Wed May 14 2008

  • discussed Final Paper
  • Jeff presented Vernier client-server solution
  • talked about documentation:
  • did course evaluation

Mon May 12 2008

  • Matt & Jim described issue with SunSPOTs (have to re-deploy code to base station every time). We proposed solution where base station code is deployed to a remote so it will run that code on power-on. Maybe base station itself has flash and can accept code in this fashion
  • Ryan showed Super Cricket stuff
  • fixed bug with assignment of sensors to streams in Wizard
  • tested 6 simultaneous sensor streams
  • talked about CricketTemperatureSensor class -- Ryan will refactor his work on this
  • Will described internal (to server) architectural change he is planning

Wed May 7 2008

  • looked at Dan's new batching code for uploading to server in iSENSE/dashboard/
  • had a long conversation about how to solve network congestion issue (e.g., launching threads to upload pieces of SVLs, uploading data only at end of collection vs. continuously)
  • decided in needs to be thread-safe (using synchronized methods) -- one method will chop off a StampedValueList to be uploaded, another method will upload it
  • Dashboard applet should make sure upload thread has finished before allowing the applet to terminate!

Fri May 2 2008

  • introduced Mantis
  • revisited workflow diagrams of Experiment Manager (below) – click to enlarge

Fri Mar 28 2008

  • Dan & Ryan demo'ed working Dashboard which retrieved data from Cricket and stored in DB
  • We designed a Session Setup Wizard for the Dashboard that will walk users through attaching required data streams for their selected experiment to the physical sensors they have available; Dan & Ryan will work on this
  • We designed a Create Experiment sequence that will be implemented as a web interface using Will's JS APIs; Will will work on this
  • Justin explained the process of choosing an experiment and then session to be graphed or mapped. He and Dan will work on this.
  • the Flot demo now shows live data we collected in class today!!

Mon Mar 24 2008

  • dashboard release will install 1 stream of data hard coded as temp
  • then consider adding multiple sensors
  • think about how to use JS instead of Java for dashboard
  • definitely build JS interface to viz tools
  • look into new Google viz engine

Mar 10 2008

system block diagram -- click to enlarge

Mon Mar 3 2008

  • 8080 server is presently running on 9090, Will will put it onto 8080 later today
  • redirects to port 8181, where the Tomcat server lives
  • sudo screen -r on isense machine will bring up the SenseServer.jar session (in will's directory) if it is running
  • we should probably change the tomcat username and password -- Will will do this
  • we looked at the Simile stuff -- very promising. Justin is going to figure out how to read XMLs into this.

Mon Feb 25 2008

  • Will showed servlet technology for server client
  • tried to debug Ryan's serial port setup

Wed Feb 20 2008

went over Will/Matt's DB design and API design; see:

Tue Feb 19 2008

Fri Feb 15 2008

  • Naming discussion
    • Fred pointed out issue of agreeing on meaning of terms (session vs. experiment)
    • See Chap Getting Design Done in Rosenberg
    • “You know you're in trouble when your project has a component called the 'superwidget'!”
    • Matt said this happens often in his co.; the docs person has trouble getting engineers to agree on what they're calling things
    • Dan said in his co. they useReallyLongVariableNamesToDisambiguateAndSolveThisProblem
  • next Mon is holiday; next Tue is class
  • Sam Christy will visit on Fri next
  • Integration demo for Fri
  • Ryan/Dan showed Dashboard app. Comments:
    • when user creates session, it should go straight into data-capture/record-play screen
    • user can hit start/stop to create discontiguous sample streams within the session
  • We talked about server API
    • Will/Matt will get together to spec out first live API
    • Jeff will have stuff to show us that we'll figure out how to incorporate later
  • Will showed SunSpots
    • They are cool
    • We must use them

Wed Feb 13 2008

  • went over dashboard and its MVC arch. Realized that Start/Stop button and Sampling Rate should be located outside of sensor-specific displays:

    * looked at DB design of session-data-type:

Monday Feb 11 2008

  • I added the DemoCode page, and I put my network protocol demo code on it, along with a PDF of the network protocol. (Will)

Friday Feb 8 2008

  • Dan/Ryan: check out Blake's code, learn how to build it, will use own Macs, start working on serial, minimal MVC that shows real time data, random data that takes up real time (e.g, 1 minute) with selectable sample rate. Post applet online somewhere.
  • Tor/Jim: checking out Blake's code (see DB stuff), learn how to build applet, need trust for arbitrary net connection, post it somewhere so people can play with it.
    • Jim: display average of all Lowell/Dracut pts, or, latest point, or # of points or something else on Google Maps.
    • Tor: time-series graph on SIMILE
  • Will/Matt
    • create network protocol, prototype client/server classes, have code that can compile, think more about DB design
    • when is set up, install PostgreSQL
    • can prototype on icodeproject in meantime using existing MySQL DB

Wednesday February 6 2008

Monday February 4 2008

Friday February 1 2008

First meeting