventual HTTPS) it could be a useful "hack" for the kinds of situations mentioned in the readme. I wonder how facebook and the telecos will feel about this. All in all, a cool concept but unlikely to work in practice if either facebook or the telecos do not want to.
It would be cool if a company such as facebook supported this. Seeing them as an enabler for free/uncensored internet, given that the country was chosen for a internet.org campaign.
If Facebook wanted to support it they would probably just implement a standard proxy server that would be more efficient. If anybody can handle the traffic it would be Facebook, but I bet their chat services are not really tuned for this kind of usage.
I tried something like this over gchat but they throttled pretty soon so didn't work well. But it worked for me over Firebase. Here is simple project to demo how to do VNC over Firebase: https://github.com/vskarine/fireport
> Gumbo was initially designed for a product that worked with trusted input files only. We're working to harden this and make sure that it behaves as expected even on malicious input, but for now, Gumbo should only be run on trusted input or within a sandbox.
That's mostly historical at this point - Gumbo passed a security review as of 0.9.1, has had several rounds of fuzz testing, and is used by a bunch of other libraries now. It's been struck out of the README at HEAD, though I think it was listed in several earlier versions' READMEs.
This is the kind of hacking I find to be abusive. Finding clever uses for things is fine, but not when it's misuse with potential for harm. This reminds me of when someone exploited TinyURL to make TinyDisk.
I think this is kind of abusive of FB's infrastructure TBH. It must be hard to keep real-time messaging up for ... okay, I was going to say for 1 billion users, looked it up and it's 1.3 billion monthly actives - 890 million daily actives (Jan 2014 figure). So tell you what, on second thought I think they can handle your data.
however there's a good chance you'll cause some infrastructure problems since I'm sure there are assumptions baked in, regarding how often a user will send new messages, queuing, caching, etc. so I'm still not totally sure how I feel about this.
at the end of the day FB is just a php script. i wouldn't be surprised if you break it.
Or limit how much data you can send over chat. Say max 10 bytes per second sustained (higher burst) allowed. Should be enough for normal chat, while limiting infrastructure problems.
I totally support this "abuse" of their infrastructure if it makes them realize that giving free access to their own social network via Internet.org isn't a charitable endeavour.
Couldn't you just use Facebook's "Fetch as FacebookBot" feature and just fetch the pages yourself? Example: https://developers.facebook.com/tools/debug/og/echo?q=https%... I was thinking about doing this since on my network Facebook access is free but Internet isn't. (2nd world country, but barely)
It looks like it is more general purpose, than just html browser.
Potentially a problem that internet = web, at zeroth approximation.
How does they rate limit this thing?
To me, the worst part of this is that there are places where Facebook is free but you need to pay for an internet connection. If the average user can only access the mindless corporate money-making machines and not the real internet, we are moving in the wrong direction.
What constitutes the real internet - the rest of it ?
They went from having no free internet to having Free facebook + other core sites like wikipedia through Facebook's internet.org, but it's moving in the wrong direction?
>If the average user can only access the mindless corporate money-making machines and not the real internet
They can, but it's not free, and a lot of people have no other option but to use what is free. This corporate money-making machine just provided pure value to a ton of people, there's no downside to having this additional option available.
there's no downside to having this additional option available.
Sure there's a lot of value in communicating with other Facebook users, but if you can't see the possible downside of having a large group of people's first and only connection being through a corporate-controlled view of the world, you're not using your imagination.
Can you actually make the argument that access to Facebook/Wikipedia/etc is worse than no Internet whatsoever?
Wouldn't it prevent people interested in creating an ISP in any area that has "free" facebooknet? Especially if the services provided by facebooknet is most of which people would want with an ISP anyway?
So not only does it prevent competitor websites, but it stops competitive ISPs, too. It's especially bad if this coporate network is censoring information. No opportunity to even discover non-filtered information outside the info bubble. This triple whammy is the consequence for those involved or not.
Are those competitive ISPs going to give the audience we're talking about access to the net for free as well? Because for a lot of people in this world, they will never be connected to the net unless they don't have to pay; they simply can't afford it.
That's why we need government or non-profit programs, giving general access to the net. Because only then it is a level playing field, and the next Facebook can be from Brasil.
This is only an intentional effort to control and shut out competition (e.g. future Facebook competitors) and has the risk of killing local ISPs as a by-product.
tl;dr It only serves Facebook. Wikipedia et al. are just thrown into the mix to make it look like a humanitarian effort.
Why would they need to be "free?" Not everyone needs personal internet to benefit from a free (as in freedom) internet being available in their area. Unfiltered information spreading in a community is almost guaranteed if it exists.
How on earth are we going to get them innovative/affordable internet solutions if no one can compete to get there?
It would be somewhat contrived, but you could argue that the shown ability of Facebook to influence the moods of a population via filtering content could be used to oppress underprivileged people.
Yes, absolutely, and it's not a new idea, e.g. "sending shirts destroys local textile economies by flooding the market with free goods and undercutting local t-shirt producers" [1].
1. http://freakonomics.com/2011/02/15/what-happens-to-all-those...
This allows those communities to dedicate resources that would have gone towards shirt manufacturing into something more valueable. Textile economies are a mostly waste of human ingenuity and time, and have been since the Jacquard Loom at least.
Within our lifetimes, this reasoning will apply to more professions than you might think. "Dignity is more important than wealth" and all...
Free Facebook means the ability to send and receive messages with ~1 billion people at no cost. That's hugely valuable to a lot of people without the money for a general-purpose Internet connection.
And it will destroy the prospects of a healthy local internet economy where everyone can publish information or compete on a level playing field.
Back in the day when Skype was a pretty cool peer-to-peer encrypted mesh network, that successfully evaded firewalls and traffic analysis, I wrote a tool to proxy Internet traffic ontop of the Skype mesh. I was quite young when I wrote it, but I managed to dust off an old demo screencast I made (and convert the swf to an mp4), which gave me a laugh. If you're curious:
http://data.zx2c4.com/skypeproxydemo.mp4
That is very cool. Does Skype not have a direct peer to peer connection anymore?
Supposedly since the Microsoft buy-out, everything is gradually becoming centralized.
I've been wondering if something like this could be helpful to breach the great firewall.
Find some high traffic site in China and smuggle in content via various site features such as chat, messaging, forums, etc.
The content could be obfuscated to appear benign while embedding hidden data in text, images, or binary attachments.
I don't know the specifics about how the surveilance is implemented, but embedding encrypted payload in benign files is a neat idea to bypass internet censorship, at least in principle.
Not exactly the same, but one of the ways to bypass: https://realworldcrypto.files.wordpress.com/2013/06/shrimpto...
Inconvenient Message Detection: https://github.com/DickingAround/InconvenientMessageDetectio...
FYI github renders readme.txt as preformatted code, but your readme.txt is not preformatted, so it is very hard to read (one line per paragraph).
That's how the Tor network manages to bypass the Great Firewall. Google "Tor Bridges" if you want to get some fine examples. I think the newest tor bridge can smuggle the traffic over Google servers.
Links for whoever is interested:
meek-google: https://trac.torproject.org/projects/tor/wiki/doc/meek#Googl...
meek-amazon: https://trac.torproject.org/projects/tor/wiki/doc/meek#Amazo...
This looks similar to the "IPv6 over Social Networks" RFC, published 2009-04-01:
https://tools.ietf.org/html/rfc5514
Wonderful proof-of-concept!
Humanity needs steganography-as-a-service. But think how difficult it is to implement in a regime where all networks are monitored and even Android phones are assumed to be backdoored. Most services will re-encode data during transmission, potentially munging whatever secret happens to be embedded. And more generally, its the metadata that often de-anonymizes, not the data itself. If all the dissidents one day begin communicating via Lutheran Insult memes, its simply becomes a matter of picking one up and applying "rubber-hose" cryptanalysis to put an end to the party.
This is really cool! In so many 2nd and 3rd world/poor countries internet is a luxury, as the price of bandwith is so high. Philippines for instance, a country that mainly provides connectivity through limited 3G and LTE, have a monthly price of about 30 beers from the local bar. In comparison, I pay around the price of 4-5 local beers, for 100/100 fiber.
Its sucks so hard that people cant afford to get knowledge from internet. Especially considering they would benefit greatly as schoolbooks are expensive and in some countries, censored.
Hack author here, in my country, Paraguay, you get 10 MBPS for ~140 U$ per month. I discovered the Internet through my father, because he is a journalist and the press had really early access to this kind of tech, 20 years later I'm still crazy about it. Thanks for the comment!
Wow! Thats really expensive! Hard do access stuff often sparks creative ideas and solutions :) Keep on hacking!
~70USD for a symetric 3Mbps static-IPv4/6 connection in Argentina.
I believe asymetric normal connections cost about ~40USD.
A beer is about 4USD, IIRC, if that helps the comparison.
Facebook limits number of messages you can send. After passing some secret threshold, your profile would be put in 'restricted' mode.
I once read a story about tunneling IP over DNS. Something about free Internet/DNS for Microsoft-Update. But I can't find it.
You are thinking of iodine.
http://code.kryo.se/iodine/
No, what I read was from the 90s. But the guy probably did what iodine does.
I remember possibly the same article - it was a guy who noticed that airport wifi allows DNS even if you haven't paid.
Yes, they have to because of how Windows detects connectivity (NCSI).
Perhaps http://dnstunnel.de/ ?
Yeah, I've played with that a bit..it was fun. It worked in airports, as DNS was available, even though you had to pay for actual surfing. Not surprisingly, it was quite slow.
Some mobile operators offer unlimited Facebook and whatsapp traffic doesn't count towards the monthly data transfer limit.
I've heard this is also true for Twitter and a few other social networking sites, so you could potentially tunnel your traffic through them too.
However, these tunnels are going to be relatively low-bandwidth (especially if they start throttling it) and high-latency, so the "unlimited" would in reality be "how much can you transfer continuously in a month?"
Interesting, but wouldn't you need a server somewhere to act as the other end of the tunnel, for which you'd have to pay for hosting and/or bandwidth anyway?
Sure, you need that end!
But hosting and bandwith for servers is extremely cheap if you don't care about reliability and trendyness. You can get a vps with 1Tb bandwidth/month for $4 (probably even less if you look around more than I did). Try getting mobile bandwidth for anything near that price.
I am pretty much a beginner dev, so can anyone briefly explain how this works? It's not really clear to me from the code.
It's mentioned in the README that packets are being sent as base64. Let's say I would like to send a GET request to google.com. How does Facebook Chat serve me with google's page?
As far as I understood:
- They are using virtual network kernel device (TUN/TAP, https://github.com/matiasinsaurralde/facebook-tunnel/blob/ma... )
- encode it as base64
- communicate with the http/web facebook client https://github.com/matiasinsaurralde/facebook-tunnel/blob/ma...
Between that they do authentication things at facebook with "gumbo". And later on everything the other way round at the server side.
So if you GET google.com your traffic goes through a virtual ethernet like device. https://www.kernel.org/doc/Documentation/networking/tuntap.t...
There is no cryptography involved, they could try to use openvpn to enable this feature. https://community.openvpn.net/openvpn/wiki/BridgingAndRoutin...
I was just reading about gumbo parser yesterday! h/t @nostrademons
Very nice hack... its just fun to do something like this.
Reminds me when I used to fiddle w/ "phreaking" when it was really easy. I didn't think about ethical implications - just got a really big thrill when I got a free call.
Thanks! I'm glad to see it used for things...when I wrote it (and decided to open-source it), it was very much a "Well, I have one immediate need for this, but it seems like people could do a lot with a robust HTML5 parser with a simple API."
interesting given FB keeps chat history forever
I wonder if they have any flood protection...
Also HTTPS packets should be quite safe.
It'd be more interesting if this was done over Google's chat, since Chrome trusts the Google Internet Authority cert...
What makes this interesting is the fact that Facebook chat traffic is subsidized by Facebook and free of charge in some countries as part of their anti-net-neutrality "Internet.org" initiative. This project allows you to hijack the Facebook subsidy for purposes they didn't intend. Of course, if it becomes popular then Facebook will have to block it, emphasizing the self-serving commercial nature of the subsidy.
Internet.org might not be net neutral, but that doesn't make it bad. Much of the world isn't the US, and differently structured markets need to be considered differently.
So if you are in an impoverished area your packets are different from first world packets, magically making it harder to send them to 207.241.224.2 (archive.org) than 173.252.120.6 (facebook.com)?
No, that is not the case. Instead, Facebook will pay for you to get Internet if that is Internet that only lets you access Facebook. And that is not an Internet at all. The infrastructure, the physical reality, means that if you have any connection, you can have all connections, and in these circumstances only profit hungry greedy monsters would consider effectively blacklisting every site but their own to guarantee them revenues to show to shareholders for providing network connectivity to you - or I guess it should be more appropriately called Facebook connectivity.
They're not blacklisting anything.
The only bad thing they do is call it "internet". It's not internet access. Free access to things like wikipedia is a good thing, as long as it's not taking money away from actual free internet initiatives.
Price gouging actual internet access would also be bad, but that would be someone else doing it, and I don't think it's happening here.
Did you know that yombinator.com o ly allows you to transmit comments on HackerNews posts, subject to heavy moderstion, not free and open internet speech?
I didn't know that infrastructure was free to build and maintain. I wonder why WorldCom and Global crossing went bankrupt, if they had no expenses.
I think that internet access to a site like Wikipedia is incredibly valuable to those with no internet access.
If Facebook wants to pay for that, and the cost is that they also get access to Facebook then my view is that is a net good.
Here's one telecom's implementation of this proprietary Internet.org app showing free tier fb/messenger the photos are still displayed just obscured until you pay for a data plan (notice it uses free.facebook.com which is complicit in enforcing this. https://youtu.be/sUGZdFXs22o?t=2m16s
Guess no embedding creative commons university pdfs in wall posts.
I did something very similar to this last summer, written in Python. No idea what state the code is in at this point because it's been months, but here's the code I had sitting around: https://gist.github.com/schlarpc/86e20aa3f3aaf3d78269
Really cool, I'll give it a try soon!
With a built-in encryption layer (on top of eventual HTTPS) it could be a useful "hack" for the kinds of situations mentioned in the readme. I wonder how facebook and the telecos will feel about this. All in all, a cool concept but unlikely to work in practice if either facebook or the telecos do not want to.
It would be cool if a company such as facebook supported this. Seeing them as an enabler for free/uncensored internet, given that the country was chosen for a internet.org campaign.
If Facebook wanted to support it they would probably just implement a standard proxy server that would be more efficient. If anybody can handle the traffic it would be Facebook, but I bet their chat services are not really tuned for this kind of usage.
It would certainly be interesting if facebook started to provide proxy access
They sure as hell would love to see all of your traffic. Facebook wants to be your ISP, no actually, Facebook wants to be your internet.
The question then is, do you trust facebook more than your isp?
For those of us with Comcast, it's a pretty close race...
They're also basically the ones responsable for restricting that same internet access.
If they wanted you to access those other sites, they'd have given you access to them.
But why would they? As-is, they have no competition. "when there's only one choice, there's only one vote!"
I tried something like this over gchat but they throttled pretty soon so didn't work well. But it worked for me over Firebase. Here is simple project to demo how to do VNC over Firebase: https://github.com/vskarine/fireport
Dear Australian Government,
Please add this to your list of reasons why your wrong headed "meta"-data retention plan can not work.
Rory
Uses Gumbo... from the README
> Gumbo was initially designed for a product that worked with trusted input files only. We're working to harden this and make sure that it behaves as expected even on malicious input, but for now, Gumbo should only be run on trusted input or within a sandbox.
That's mostly historical at this point - Gumbo passed a security review as of 0.9.1, has had several rounds of fuzz testing, and is used by a bunch of other libraries now. It's been struck out of the README at HEAD, though I think it was listed in several earlier versions' READMEs.
While I find this project awesome, but I don't think telecos or even facebook will allow this in long run. I think that it even breaks facebook TOS.
https://github.com/pasis/pppoat
Base64 encoding is such a useful tool. I used it to transfer binary over IRC once.
Why not just use DCC?
What is DCC, in this context?
I tried to find it in google, but the acronym is very ubiquitous.
First hit of this query: https://www.google.com/search?q=irc+dcc
DCC relies on all parties to have a specific port open and even then if the connection drops all hell breaks loose.
That's a really neat hack, love it.
Wow didn't see this was so old
We need screenshots here! :D
9984894172
Nice hack
nice
This is the kind of hacking I find to be abusive. Finding clever uses for things is fine, but not when it's misuse with potential for harm. This reminds me of when someone exploited TinyURL to make TinyDisk.
No, please no this is terrible.
Why?
rush I jadhav
I think this is kind of abusive of FB's infrastructure TBH. It must be hard to keep real-time messaging up for ... okay, I was going to say for 1 billion users, looked it up and it's 1.3 billion monthly actives - 890 million daily actives (Jan 2014 figure). So tell you what, on second thought I think they can handle your data.
however there's a good chance you'll cause some infrastructure problems since I'm sure there are assumptions baked in, regarding how often a user will send new messages, queuing, caching, etc. so I'm still not totally sure how I feel about this.
at the end of the day FB is just a php script. i wouldn't be surprised if you break it.
> at the end of the day FB is just a php script
Facebook Chat isn't. It's primarily Erlang/C++.
https://www.facebook.com/note.php?note_id=14218138919
Facebook chat stopped being Erlang quite some time ago.
"FB is just a php script" was just a joke :)
Maybe if it becomes popular, it'll make facebook rethink wanting to store everyone's messages until the end of time.
Or limit how much data you can send over chat. Say max 10 bytes per second sustained (higher burst) allowed. Should be enough for normal chat, while limiting infrastructure problems.
If facebook can see this traffic, they would love it, not discourage it.
I think a good implementation of this would encrypt the traffic between the proxy server and the client.
Hahahah nope.
It's one of the more useful ways facebook infrastructure is put to work.
I totally support this "abuse" of their infrastructure if it makes them realize that giving free access to their own social network via Internet.org isn't a charitable endeavour.