retweetradar.com - One days work from concept to launch with Google App Engine... Scobleized!Dec. 22, 2008
So I was getting a little restless and decided to work on another Google App Engine app, this time without all the Python learning curve and with some real world experience with the Google tools under the belt from http://spy.appspot.com. The goal was to see how fast I could go from concept to useful application leveraging a few of the revolutionary tools we all have at our fingertips today. In the end with about one days work an app was released, likely with a few bugs, but some fun things happened and I believe a bit of foreshadowing of the future of building web apps.
The ConceptI had an idea that I really wanted to extract meaning from a large-ish set of social media messages in near real time and visually trend it over time, showing topics heating up in conversations even while still slightly off the conventional news radar. We've all seen lately the groundswell social media conversation routinely outpacing tradition broadcast media in speed and openess of accounting, this app would be yet another way to put your finger on that pulse.
The class of information that looked interesting was "retweets" in other words messages where Twitter users were quoting someone else's post, essentially saying they liked it, or spreading the word. Tech blogger Robert Scoble called out the same last Friday saying "[Retweets] ...by the way, great place to find news!"
So now to the how, obviously Google App Engine was my choice for delivering this application, it's ease of access, familiar development environment, ease to scale and obvious lack of initial cost make it a winner -- Paul McDonald and Tom Stocky being very great folks to talk to, among other Googlers on the App Engine team I am sure, helps as well.
Next I needed the information, Twitter's great search API acquired with the Summize purchase makes pulling tweets in near real time a non event. Now the value add, the real magic in the idea was to pull relevant information from the tweets and trend them, for that the lesser known Yahoo Term Extraction API was chosen, it's simple interface, reliability and quality results. So the parts were chosen... we have the ability to capture social media posts, process out key terms, save, count and display them on the web... very nice. So in the equivalent of one days work over the last two days I took these parts and pieces, shook them up with jQuery and released http://retweetradar.com to the world.
The OutcomeThe outcome was astounding http://www.retweetradar.com launched publicly via a Tweet and FriendFeed posts at around 8pm CST on Saturday, given the work done on Friday and Saturday certainly no greater than one common work day from start to launch. Once Sunday came around I tweeted the news to a few folks, Dan Zarrella was one he has done work on the data behind viral messaging on Twitter another of note was Robert Scoble to close the loop on his tweet about the value of retweets... then things took and unexpected turn... an inconspicuous but cool "like" came in from Scoble on Friendfeed.
Next thing I know Robert Scoble is using me as an example in a great blog post about events the mainstream media and even Internet media miss...
I love developers who try new things out. Check this out. Is this on TechMeme? No. Plus using Google’s App Engine, which is another trend we’re tracking: cloud computing.
I couldn't agree more with the topic of that blog, even if I wasn't involved! You did read it right?
This isn't about launching another web app...The fact this is possible is the real star here, the ability for anyone to bring an idea to the web in a highly reliable and scalable way for little to no initial investment shows a bright light toward the future... folks used to say the same for picking up some shared hosting for $5.95 and knocking out some PHP but realistically that app tipped over the second it saw real traffic and was hard to say whether it would be up day by day based on the other people, number of whom soaring each month, who shared your hardware.
Google App Engine and Open APIs clearly show where the web is going, the playing field is leveled... get out there and deploy those ideas you say you have scrawled on those napkins that you say are going to revolutionize the world, the tools you need to prove it are out there right now.
So, draw your own conclusions to the viability of my new app retweetradar -- No, please do! Contact me with feedback on whether it is useful and any ideas to make it better -- it's really not a huge concern long term... but I don't believe it is possible to argue that these tools be it social media, cloud computing or whatever you want to call them are truly leveling the playing field, flattening the world and company hierarchies and making whatever the future of the web will be possible for anyone.