Connor Bohan's blog

 The Data Annotation Interface is as done as its going to get before presentations. I'll have a week and two days to fix it up after that's done but from now until Friday its paper writing and Powerpoint making time. As for new features there are two: the interface is now (mostly) compatible with the data explorer's POST/GET variables. There's still a bug that prevents previously made comments from showing up when you click on them but that can be fixed when I get back. I also added Internet Explorer 8 compatibility.

 After meeting with one of my future users (there aren't very many of them) I've been given a few more things to add to my interface. However, a ten day break to go to Italy has not given me much time to actually add the new features. The only real change I've made since my last blog was to add little calender symbols by the start and end time shufflers (what I've taken to calling a series of buttons that allow a user to adjust the start and end time of the graph).

 This week has been mostly bug fixes. Charles attempted to use my annotation interface and would a few problems that I have been remedying. It appears that the test data that I was working with had different properties than some of the other data. This resulted in the interface sometimes failing to grab the deployment id. Without the deployment id the interface had no way to get previous comments on the data or make new comments.

 This is my second year being an intern a CMOP and like last year I'm working with Charles Seaton's data annotation project. I've switched mentors to Alex Jaramillo but that was truly the only change. My senior mentor, David Maier, is the same and I'm working with my old code.

 As of Friday 7/15/11 all the basic functionality of the commenting form is in, in theory. In practicality, there is probably at least one critical corner-case bug that I don’t know about. But this week will be centered around preparing for a preliminary launch. This week’s new features include: a beautiful graphical start and end time selecting system, the ability to show/hide the various parts of the form, comment submission times, and a new primary-key free annotation table.

 I spent the third week largely developing an interface to allow users to select which data the wanted to comment on. This involved getting station, offering, and variable data from the database and populating <select> tags with it. This task also got me familiar with writing to the backend image-writing service the creates plots based on the url you give it. This will come into play later when I have to not only write these urls but read them. Finally, I finished getting Drupal user information which is required when submitting comments to the database.

 This past week has been less than productive due to a complete inability to recreate the CMOP database on my machine. Tuesday through the Thursday was spent fruitlessly trying to rectify this problem. However, due to some issues mainly regarding plpythonu and getting it to work with the Postgres database on my computer, we decided that is would be easier just to give me read/write access the CMOP test database. Friday was spent mostly looking over the database and already-in-place code and now it’s time to get back to coding.