We reported on Google's LiquidFun V1 earlier in the year as, as much fun as you can have without actual liquids getting involved. Now we have LiquidFun V1.1 and you can have as much fun in a browser.
In case you aren't sure what a particle simulation is all you need to know is that you define the position and properties of a set of point particles and the program will animate them for you. If you don't apply constraints to keep the particles positioned relative to each other you get a fluid:
If you apply some constraints then the particles "flow" as solid elastic blobs and you have soft 2D objects:
If the two videos have piqued your interest then visit LiquidFun and try the realtime demos out in your browser. Also try the games to see the sorts of things that you might use LiquidFun for.
Also new is official support for iOS for the testbed and the EyeCandy app. The particle simulation has alsow been optimised and it can now make use of ARM NEON SIMD hardware if the device has it - A8 up. There are also lots of bug and stability fixes and new functions.
There is also the provision of two open source games to show you have it is done - currently only available from the US Play store. VoltAir is a C++ platform game.
The second game is LiquidFun Paint which is written in Java to show you how to use the SWIG bindings. Given that SWIG is a development tool to allow a range of languages to work with C/C++ the chances are that the same techniques could be used for PHP, Perl or Python to name just the languages starting with P.
Since it was released LiquidFun has been use by a number of commercial games and built into examples of particle simulation.
As I said about versions 1 if seeing LiquidFun in action doesn't inspire you to write something I don't know what is going to get you moving. With advances like this there is a lot of mileage in 2D games even post-flappy bird.
An advanced level course in machine learning has been added to Udacity's catalog. Over the course of four lessons you'll get to understand and use Google's open source TensorFlow tool, which can [ ... ]