Author Archives: pgfeldman

Tom DeVito 10.28.2011

Start: 10:10

  • Keep forgetting to write out my objectives at the beginning of the day.  Will try to remember tomorrow.
  • Made some more changes to the Data Dictionary to independently identify and send remote data.
  • Continued to work on the Audio Master.
  • Changed the sound controller class on the arduino side to AudioElement.  Working on changes to make it work with the Master.

end: 6:10

Thomas DeVito 12.27.2011

Approximately two hours were used to try to resolve our server issues.

Start: 10:20

  • Retooled the audio controller on the arduino side to work with the Audio Master on the pc side
  • Started working on the Audio Master.

End: 6:20

Kristi – 12.28.2011

Mike and I met with Tangie Hall today re: Visibility documentation for both the Project Assistant and Project Manager.

Tangie mentioned that she has been unable to use the Project Assistant as there was confusion about the splash page (i.e., the flashing buttons indicating that the page was down – avoided Tangie and her staff to use the page).  Suggestion was made to make the splash page appear more like a traditional logon/password page with a tutorial button.

Tangie is taking back a set of documentation for her to mark-up while she goes through the Project Assistant.

She also suggested that she would like a “print report” like option for both tools.

Overall the meeting was very productive – despite the inability for her team to use the tool to this date.  I am hopeful that we will work well together and accomplish much in the upcoming weeks.

Phil 12.28.2011

8:00 – 4:30 VISIBILITY

  • Watching the OReilly video OpenGL and 3D in HTML5. If you have an ACM membership, you can watch it from your Safari bookshelf. You do have your FGM-provided ACM membership, don’t you?
  • Looks like picking is going to be an issue, but we can do it old-school: http://www.khronos.org/message_boards/viewtopic.php?f=43&t=3336
    • If you use C3DL, picking is supported. Hmmm.
  • Learning WebGL cookbook
  • WebGL Cheat Sheet (API as of February 2010)
  • glMatrix (appears to be the fastest library?)
  • C3DL looks promising. It even explains why Mike’s firefox WebGL was turned off. Going to read this one tomorrow.
  • 11:00 meeting with Clif S. and his gang
    • Showed the various VISIBILITY apps. It looks like they should be useful
    • Discussed WebGL development option

Phil 12.27.11

8:00-4:00 VISIBILITY

  • It’s going to rain today and my new hardware aches. This “feeling the weather in the bines” thing might be real after all.
  • Working on a simple animated gwt app. I’m trying for this.
  • Having all kinds of problems with our server. Using vi on a 20 line file takes between 4 and 12 seconds to open. Tom is opening a ticket with Steadfast.net.
    • It appears that this is a problem with Steadfast. They are doing some hardware migration that is messing with us
  • I have the gwt app running in hosted mode, but I can’t get it running as a war file. I get an error saying that the app may need to be recompiled. Hmmm.
  • Did make contact with Clif – his folks are coming over Wednesday at 11:00. Didn’t get ahold of Tangie.

Phil 12.26.11

8:00 – 4:30 VISIBILITY

  • Happy Boxing Day!
  • Tried sending an email to Clif S. At all the various permutations of his name and none of them seem to be right. Hmmm.
  • fgmdev.com is still responding really slowly. http://www.fgmdev.com:8080 seems quick though, so the Tomcat server is running well, but lampp isn’t?
    • At 9:30, things are considerably faster.
    • At 9:31 things are slow again
    • Had to start up svn. What is going on with our server!!!?
  • Starting with the gwtgl documentation.
    • And while we’re at it, time for a new version of Eclipse (3.7.1 – Indigo). Trying a 64-bit version again… I also copied it to fgmdev/exchange/EclipseDistros if you don’t want to wait 17 minutes for the download
    • Tried the tutorial, which didn’t work (and from the comments, doesn’t work for many people). Checked out the code from https://gwtgl.googlecode.com/svn/. That kindof works…
      • Only check out GwrGl and GwtGl-techDemos. GwtGL-examples doesn’t seem to have anything in it, and competes for the same namespace as techDemos (which checks out as “examples”)
  • New version of GWT In Action is available on fgmdev/exchange/Books

Tom Devito 12.23.2011

Start: 9:30

  • Talked to Phil about how to design the system
  • All processing should be done on the PC side.  The Arduino just receives commands.
  • Changed the Data Dictionary to recognize remote data elements and send them automatically
  • Grabbed the command and response class I made earlier and put them into the new project.  Add command verification and uid.

End: 6:00

Phil 12.23.11

9:00 – 4:00

  • Prepping for interview. It went well. They want to visualize lots of data.
  • In thinking about the above, I’ve started to worry about the graphing capability within GWT, which is not really even as good as Flex. I’m starting to think that we could use GWT for what it’s good at – text fields, comboboxes and such, while developing standalone charts in WebGL (3D). The reason for this is that once we’re in the 3D realm, drawing cool things actually becomes easier. Shiny is done with reflection mapping as opposed to layuers of vector art. Hardware accelerated 3d is capable of showing tens of thousands of points in a frame – orders of magnitude more than 2D systems
  • I asked Mike to try putting a gwtgl app inside jebgl to see if its possible

Tom DeVito 12.22.2011

Start: 10:20

Goals

  • Finish the main on the arduino side with all changes accounted for
  • Finish AudioTactileExec on pc side.

Tasks Completed:

  • Fixed all the variable definitions and datadictionary entries to have entries for each fingers collision, state, and pressure.  Before everything was bundled in state
  • Added the new variables to the loadbuffer method which prepares the buffer for sending
  • Did the same stuff on the PC side
  • Fixed the main loop on each side so that they recieve data, process data/run simulations, prepare the send buffer, send the buffer, and then clear both receive and send buffers.  I had it running in a different order before which I think was causing the data to be overwritten before it was processed.

End: 6:20

I spent about 30 minutes trying to figure out an office internet problem.  This time is not charged to the project.

The slowness of the blog Dong observed, seems to have gotten worse.

Phil 12.22.11

9:30 – VISIBILITY

  • The days are getting longer!
  • Downloaded SmartGWT. I talked to Dong about the risks of mixing them, as well as the advantages/disadvantages of each. He thinks that GWT is easier to work with.

Thomas DeVito

Start: 12:00

  • I either forgot to set my alarm or did 3 math problems in my sleep.  I guess I should up the difficulty on the math problems required to turn off my alarm.
  • Wrote out a summary of how the software is supposed to work to help declutter my approach to getting it working.
  • While writing this I have decided that starting tomorrow, I will write out my goals for the day on the blog, before I start working.  Too often do I jump between things and end up confusing myself.  A checklist of goals should help a lot.
  • Worked on the default simulation.  This is the same simulation I have been using but now it has to run on the PC side and send the collision data back to the arduino.
  • Decided to make collision an independent struct so that magnitude and type can be bundled together.  Made changes to accommodate for this in the simulation, sound controller, and main data dictionary stuff.

End: 6:00

Dong Shin 12.21.2011

  • developing PPM with GWT
    • spent past few days creating UI mimicking PPM
    • mixing GWT with SmartGWT is not very good… Some dynamic object creation throws exceptions that are hard to find where it came from.
    • getting list of projects from database
    • displaying list of projects in SmartGWT
      • special parsing required to use their ListGridRecord

Phil 12.21.11

9:30 – 2:30 VISIBILITY

  • Hmm. Can’t seem to edit an entry when using IE. Odd. May poke at this later…
  • Heading over to the fort for the first time since I broke my leg. Hopefully my glasses are still there
    • Talked to Tangie. We’re going to try to find some number (10-ish) of non TST users that we can get to enter data before the end of the year. If that works, then we’ll try to get all the non TST folks by Jan 17th.
  • Went poking around looking for drawing capabilities in GWT. Canvas is supported by GWT and appears to be supported in all browsers starting with IE9

Tom Devito 12.20.2011

Start: 10:00

  • Added a temporary simulation class to act as a placeholder til we have a real one.  This could also be used to process the inputs and outputs for the animation.
  • Trying to figure out the best way to manipulate the data.  Currently what I have is a call to recieveData, which imports the data from the buffer into the appropriate data dictionary entries.  Then I run the simulation so that the data gets updated as needed, it also grabs user inputs.  After that it sends the data to the arduino.  I am starting to think that maybe I should seperate the status struct into two parts so that the arduino sensor data doesn’t get overwritten when the pc sends data.
  • Made sensorData independent from the state struct so it didn’t get overwritten when the comes back from the pc with changes.
  • Took name out of the state.  If I need the name for the output, I can get it from the dd entry name.
  • Fixed collision processing in the SoundController class.  The collision processor accepts integers between 0-900.  These values can be negative because it takes the absolute value of the collision data.
  • Decided to make a simulation base class which all simulations will inherit from, instead of the placeholder I started.  Some common variables all simulations will share is max position, scaling factor, deadzone,  sensorData input, output collisions.

End: 6:00

Tom DeVito 12.19.2011

Start: 10:00

  • I was sick on Friday so I am continuing the work started on thursday.
  • Added methods for getting the state and command struct pointers
  • Added a data dictionary entry for each fingers state and command struct
  • Making changes to the pc side to accommodate the new structs.

end: 6:00