2015-03-06 – Post-Deadline Stress Disorder

Alex: With the IMERSA deadline and demo behind us, it was time to pick up some of the hacked pieces of code that were created to make everything work and clean most of them up. Furthermore, to have access to more hard disk space (and more processing power), I moved this server from a virtual server hosted by the university to a dedicated, stand-alone server. A lot of work that will hopefully go unnoticed for everyone.

Joakim: This week there hasn’t been much development made. The week has been spent catching with my other projects which had been put on hold before our OpenSpace demo. Have had the time to ponder the upcoming scale graph thou!

Michal: With the many issues we’ve had with the geometry I decided to rebuild the model from scratch. The original had a multitude of issues, non-manifolds, lamina faces, 4-sided faces, edges with zero length, faces with zero geometry area, non-planar faces etc etc. With the rebuild we now have a lightweight model with good edgeflow, nonintersecting geometry and all one-part geometries are now logically fused. In essence, with all the bad topology gone the model has been reduced from 73 to 17 MB. Now on to UV-layout and texturing.

Anton: My first week on board have been spent on getting things running, loading a model for the 67P comet and creating its path. I still have not found data for the full trajectory but the result is satisfactory for the time period I do have.

2015-02-27 – IMERSA

Alex: Everyone has been very focussed on getting the demo at the IMERSA conference up and running. Among those changes on my ends have been mostly usability; adding new properties to change, adding an on-screen GUI for the live demo, and cleaning up and hardening the code base. Joakim and I implemented a temporary hack for switching coordinate systems until a proper mechanism is in place so that we do not run into floating-point precision errors.

Joakim: This week has mainly been spent performing hotfixes and testing the software for the demo this weekend. Amazingly satisfying to see it all come together and feel that we have a very nice piece of software to show off!

Michal: Intitially I began with changing the modelloader upon discovering that there are issues with the uv-coordinate mapping. The cause of the problem lies in the geometry itself as the model has been exported through 3 different softwares. Although now we have a modelloader thats capable of stringing together a set of geometries. Other things worked on: Image sequencing has been improved along with instrument switching, figuring out issues with the jupiter run along with various improvements to the fov class and fixing of the obj geometry.

2015-02-20 – First!

Alex: Most of my week was dedicated to either fixing the webpage, updating the build server, or cleaning up various code. In the rest I managed to include a rendering of the Milkyway as a backdrop as well as finally getting the Stars to render in a quality level that is acceptable. Additionally, we managed to test our setup not only on our dome in Norrköping, but also in New York (with success on our end and not-so-much success on the other side of the ocean unfortunately).

Joakim: Why can’t they just get along? Who would’ve imagined that network synchronization of multiple computers running at different frame rates could be such a chore. I can happily say that after almost a weeks hard work it finally seems as if all of the cluster is in perfect agreement and playing nice.

Michal: Examined the kernel timeline and determined where New Horizons captures data from Pluto and Charon. Added refinements to the image-sequencing class, FOV line draw and cleaned up a lot of the mod files. Also, fixed some issues with the playbook read and added functionality to read Hong Kangs pre-parsed playbook.

First entry

This is the first entry in the development blog which will highlight the code developments for each week. This way, you are able to get an insight into the current development of OpenSpace. It is very appropriate that the polish of the WordPress webpage is the first entry here with many more to come.

1 2 3 4 5 6 7