You’re in your best chair listening to music when your phone rings. You open your hand, hear a friendly sound notification, then close your hand and move it downwards. The music volume follows your hand downwards. Would be awesome, right?
Here’s a sketch on how a gesture recorder for Kinect might look. Click to view in Google Docs.
Last month, we outlined an initial proposal to address the inadvertent sharing of User IDs (UIDs) via the HTTP Referral header. Further, we announced our intention to provide a unique, but anonymous mechanism that developers can use with third-party service providers. Over the past few weeks, we have received feedback from the developer community, which we have used to refine our proposals. Today, we are providing an updated proposal for handling inadvertent UID sharing and releasing a new mechanism for legitimate and anonymous sharing with third-parties.
Iframe POST Proposal
Currently, we pass iframe applications the ‘fb_sig_user’ query string parameter in the URL. This allows the application to identify the user and create customized, social experience. Due to the way browsers work, this information in the URL can be inadvertently passed in the HTTP Referrer header when someone clicks a link within the iframe.
Our initial proposal to address this issue used encryption as a means to protect against this inadvertent sharing, but still passed this encrypted UID in the URL. After talking with the community, we have updated our proposed solution to use a different mechanism that provides better protection for users while minimizing the impact on existing applications and eliminates the need to use encryption libraries.
In short, this new proposal embeds the UID in a HTTP POST body ensuring that it will not be exposed in any HTTP Referrer header whatsoever (encrypted or otherwise). We do this by creating a <form/> element targeted at the application Canvas URL:
<form target="canvas_iframe" action="http://example.com/" id="canvas_form">
<input name="fb_sig_user" value="1234" type="hidden" />
</form>
<iframe name="canvas_iframe"></iframe>
<script>
document.getElementById("canvas_form").submit()
</script>
This change will require minimal effort for developers and addresses the feedback that we have received to date. You can learn more about this change on the POST for Canvas page.
This is still a proposal from us and we welcome feedback on the approach. We have, however, implemented this change, so developers can understand how this proposal will impact their applications. You can test this change by turning on the “POST for Canvas (Beta)” migration on the Advanced tab of the Developer Application. We plan on gathering feedback over the next two weeks and then determine if/when to require this mechanism for all applications.
Third-Party IDs
Facebook has never and will never sell user information, and we expressly prohibit Platform developers from passing any data from Facebook to data brokers. There are, however, practical reasons that developers will need to share information with a legitimate third party. For example, a game developer might need a way to identify that a user has taken an action on a partner website (e.g., completed an offer) so that the person can receive virtual goods. For this reason, we have developed a mechanism for developers to obtain an anonymous but unique identifier for their users, which we call a third-party identifier.
Developers can obtain a third-party identifier through either the Graph API or FQL:
Graph API
Get the third-party identifier from my User object:
http://graph.facebook.com/dmp?fields=third_party_id&access_token=[access_token]
Get the third-party identifier for two users:
http://graph.facebook.com/?ids=dmp,vernal&fields=third_party_id&access_token=[access_token]
FQL
Get the third-party identifier from the user table:
https://api.facebook.com/method/fql.query?query=select third_party_id from user where username ="dmp"&access_token=[access_token]
Get the third-party identifier for two users:
https://api.facebook.com/method/fql.query?query=select third_party_id from user where (username ="dmp" or username="vernal")&access_token=[access_token]
These identifiers may be shared with permitted third parties, but like all data from Facebook, they are also prohibited from sharing with data brokers. We encourage developers to move to this mechanism quickly, and we will require it as of January 1, 2011.
If you have comments or questions about either the iframe POST proposal or our third-party identifier feature, please use the comments found below.
Mike Vernal is Director of Engineering for Facebook Platform.
Have you heard of CCN? CCN is an open-source implementation of “content-centric networking” or more commonly “named data networking.” It’s a technology being actively developed by the Palo Alto Research Center (PARC), formerly Xerox PARC, the birthplace of computing mainstays like the PC, Ethernet, laser printing and the graphical user interface.
So what’s CCN? It’s an alternative idea about how computer networking should work – and it could very well one day be the future of Internet communications, most importantly, mobile networks.
Sponsor
If testing out experimental, next-generation network architecture is the kind of thing that gets you excited about technology, then you have to get an Android phone. PARC has just announced that Android is now a supported platform for CCN testers.
To find out more about CCN’s development, progress and potential, we spoke to PARC’s Jim Thornton, project manager for CCN, and Business Development lead, John Tripier. Below is a summary of what we learned.
IP, or the Internet Protocol, is the backbone of today’s Internet. When it was invented, it was created with the idea of computers talking to each other, sort of like a telephone system for computer networks. But that was many, many years ago, in a much different world. Although the invention of IP has led to major innovations in terms of communications, it no longer represents the best technology for how we use the Internet these days.
If you ask people what they think about the Internet, or how they use it, they’ll tell you about watching YouTube videos, socializing on Facebook, checking email, etc. A lot of what people do involves viewing and accessing content. IP allows you to make a request for that content by typing in a URL (Update: as a commenter points out, I should clarify – the URLs are further up the stack, built on TCP/IP). The URL then resolves to an actual address on the Internet, that being the location of a physical server in a data center somewhere, connected to the Web. Of course, that’s a simple explanation – large sites have more than one server and more than a single IP address. But the concept is the same – the data travels from one endpoint to the other.
CCN, on the other hand, is an attempt to take what works about IP and modernize it for how we use the Internet now. Like the URL system built on top of IP, it too wants to introduce a simple way of asking for content. But instead of URLs, CCN uses names. The difference is that with CCN, the content being requested doesn’t have to come from only the originating server – it can be housed on other nodes on the Internet. CCN will deliver the content from the node closest to you. This could be a router maintained by your Internet Service Provider (ISP), for example, which would be much closer – physically – than a server in a data farm thousands of miles away.
Not only is the distance between content requester and content storage closer, it would also lessen the congestion on the network as a whole. Instead of everybody hitting up a single site for a popular item – say, a YouTube video of an adorable kitten, for example – they would just pull from the nearest copy.
Although CCN has multiple use cases – consumers, enterprise or mobile carriers could take advantage of CCN on their networks, it’s the mobile carrier application that’s of interest to us. Carriers have the greatest need for a solution like this, as we move into the new computing landscape where connected devices, smartphones, tablets and portable computers are the norm, and exist in greater number than the PC ever did.
The mobile carriers themselves are struggling to adapt, building out new network infrastructure, but not fast enough, it seems, to keep up with demand. Some have proposed measures to prioritize certain content streams over others, but this goes agains against one of the core principles Internet proponents hold dear: network neutrality. Google and Verizon, for example, reportedly teamed up to push lawmakers into allowing content creators to pay for prioritized content delivery. They later denied this was the case, but many are still skeptical, especially given the number of loopholes in their proposed plan.
The good news is that with respect to network neutrality, CCN is neutral. Carriers could avoid the traffic congestion that occurs around popular pieces of content, by allowing networks to move that content just once across the line, then serve it up to end users far downstream from the original source.
CCN could one day roll out alongside or on top of IP – it wouldn’t have to be a case of either/or. For use on the Internet, all that’s needed is its integration into the routers and networking equipment that move data packets across the net. That’s no small thing, however. CCN is likely years and years away from real-world use. But clients testing PARC’s software solution, the open source technology hosted at CCNX.org, are just 18 months to 2 years from using it commercially.
We YouTube video viewers will have to wait a bit longer, unless Google were to get involved, for example. There are very few companies of Google’s size that could actually implement a technology like this on a large scale…which is why it’s interesting to see that PARC’s CCN software is now available for testing on Android devices. That wasn’t done at Google’s behest, mind you, it’s just that Android is based on Linux, and Linux is the OS of the Internet.
Dealing with traffic congestion isn’t CCN’s only advantage. It can also take advantage of any radio on a device to transfer bits, it has a robust security model in place involving cryptographic technology, it’s flexible, and, from an end user standpoint, it would be simple to use. Instead of having to practically become a network administrator to set up your home’s network, for instance, configuration would be much easier to manage and security would be built-in.
Researchers interested in exploring the possibilities of CCN on mobile can download the new code for use on Android devices, and then, as PARC says, “work on the bleeding edge of technology experimentation and adoption.”
People say that I never say anything bad about the iPhone. Here’s something: its notification system is awful. In particular, I’m talking about the way notifications are managed. They aren’t. Any Push Notification that comes in overrides another one. Apps can get badged with updates, but you’ll have no idea what’s new. That’s why Boxcar is a vital app.
We’ve written about Boxcar a number of times over the past year or so. They’re now launching version 4.0 of the app, and it’s by far the best yet. The entire app has been overhauled from the ground up. The result is a killer management system for Push Notifications. Apple needs to buy this company.
While previous versions of the app gave you a time-ordered stream of your notifications, it was somewhat daunting to look at. And you couldn’t do much from in the app itself. The new design makes everything much easier to manage and more visually appealing. For example, notifications from Twitter now show up complete with the user’s profile picture. And you can retweet or reply to any of those message right from within Boxcar.
The app itself has taken a design cue from the excellent Facebook iPhone app and features a home screen with nice looking big buttons for each service you have activated. There is also now Retina Display support if you have the iPhone 4. All of this works on the iPod touch and iPad as well.
And 4.0 brings Support for a host of new services. Google Buzz, Google Voice, Github, App Recommendations, and a few others are all now included and can be installed with a few easy steps. These App Recommendations are done by Boxcar themselves.
Boxcar is now a team of 5 people based in New York City. To date, they’ve delivered over 500 million notifications to iPhones, iPod touches, and iPads.
Boxcar is a free download in the App Store (though it’s $4.99 if you wish to disable in-app advertising). You can find it here. Version 4.0 should be live shortly.
Urban Airship, a company that provides push messaging and in-app purchase content and notifications for thousands of mobile apps, just announced that it has closed its Series B funding round for the amount of $5.4 million.
This brings Urban Airship’s total venture capital investment to $6.5 million. The round was led by Boulder’s Foundry Group. True Ventures and the Founders Co-op — both of whom participated in Urban Airship’s Series A in February — also joined the Series B.
You might not be familiar with Urban Airship by name, but if you have ever used an iPhone, BlackBerry or Android app that offers push notifications or in-app purchase options and notifications, you’re familiar with what Urban Airship does.
Push notifications are a pretty common part of modern mobile applications. It’s a great way for an app to communicate with the user without necessarily running in the background, it can be used in lieu of SMS messaging to deliver news and content, and it can also alert users of newly posted content, new deals and promotions — even geographically relevant information.
Urban Airship created a platform with a goal to make adding push notifications to BlackBerry, iPhone and Android apps easy and seamless. Its AirMail Push product is already in use by apps like Dictionary.com for iPhone [iTunes link], Gowalla and even Mashable’s iOS apps [iTunes link].
Urban Airship also offers an AirMail Inbox service that can stand in for MMS or standard e-mail blasts. It’s a way to transform push notifications from opening an app to instead opening a message that looks more like an e-mail or graphic rich page.
Urban Airship also offers tools for in-app purchases, making it easy for users to be notified of their in-app options and confirm their purchases. For content publishers that offer subscription content — think magazines and newspapers — these notifications can also be utilized to let users know when an issue is available for download or when another component of their subscription is ready.
Newsweek recently started using the In-App Purchase subscription module in its iPad app.
The big advantage of using something like Urban Airship, rather than implementing your own push notification services, is time. Apple, as an example, has very specific rules about how push notifications can be used. Additionally, maintaining your own server that can deliver persistent, real-time messages and scale based on traffic and usage can be costly and time consuming.
Because Urban Airship already has the backend in place for the three biggest smartphone platforms — in terms of app usage — it’s a compelling option for developers. Additionally, the first 1 million push messages sent each month are free.
The platform agnostic nature — at least from a developer standpoint — means you can use one service to power and manage your push notifications, rather than having to maintain separate servers or services for each device type.
The real potential for push notifications, we think, goes beyond letting users know what’s new in an app or even if something is available for purchase — it’s all about location.
As we’ve noted with apps like Foursquare, Loopt and SCVNGR, location-aware apps have real potential. Location allows content publishers to give context to their messages. For instance, if I’m downtown and I get an alert that X store is having a sale, and that message is delivered based on my location — that message has more meaning and is more useful.
That’s one reason that Living Social is using Urban Airship to deliver its deals data.
Having a server-side platform that can deliver customized messaging and offer in-app purchasing options is a valuable service. By positioning itself as something that can be used on multiple platforms and offering affordable platform tiers, we expect Urban Airship to continue to succeed.
More About: android, blackberry, in-app purchases, iOS, push notificaitons, urban airship
For more Mobile coverage:
