Reddit Goes Open Source
Posted by James | Filed under Social, Web
Reddit has decided to open source their entire code base, allowing developers to see exactly how the social news site works. With the move, Reddit is encouraging developers to submit their own code and extensions for improving the site. Of course, they are also potentially opening up the site to serious gaming, since developers will be able to dissect precisely how Reddit’s algorithms determine what is popular and makes it to the homepage.
Steve Huffman, co-founder of Reddit, thinks this transparency is something users deserve, and is a strong competitive move against leading social news site Digg. “Digg has struggled to stay transparent with their users,” he said. “Social news in general has hid behind algorithms, which has caused some consternation amongst users. Users don’t get why things aren’t showing up on the front page.” Of course, he is referring to the fact that Digg’s algorithms are completely secretive, often creating frustration amongst users when stories with a lot of Diggs and comments are left off the homepage, but seemingly less important stories find their way to the front.
Much like Facebook’s fbOpen, the idea of going open source isn’t so much as to encourage developers to build their own Reddit clones. In fact, Reddit is using the same licensing structure as Facebook – CPAL (Common Public Attribution License) – which stipulates that anyone who uses the Reddit code needs to make their changes available to everyone, as well as acknowledge they are using it.
Reddit’s new site for developers – http://code.reddit.com – is now available. At this point, the race is on to see who can figure out the algorithm first, and it’ll be interesting to watch and see how heavily people try to game the site once it becomes public knowledge. It should be noted that when the company launched its new re-design a couple of weeks ago, within minutes the most popular story on the site was one about how badly it sucked, encouraging Reddit to bring back the old version!
Tags: algorithms, Common Public Attribution License, CPAL, digg, facebook, Facebook Open Platform, Open Source, reddit, social news
Firefox 3 Release Candidate 2 Released
Posted by James | Filed under Web
Firefox 3 Release Candidate 2 is now available for download, but read the release notes first so you can familiarize yourself with the known issues.
Some advice from that issues page:
- Don’t install Firefox 3 and Firefox 2 in the same directory.
- The Crash Reporter may not activate if Flash causes a crash.
- Delete an entry from history, then clear private data, and they know Firefox crashes already.
- You may need to reset some customized toolbars.
- Some Web pages don’t properly detect Silverlight.
- There’s not yet a Windows Media Player for Windows Vista users.
- DNS lookups may be slow on the Mac for IPv6 users.
- Linux users may have to reconfigure their mice.
Remember that this is beta code. That’s cool because use of beta code is one of your privileges — some might say obligations — as an open source user. The more beta-testers code has the more bugs are found and the better the final release.
Beta testing is not only troublesome fun but it gives you insights into the development process, which can lead you deeper into the open source world.
And don’t forget to read the bug filing instructions before reporting to Bugzilla.
Now get out there and find what’s wrong!
The final ship date for the software is still scheduled for later this month.
Tags: beta, browser, firefox, mozilla, Open Source, software
SearchMonkey Custom Results Pushed to Users
Posted by James | Filed under Search, Semantic, Web
Yahoo!’s SearchMonkey platform got a little more public today with the unveiling of the Search Gallery – the platform’s official application repository. The gallery already has been open to developers and curious bloggers for a couple of weeks, but Yahoo! is now pushing it to the public at large via a “Customize” drop down menu on all search results. In addition, starting today developers can share applications via external links even if they haven’t yet been approved for inclusion in the official gallery.
"This is the first phase of a larger plan to provide opportunities for viral distribution of SearchMonkey apps," said Yahoo! Search Product Manager Amit Kumar on the Y! Search Blog. "We’re continuing to develop new ways to surface and share useful and high-performance applications in users’ search experience and more broadly on the web, so expect more in the near future."
As of launch, the gallery contains 39 approved applications. These range from apps enhancing Yelp! and LinkedIn results to one that provides a code reference for Ruby related searches.
SearchMonkey has the potential to be very disruptive in the search space. It gives web developers the ability to enhance the display of search results without the ability to influence search rankings. Who better to know how to best display content than those who created it? Unfortunately, the search applications I tried out today mostly didn’t seem ready for public use.
As you can see in the screenshot above, the Yelp! application that I installed didn’t really enhance search results beyond slapping a logo next to the URL. In theory, my search for the Petite Deli in San Francisco should have yielded a Yelp! result with the deli’s address, phone number, rating and links directly to user reviews and photos. The IMDB application (which admittedly appears not to have been created by IMDB) gave me connection errors, and the Last.fm app behaved similar to the Yelp! application for me — it enhanced nothing.
Did Yahoo! jump the gun on pushing the Search Gallery public? Or perhaps have their gallery approval standards been too low and these apps just haven’t been tested thoroughly enough?
Tags: customisation, Search, searchmonkey, yahoo
Welcome to Jimmahdigital.com, personal blog of James Simm, Wigan/Manchester-based .Net developer.
