Today, Terry, Philip and I discussed the system in general, and Philip demonstrated his progress with gnuplot in generating graphs. More specifically:
- Philip showed us his (impressive) graphs and the code used to generate them from the readings files.
- We discussed dropping the "0 mm" reading once we've got the start-up sequence implemented.
- We discussed speed requirements.
- There aren't any strict requirements because gnuplot is pretty quick even on the NAS box.
- We can generate at least some of the graphs ahead of time (strategy not yet decided).
- The format for the readings when pulled out of the database will be very similar to the readings.csv files - this seems to be a good format.
- We discussed how to integrate gnuplot with the framework - we can just launch a subprocess for this.