The U.S. government’s DARPA (Defense Advanced Research Projects Agency) has granted $11 million to Rice University for their continued work on a project that is essentially an auto-complete tool for programmers. Described as a massive database of open-source code, PLINY aims to allow programmers to finish their software much more quickly through a simple search.
The name of the project comes from Pliny the Elder, a man who wrote one of the world’s first encyclopedias during the Roman Empire. This name is not used lightly, as PLINY’s developers are attempting to build one of the world’s deepest repositories of code, giving the programmers the ability to write the first few lines of their software, and have the rest automatically appear at the touch of a button.
This is an incredible undertaking, and one that almost seems impossible to imagine being successful. Swarat Chaudhuri, an assistant professor at Rice University and co-developer of PLINY, says that the code that is automatically pulled up “should work seamlessly with the code that’s already been written.” And remember, this is open source code we’re talking about, so who knows how well it’s been updated/maintained, or if it will be compatible several years in the future.
A sample demonstration from Chaudhuri shows him filling a hole in a piece of paper as an example of completing the missing work from a programmer’s project. PLINY is shown searching through billions of options, and using the closest match to fit with a bit of editing. The project is still a long way off from completion, and DARPA’s funding will allow a team of over 12 researchers at Rice University to continue working on PLINY for the next four years. Sounds like it will at least work better than the auto-correct we have on our smartphones today.