MarkUs Blog

MarkUs Developers Blog About Their Project

Meeting minutes – October 30rd 2009

without comments

The meeting started at 12:02 and ended at 13:02.

Meeting Agenda

  • Evaluation schemes
  • Current progress
  • Start evaluating how much work is left on each feature and plan accordingly

Joel Spolsky’s Blog Post about Student Projects

  • Karen was disappointed to hear him praising Greg’s initiative in one breath and then complaining about how students don’t produce anything in the next.
  • Karen also said that MarkUs wouldn’t exist without the work done by students on course projects.
  • Karen is confident that we will be using everything (or almost everything) that we are working on this term.

Evaluation Schemes

  • Karen wanted to see the Ulaval students evaluation scheme.
  • Ulaval students posted it on the blog 2 minutes before the meeting.
  • Karen was wondering if Tara, Fernando and Farah evaluation scheme had a team component or it meant to be a team mark.
  • Tara said that since Fernando and her will be working on the Note system, a team mark for this feature is appropriate.
  • Severin said that a team is the whole team, not only who you are working with for your individual projects.
  • Farah proposed that the “Overall Process” section should be a team mark.
  • Karen said the features could also be treated as a single mark for the group that worked on them.
  • Karen is inclined to treat the maintenance marks as a separate feature.
  • Severin had the idea that team evaluation could be done by your peers rather than Karen.
  • Gabriel proposed that it could be both (peer evaluation and Karen’s).
  • Karen will always reserve final say (except when she has to defer to another supervisor).
  • Karen would like everyone to write up answers to the questions posed by Severin (or an altered list if we decide) and she will use that as one component of the evaluation.
  • Severin said (in his blog post) that Mike should come up with a team mark after each individual answered questions about team performance.
  • Mike said he knew about that part but he haven’t yet spoken with Greg or Karen about it.
  • Karen is extremely happy with the work that everyone has been doing.
  • Karen will be consulting with Mike on how to assign the grades, because he has been more closely monitoring weekly progress.
  • Karen said that the evaluation plan is now what everybody proposed.
  • Severin said that anyone with other questions for peer evaluation should comment on his blog post.

Current Progress

  • Tara and Fernando did a meeting and they will make mock-ups by Tuesday to have a feedback by next week’s meeting.
  • Farah is currently working on finishing the creation form and it’s going well so far.
  • Tara proposed a good idea on the ReviewBoard : all interface related review requests should include at least one screenshot.
  • Everyone liked the idea.
  • Gabriel, Mélanie and Simon found solutions to our design problems before the meeting and everything is great now.
  • Severin said that we should work on a backup scenario for markusproject.org because the server’s importance seems increasing.
  • Fernando was wondering what was used to generate documentation. He said an error occurs when he runs “rdoc”.
  • Severin believes there is an option to use a directory outside MarkUs source to generate documentation and it should resolve the problem.
  • Mélanie proposed to look at the database index to improve MarkUs performance because she found missing index. She is good with databases.
  • RDOC is used to generate API documentation
  • Other documentation should be posted on the blog or wiki.
  • Mike said that MarkUs’ application help/documentation could work like the ReviewBoard. Help pages are hosted on markusproject.org and linked to from the application.
  • Karen proposed it to be a major component for next terms’ students.

Evaluate how much work is left on each feature and plan accordingly

  • Simon said that Ulaval students will evaluation what they plan to do and the time needed for each tickets immediately after the meeting.
  • Karen wants to see an estimate of how much is left to do for each features into the punchlines.
  • Fernando had some problems with unit testing the logger.
  • Simon said it is harder because  the logger uses environment constants and didn’t find a way to mock it.
  • Gabriel said a solution could be a configuration object.
  • mikeg1a found a way to mock constants and gave us a link : http://www.danielcadenas.com/2008/09/stubbingmocking-constants-with-mocha.html
  • Mike told Simon to check out rake tasks (lib/tasks/demo.rake, populate.rake, and results.rake) because there could be some fixups to do for the “marking_scheme_type” in the assignment table.
  • Simon said he will also make a migration to put the “marking_scheme_type” to “rubric” for all the existing assignments.

Other questions

  • Simon was wondering if there were a weight property in a flexible criterion.
  • Karen said no. There is only a max value.

Written by simonlg

October 31st, 2009 at 4:05 pm

Posted in Uncategorized

Leave a Reply