2015-05-15 – Pluto-Palooza

Alex: In preparation for the first public pre-release, I was busy with preparing the website, including Google Analytics to get some impression about OpenSpace’s popularity. Furthermore, on the code side, I implemented the parsing of the REX instrument on-board New Horizons and implemented a basic rendering of it, which was further improved on by Anton. Last but not least, preparing distribution packages on Mac is *way* too complicated for my taste.

Anton: Before our big presentation I was thrown between run-throughs, meetings and last minute fixes to make the demo look as good as possible. I was also on the Pluto Palooza event at The Explorers Club on the upper east side to see what we would expect the scientists talking about before/during the presentation. It all turned out great and we got tons of compliments of what we have achieved so far!

Tomas and Emil: We are up and running at Goddard! We have started looking into how we can optimize the multi-res bricking volume rendering, which is going to be a big part of our thesis. Previously the volume rendering was performed in two render passes per frame. A dry-run was carried out by the GPU in order to select the bricks required for the second step – rendering. Our first experiment was to move the brick selection to the CPU to minimize the amount of data being sent back and forth between the GPU and the CPU. This seems to have resulted in a slight speedup, especially when the volume takes up a large portion of the screen. We are further looking into how we can make use of the data from the first step instead of doing the brick lookup twice. This will keep us busy for a while!

2015-05-08 – Ramping up to a big prerelease

Alex: With the upcoming teacher’s release in full effect, my work was focussed on finalizing the Timeline GUI application and the Mac support. Some minor quarrels with Mac are remaining, but they will be fixed during the weekend. Last todo items on the list: Testing everything extensively and preparing distribution packages.

Joakim: This week I’ve had to spend time on other projects so development has sadly been mostly still for my part. Hope to get more done in the following week.

Anton: I am in the middle of preparing for the big presentation of OpenSpace at PlutoPalooza in the LeFrak theatre at the American Museum of Natural History next Thursday, May 14th. The work has so far been mostly testing the setup on different machines and deciding what to show and how to show it. I have however also been working on how the path is drawn to be able to show hour and 15 minute marks along the trajectory of New Horizons.

2015-05-01 – Two Minutes (well, weeks) to Midnight

Alex: This week of mine was very occupied with increasing the Mac OS support for OpenSpace. In this week went went from “not compiling” to “showing some basic rendering”. Not a bad start after OS X was abandoned a few months ago. This time, it will be different!

Joakim: This week I’ve spent most of my time debugging new features and merging two major branches of OpenSpace. It’s getting there but progress is slow…

Anton: I have finally had some time to work on visualizing the Dawn mission. I have now successfully added the field of view for the framing cameras and started with image projection with a subset of images from Vesta in August 2011. Even though the comet is currently approximated by an ellipsoid, the results are good so far.

2015-04-24 – Ramping up to the next Release

Alex: I have been very busy with getting back into Qt 5 for programming the preliminary user interface that will be released in the upcoming weeks. The work on the communication between the main OpenSpace instance and connected GUI instances. It has progressed pretty far and some finishing touches are required.

Joakim: I’m finally back to work on OpenSpace after a long and much needed vacation. This week I have spent most of the time getting reacquainted with the code, I’ve also started looking into refactoring and improving our planetary-projected textures

Anton: This week has been full of merges, version control conflict solving, bugfixes and cleanups. No major new features were implemented from my part but I do now have a stable version containing work from several branches.

2015-04-17 – Preparing for Distribution

Alex: Most of my work week was dedicated to finishing up the SceneGraph restructuring work and integrating all changes that have happened in the meantime so that nothing is broken by my commits. After that, I started the work on a preliminary GUI application that we want to distribute alongside OpenSpace to enable control for the New Horizons encounters.

Anton: I have implemented a plane that the images taken by LORRI gets projected to at the time of capture at the position in space where the camera was aimed. This allows us to show stuff like the Europa rise from Jupiter as New Horizons saw it and outgassing of the planets and moons examined during the mission.

2015-04-10 – Back to Normality

Alex: Most of my work this week as been dedicated to rewriting the module loading system. It now understands dependencies and can initialize (and evaluate) scenegraphnodes in the correct order. And I got to learn a lot about topological searches in directed graphs!

Anton: A better part of the week went to finishing up the work on estimating positions and rotations when data is missing and how to show to the user that the position might be inaccurate. I did have another successful teleconference. This time with a couple of guys from Madrid involved in the Rosetta mission at ESA. I am currently researching the new information we got from them and texture projection for upcoming tasks.

2015-04-03 – Happy Easter Workation!

Alex: For this week, I started investigating methods to deal with multiple installed graphics cards on a system. It seems like there is not much you can do unless you are on an nVidia Quattro card. Furthermore, I started the work of supporting Mac OS once again. Due to the current limitations of OpenGL 4.2, this will require some fall-back solution for the order-independent transparency.

Anton: The most interesting thing happening this week was a teleconference with some JPL/NASA people who have been working closely with the DAWN mission. They seem excited about what we are trying to do and want to help us as much as possible without spreading ITAR protected data. To be able to handle missing data intervals I have been digging deeper into SPICE kernel structures and the current positioning of bodies in OpenSpace. I have tried a couple of different approaches and have just implemented a simple interpolation method for when there is a gap. I have also been working on a way to define whether the existence of an object should be time dependent.

2015-03-27 – Timeline Scrubbing and DAWN

Alex: Nothing new this week again from my side. Just the growing wish to get back on the implementation train again and continue to contribute to the codebase in anticipation of the big July event.

Michal: Have been working on extending the ImageSequencer class with the capability of returning a set of images, so that we can skip in time. The new logic is to allow each class with projective capabilites to keep track of its own subset and do a check against a min-max range for each sub-subset, which should minimise lookup and reduce lag significantly. Also, this will facilitate the implementation of MVIC and allow support for an arbitrary amount of missions. Ultimately, I’m focusing on stuff that is most essential for the next programmer in line as I’m wrapping up to move over to work on the Explorer project.

Anton: Due to my lack of data on the Rosetta mission I decided to put that on hold and start looking at the Dawn mission. At first sight, the SPICE kernels seemed to cover the Vesta orbits but some random errors occured. After some research I found that the probe was actually rebooted and unavailable for a few short periods of time during the orbits. The temporary Dawn dummy model was devided into 3 parts, making the solar panels rotate separately from the body to be able to visualize aiming of the instruments and have the solar arrays face the sun to provide energy for the ion thrusters. I am currently considering options for handling periods of time where data is missing.

2015-03-20 – Paperplanets

Alex: Another successful work of working on the SciVis paper whose deadline comes closer and closer (March 31st). So noteworthy has happened this week and I hope to be back on the development track in April

Michal: Been working on the ImageSequencer class to make sure all images get projected. Problem is that the playbook (unlike FITS header files) contain no info about which target is actually aimed at, which at this point is necessary (might make own playbook). Tried setting up external application for slider. Other than that, been going over everything I know with Anton, outlined practical solution for how to project to arbitrary object and we discussed various ways of extracting the projection capability from RenderablePlanetProjection class.

Anton: I have looked into the creation of the planets, and introduced the triaxial ellipsoid shape to make the geometry of the bodies more accurate. I have also fixed the texture mapping of the planets to make the longitude axis align with reality and started looking at image projection to arbitrary objects which will probably be the main focus next week.

2015-03-13 – Return to Normality

Alex: I am currently caught up in writing a scientific paper for this year’s IEEE Vis conference, so I have not been able to work on many things this week. Among those things that I could do was preparing an FTP server for managing big data storage, as well as updating software on the webserver.

Michal: This week I have been stretched across many different minor tasks – polishing out the very last issues with the uv maps, prepping OpenSpace for the meeting with NOVA producers, and have started looking at how to write the time-interpolation for OpenSpace. Alongside this I have also been discussing with Anton various strategies on how to approach the modelloading, issues needing fixing and detailing out parts of my work and briefing stuff regarding general opengl functionality. On Wednesday during a screencapture session my machine suffered a disk failure and was set back a day, now with all the small task out im going to foucs on the timeline.

Anton: This week I’ve been examining a few existing models and SPICE kernels for Rosetta with mixed success finding suitable ones for OpenSpace and preparing for possible extension of our object loader. Together with Michal I have also been making small adjustments and preparing for a demo this Friday at the museum which turned out great!

1 2 3 4 5 6 7