In this edition of our open source news roundup, we take a look at Yahoo releasing software for deep learning, the Linux Foundation picking up RethinkDB, and more.

Open source news roundup for February 5-18, 2017

Yahoo open sources TensorFlowSpark for deep learning with big data

In the latest of a long parade of releases from tech giants, Yahoo! announced that it’s open sourcing TensorFlowOnSpark, software that makes Google’s “TensorFlow open source framework for deep learning compatible with [Yahoo!’s] data sets that sit inside Spark clusters.”

Using TensorFlowOnSpark will enable developers to create deep learning models that run on large clusters of computers. According to Yahoo’s machine learning team, “changing fewer than 10 lines of Python code are needed” to get an existing Tensorflow program to work on TensorFlowOn Spark.

The source code for TensorFlowOnSpark is license and is available on GitHub under an Apache 2.0 license.

RethinkDB comes under the Linux Foundation’s wing

It was only a few months ago that open source database developer RethinkDB announced that it was shutting down. But RethinkDB isn’t dead. The Cloud Native Computing Foundation (CNCF) has purchased the company’s intellectual property and given RethinkDB’s code to the Linux Foundation.

The source code is available under an Apache 2.0 license which, according to Bryan Cantrill of the CNCF, will allow RethinkDB to do better than it did under its previous AGPL license.

Udacity open sources its self-driving car simulator

Want to create your own virtual environment to test a self-driving car? You can now. Web-based education site Udacity has open sourced its self-driving car simulator.

The software is part of Udacity’s self-driving car nanodegree program, and you can grab the source code from GitHub. Writing for TechCruch, Dale Etherington notes that “Open sourcing its self-driving car simulator … provides even more base-level tools to the community working on this big, complex problem.”

IBM shares projects to make software more accessible

More and more software developers are trying to make their creations as accessible to people with physical and visual impairments as possible. IBM has helped make doing that easier by releasing two accessibility projects as open source.

The projects, AccProbe and Va11yS, can “help developers test and debug accessible applications” and provide sample source code, respectively. Using these projects, developers can not only craft accessible software, but can also “be used to compensate for environmental and physical conditions that can affect the user.” You can fork and use the code for AccProbe and Va11yS on GitHub.

In other news

Thanks, as always, to Opensource.com staff members and moderators for their help this week. Make sure to check out our event calendar to see what’s happening next week in open source.