The Saga Continues
A few people headed home after day 1, but we still had over 20 people sprinting on day 2. On Day 2, we continued to code and also had several design discussions on technology and APIs.
Transform Services: One of those discussion was between Connexions, Vietnam OER, and OERPUB (my Shuttleworth funded team) about a common API for transforming documents from one type to another (Word to HTML to EPUB to PDF, etc). Although it sounds really esoteric, it turns out that this is an area that is ripe for collaboration and sharing code, not only among these teams, but also for anyone producing digital and print books like Booktype, Pressbooks, and even OReilly. The photo to the left shows a quick sketch of common ground and this doc has a further writeup on the ideas shared. After the discussion the VOER team headed home, and one of the Connexions developers started refactoring current code to match the new shared vision.
WordPress editor embed: And the photo to the right shows the progress on getting the OER editor embedded in WordPress. It builds on an existing WordPress plugin for Aloha (the base HTML5 editor of the OER editor.) The existing plugin needed to be updated to make everything work (as all code does periodically) and that has been submitted back to the original developer. We also added in the new semantic elements drop down (not shown) and are updating to the latest Aloha to correct for a bug that Clemens recognized in the behavior.
Internationalization: Clemens Prerovsky from Aloha showed a group of us how Aloha internationalization is done. Basically they use “keys” in the code and then each plugin has a set of language specific files with key-string pairs. Translators edit those directly right now. Although not super streamlined if you have lots of new translations to do, it is simple and straightforward. We also need translations for auto-generated text that is a part of the content being created, and we discussed the added complexities that entails.
Abandoning Content-Editable: Since we had an Aloha founder in the room, Phil of CNX and Clemens of Aloha discussed the problems with trying to use content-editable in the editor. Browsers handle events in the content-editable region differently which means that the Aloha base functions have a lot of code to correct for that. It may be saner to essentially implement cursor and event handling from scratch and not use content-editable. Clemens will take that idea back to the Aloha team and perhaps we will work on it during the Aloha barcamp, June 6,7 in Vienna.
Wikipedia Therapy: The folks working on wikipedia entries got to experience editing over time (see the OpenStax College entry), which doesn’t happen in a 2 hour workshop. The wiki folks also provided feedback on how to navigate the wiki world and respond to reverts and slightly heated discussions. They worked on cross-wikipedia licensing issues, for instance when a resource from Brazil is reused in English wikipedia and the license is in Portugese.
Logging and Analytics: Connexions did more work on logging. Ed, normally consumed by planning and management enjoyed being able to develop for a couple of days. Ross investigated pwik — an opensource analytics and log analysis framework. He created a million line log file to do some load testing and initial results indicated that performance may be a limiting factor.
I hope this gives a flavor for the cross-pollination that happens at sprints; conventions agreed upon, ideas germinated, bugs identified by experts across the room, nagging questions answered, opportunities to branch out, or branch back and brush up skills and rejuvenate.