Proprietary software holdouts

The OERu is committed to using free and open source software (FOSS) for all our technical infrastructure to the extent possible. We're doing pretty well in living up to that principle, but there are a few crucial technology areas where we have not yet found a sufficiently usable or mature FOSS option. This post identifies those exceptions, as they stand in September 2016. Despite being grateful to have these proprietary applications available to us, we will replace them with FOSS in each case as soon as a viable option is available!

As a distributed organisation, with many international partners, the OERu's success depends on reliable online video conferencing technology for communication and to support collaboration.

For small scale conferences, with to up 8 participants (limited by the participant with the lowest bandwidth Internet connection), we use FOSS-based WebRTC-based directory services, like Appear.in (it's not a FOSS web app itself, unfortunately, but it uses the fully FOSS WebRTC stack built into open source browsers like Firefox and Chromium - the open source project on which Google Chrome is built and with which it shares probably 99% of its code), or those built into open source tools like our Rocket.Chat messaging environment.

Where using FOSS tools currently falls down is larger-scale video conferencing, with more than 8 participants. That is due to a feature/bug of WebRTC: end-to-end encryption of video and audio data between all the participants. The upside is that there's no central place (as exists with older technologies like Skype) where the communication data is unencrypted and can be "slurped" by "intelligence" agencies unbeknownst to the participants in the call (think the NSA in the US). The downside, however, is that the use of bandwidth increases exponentially with each new participant in a group virtual discussion.

The problem we have encountered is that existing FOSS "multicasting" technologies (where the various encrypted data streams are decrypted at a central point, and bundled for each participant, and sent onto them as a single, re-encrypted stream of aggregated video and audio data) like BigBlueButton still depend on outdated (and proprietary) desktop software technologies like Adobe Flash. We're watching BBB closely, in hopes that its community can successfully make the transition to open web standards like WebRTC (which is part of the HTML5 specification).

In the meantime, we've made the decision to pay for the use of a proprietary technology called Zoom, which is noteworthy in the video conferencing space for the fact its clients support Linux desktop environments (which is what we use at the OER Foundation, the OERu founding partner for whom I work).

Although their Linux client is clearly a poor cousin to their other clients, their developers have been admirably willing to improve its functionality (and fix glaring bugs) based on our constructive feedback (including a few video conference sessions to explain bugs we were seeing). It works well for us, and we have had good feedback regarding Zoom from our various collaborators.

Update 2017-06-14 since writing the above about Zoom, we have found an open source alternative that has almost entirely replaced Zoom for us: Jitsi Meet. It's free in both senses of the word: gratis and libre, and works very well besides. It, like most WebRTC-based video conferencing technologies, doesn't even require the installation of a separate client (except on proprietary platforms like iOS) because it's an open standard built into market leading web browsers as described above.

Also, we're going to be trying a very interesting new open source technology developed at MIT as a way of managing larger scale "webinar"-type video conferencing: Unhangouts (which, despite the name, no longer involves the similarly named proprietary Google product).

We encourage you to give them both a go!