Earlier today we relased our first 100 Alpha invites! We have chosen a range of people for this alpha phase, from developers through to designers, data mining experts, monitoring enthusiasts and news curators. What we wanted to do was get feedback from as wide an audience as possible and not just those who have a geeky interest for programming and data. We are trying to make our system as usable as possible for everyone. Hopefully we’ve got some way towards that now!
Now we do understand that there is a LOT of demand for access to our system and we will be releasing additional invites as soon as we have some feedback from our initial alpha users and are sure that it is reliable enough for higher volumes. We’ll also work on the feedback that we receive and make the system as strong, simple and reliable as we possibly can!
We have been hard at work on DataSift over the last couple of weeks culminating the next iteration of our platform and making minor fixes to the DataSift site.
WebSocket Support
The client will see new data as soon as the WebSocket server pushes it out, giving them an advantage over the various polling solutions since it’s not waiting for data to queue and pulling it all at once. So the bottom line is, it’s faster!
WebSockets are a bi-directional, low latency, simple transport layer that can be used in a variety of ways. WebSockets are also a part of the HTML5 standard, with all major browsers either supporting them already, or planning to do so. You can reach a large percentage of users with a native messaging implementation. WebSockets aren’t just limited to web browsers either, they can be easily implemented in server-side projects.
Browsers that support WebSockets:
Internet Explorer 9 (beta)
Firefox 4 (beta)
Opera 11 (beta)
Safari 5
Google Chrome
iOS Firmware 4.2 (iPhone, iPod Touch & iPad)
Language Updates
We have 4 new additions to CSDL.
New Predicates
contains_word: Matches an exact word, but not supersets of the word. e.g. interaction.content contains_word "buzz" will match “buzz” but not “buzzing”.
contains_phrase: Matches an exact phrase, but not supersets of the phrase. Like contains_word, except permits punctuation and whitespace.
New Targets
twitter.mentions: All the usernames that were mentioned in the tweet. e.g. twitter.mentions contains "tweetmeme" matches all tweets that reference the @tweetmeme user.
language.tag: The IETF language tag of the language detected in interaction.content. e.g. language.tag == "en" matches only interactions written in English.
Bug Fixes
Fixed a bug which was causing stream previews to display the following message "Error loading rule with that Stream Identifier - An error has occurred that could not be recovered from". Stream preview building should now be more reliable and produce verbose error messages.
Key to the potential growth of DataSift is having 3rd partly developers build applications on top of it. In an early effort to show that we understand developers needs we have built out our own App Store so that applications that use our platform can be easily found by our own users. The store is design to allow many types of applications, these can be mobile, web based or even code libraries that use our API. We categorize applications so users will be able to find by platform, product type (e.g. CRM, Desktop Client).
The App Store is integrated back into the core DataSift platform in a number of ways. The first is that user sign-in is still via the core DataSift site – (the App Store will be hosted at http://apps.datasift.net) application listings will also be shared into DataSift itself – so when end users go to a stream they are suggested applications from the Store to use to consume those streams (and we can even suggest those based upon what kind of platform they are using when accessing the site.) The second is that you can ‘like’ a application to put it in a list of Apps you want to be able to access quickly in the future.
Follow us on Twitter
Linkedin Group
Facebook