Codetrails's blog

Sponsored OSS Development: Making Snipmatch 2.3 fit for any language

The following is an experience report by Klaus Landsdorf and Stefan Macke, two developers working at Alte Oldenburger Krankenversicherung AG, a German health insurance provider, about how sponsored Open-Source Development made Code Recommenders Snipmatch work for their use case: managing hundreds of code snippets for several languages (other than Java).

Among other languages, Alte Oldenburger Krankenversicherung AG uses Software AG’s Natural, a 4GL programming language optimized for accessing Adabas databases. As Natural requires quite a bit of boilerplate code, over time the development team accrued a collection of nearly 1,000 code snippets that relieved us from manually writing all this boilerplate code. But a collection of snippets this large requires the right tool to manage it.

At first, Eclipse’s built-in code templates feature seemed a good solution to this problem. Unfortunately, managing a collection of snippets this large as code templates turned out to be quite cumbersome.

Codetrails Connect and Raspberry Pi

There’s been lots of news recently in the Codetrails Connect world – not the least that version 1.2 was released, bootstrapped with a rich repository of sample code.

Since June of this year, we’ve run a pilot program where individuals have contributed their code completion events to the crowd. With 15.000 code completion events in just over two months, they have build quite an extensive repository of knowledge; all implicitly collected without any effort on their part.

To make the pilot more fun, we offered 10 Raspberry Pis for participants who contributed more than 50 recommendations. Let's have a look at what our winners are using their Pis for.

Crowd-sourced Code Recommendations - 27 more days to win a Raspberry Pi

With over 10,000 completion events collected, the Codetrails Connect (formerly Woodstock) community have contributed a significant amount of data. Using Codetrails Connect, they’re sharing their development knowledge with each other on an ongoing basis, enhancing their own and their colleagues' IDE’s through intelligent code completion recommendations.  See Johannes’ post for some things we’ve learned.

Mining intelligent code recommendations just got easier: [ctrl]flow Miner 0.8 released

Codetrails’s [ctrl]flow Miner offers a flexible command line interface that allows you to adjust your data-mining job to your needs. To make these configurations even easier, we've just released Version 0.8 of the [ctrl]flow Miner with a handy configuration wizard.

Introducing Crowd-Sourced Code Recommendations at Eclipse DemoCamps 2013

The Eclipse Kepler release is getting closer. Let’s celebrate this event together at one of the many Eclipse DemoCamps this summer. And what better way to celebrate than demo-ing our latest tools that take code completion into a new direction: down the rabbit hole – and into the cloud.

Contributing to WALA to improve static analysis in the [ctrl]flow Server

WALA, the T. J. Watson Libraries for Analysis, is a powerful open source framework that drives the [ctrl]flow Server’s static analysis process for extracting usage information from your code (essentially the analyze command). Over the past few weeks, we have made several contributions to the framework that we hope will be useful not only for Codetrails but for other WALA users as well.