Present: Patrick, Seth, Luka, Lee, Vitaly
Luka:
- I'm working on fixing the CVS tree
- The new tree should be ready by Friday
- I'll help Patrick roll back Jason's code this afternoon
Lee:
- Routefinding doesn't work right now
- I'll stay in touch with Patrick to get my demo ready
- Please put me on the BMG mailing list
Seth:
- We've got some cool new Nokia phones; Patrick will be using those
- The phones should connect via Bluetooth to signs
- These brackets are OK; should maybe use locks
- We should get a C client of Jason's API up and running
- To Vitaly: "its not that bad" (lies)
Patrick:
- I should have a signage demo ready for Friday
- I also made some brackets in the machine shop for mounting the signs
- The BMG pipe should infer building adjacency
Vitaly:
- Most of my time has been spent working on the basemap stuff
- need to talk to Peter about wget, which is apparently broken on irix
- need to learn more about unix scripting languages (advice from Seth: "it can be a bit obscure")
Present: Patrick, Vitaly, Luka, Sean
Patrick:
- Pervasive computing is a good class, very pertinent to what we're doing with the signs
- This weekend I saw some active signs at JHU, will figure out their story
- Many many issues with buggy API code from Jason; think the issue has to do with triangulation caching
- Seth says: try a lazy computational structure, where triangulation is done as needed (instead of as a pre-processing step)
Vitaly:
- Wget seems to be working now, b/c of an upgrade from Adel
- Degenerate data is causing problems in basemap construction, will try to write a data cleaning application
Luka:
- The build seems to work fine locally
- My plan: to break things into different modules
- I've been looking at stairmaster lately, seems buggy
- Seth says: "This may be painful."
Sean:
- I have the camera working now
Present: Seth (via phone), Patrick, Vitaly, Mike
Seth:
- My newborn child is keeping me busy, but you can reach me over email
Patrick:
- The graphical sign half is coming along, should be ready soon
- I found some java code to help me do audio stuff, and I wrote a test program that does audio capture. Not sure how to save audio to disk.
Vitaly:
- I'm working on some visualization bugs in basemap coloring, things are now recognizable on a large dataset from campus
Mike:
- I'm plowing through the pipeline, it seems to be mostly fixed / bug free at this point
- Will add some statistics gathering eventually / soon
Present: Seth, Patrick, Luka, Vitaly, Sean, Mike
Patrick:
- Demoed the graphical sign doing 3d routefinding (looks ghetto, but it works)
- Demoed a crude java microphone app that works, with bad sound quality - maybe should get a paper cone or an external microphone
Luka:
- I was sick, didn't get much done
- CVS commits now send emails
Vitaly:
- Patchmaker (not pagemaker) is giving me all kinds of weird issues when triangles are joined
Sean:
- Demoed a version of his program, suggestions included more views of the world with/without wireframes; also trying for a different overlay color than yellow.
Mike:
- Demoed some html logging features of the pipeline
Meeting canceled due to Spring Break
Present: Seth, Patrick, Luka, Vitaly, Sean, Mike
Patrick:
- Wrote a new 2d display layer with interface
- Rewrote the networking code for the textsign / eventserver interface, reimplementing events as serialized objects
- Web screen reporting is now working - check my project research page for details
- I'll look into getting an external microphone from adel for the audio comment feature
Vitaly:
- Wrote a new rendering policy that speeds things up and is useful for debugging
- Went to NYC for Spring Break; it is very crowded
- I've been having some weird coloring bugs
Mike:
- Working on the main script currently
- Trying to generate structured error and warning messages
- Need help restricting file access to MIT only
- Seth: we should try to add end to end checks in the script (eg report 0 size files)
Luka:
- I'm working on my actual AUP now (visibility algorithms)
- I have some ideas for visibility algs (Seth: Tight bound for analytic visibility is O(n^9) (Luka is sad)
However, conservative algorithms can be much faster.)
Present: Seth, Patrick, Luka, Vitaly, Sean, Mike
Patrick:
- I have some DHCP issues with the web sign reporting all wrapped up now (thanks Adel)
- I have the "add comments" button on the textsign implemented; audio comments are saved as a .wav locally
- I'm working on a 3d visualization layer; should take 1-2 weeks
Mike:
- I've rewritten a lot of stuff in Python
- Dejunk seems to time out a lot, this is caused by data structures w/ circular references
Vitaly:
- Made progress on some coloring bugs, which were caused by (a) an incorrect comparator and (b) some OpenGL inconsistencies (Seth: It's the MS implementation? That explains a lot.)
- Seth: try coercing incoming data to a grid pt - that should fix some errors
- Demo'ed the basemap program (pretty cool)
Luka:
- Firewalk is working; I spent some time with tcl and learned about the UI
- Starting working with the visibility module and writing code