W3C home > Mailing lists > Public > public-rww@w3.org > August 2012

browserver

From: Melvin Carvalho <melvincarvalho@gmail.com>
Date: Fri, 10 Aug 2012 19:03:48 +0200
Message-ID: <CAKaEYhLW8di_7+_Xr4MdDPAKAueLRR+npH2bBcnm=nBVX9r5bQ@mail.gmail.com>
To: public-rww <public-rww@w3.org>
http://browserver.org/
browserver(Hey all! I'm currently getting Hacker News'ed, so if this
doesn't work for you the first time, please refresh and try again.)

Hello! I've got some good news for you: your web browser has just been
upgraded to a web *server*. It's responding to HTTP requests on the
Internet as you read this.

True story. You can try it yourself here:
[ see your own link ]

Here's what just happened:

   1. You used curl to request a resource from [ see your own link ]
   2. A browserver server <https://github.com/jed/browserver-node> received
   your request, and figured out that you wanted to talk to client *
   q6zh45q6k6kfn7b9*, which is actually your browser.
   3. The server then used an already-established websocket-like connection
   (thanks to engine.io <https://github.com/learnboost/engine.io>) to
   forward the request.
   4. The browserver client <https://github.com/jed/browserver-client> in
   this page responded, getting any necessary input from the browser (or from
   you). The server then forwarded this response back to you.

In other words, the open-source browserver server and client worked
together to give your browser a real address on the Internet.

So?

Well, this means that you don't need to roll your own custom code to
connect the various pieces of your web architecture to your end clients.

Instead, you can move the complexity of your app to the edges by making
your end clients first-class HTTP servers, and then use your existing
HTTP-related infrastructure to communicate with them.

For example, you could:

   - Subscribe them directly to any webhook-capable API, such as Amazon SNS.
   - Send notifications to other decoupled web services whenever a client
   connects/disconnects.
   - Simplify development by using the same familiar node.js HTTP API on
   both the client and server.

To learn more about how browserver works, head on over to GitHub and check
out the browserver server <https://github.com/jed/browserver-node> and
browserver
client <https://github.com/jed/browserver-client>.
browserver was brought to you by Jed Schmidt <https://github.com/jed>.
Received on Friday, 10 August 2012 17:04:18 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 10 August 2012 17:04:18 GMT