Scott mostly worked on implementing the imageEventRouter and its corresponding GUI. By the end of the month the imageEventRouter was 90% implemented but he found that the renderer was not drawing to the visualizer. Scott is awaiting a meeting in May when the team is back, to discuss renderer changes that might be necessary. He went on to implement the full image GUI, and parts of the TwoDData event router and gui components as well. Additionally he updated the windows build to accommodate the new changes, and worked through a variable selection bug with Alan, since the image renderer had no default 2D or 3D variable.

John refactored the XmlNode class and a number of subordinate classes to simplify the code and improve structure.

John also ported the ShaderMgr class from vapor 2.x to vapor 3.x.

John committed a first version of the TwoDImage renderer. John is now waiting for Scott to complete the GUI and expect to then have to debug the class.

Scott adjusted the scheduling of the XSEDE webinar to June in order to avoid an NSF audit of the Mesa Lab. He further tested Vapor on Stampede’s filesystem and Maverick cluster so users could follow along with his lesson plan.

Scott began to do the final renderings and production of the coral visualization for submission to XSEDE 2016. He generated a 10­part storyline of what needed to be rendered after meeting with Frederic Castruccio, the developer for the CT­ROMS model. He also worked with Matt Rehme to produce an introductory sequence that showed the Global Conveyer Belt, which was outside of the scope of the ROMS data he was rendering. He has been using a trial version of Adobe Premiere Pro instead of ffmpeg for video editing