MarkUs Blog

MarkUs Developers Blog About Their Project

Archive for February, 2011

Meeting Minutes for February 24th, 2011

without comments

Bertan is working on the yml upload. He is testing the code to understand more about it.

Danesh is trying to implement an API for creating new users. Karen supplied him
with few links to help him with the HTML status codes.

Tobi was having problems in the login page due to session.

Karel is going to collaborate with Ibrahim on the testing framework.

Misa got a ship it, she is waiting for vivien’s changes. Once that is done they
will merge thier code. After that they will be testing and doing a screencast.

There was a discussion of markus being involved in GSoC. Developers seemed to
be interested in doing it.

Good luck on Yansong’s midterm!

Written by BertanGuven

February 25th, 2011 at 3:33 pm

Posted in Uncategorized

Punchlines for February 24th, 2011

without comments

Misa Sakamoto

Status:

  • all tests now work
  • pull request to git

Next Steps:

  • waiting for request to go through
  • need to check with Vivien on her status for changes requested on review board

Roadblocks:

  • none

Karel Kahula

Status:

  • Just finished with hopefully the last of my problems with Git.
  • Worked out my bugs from last week for issue 144.

Next Steps:

  • Completing unit tests for issue 144.
  • Collaborating with Mr. Shahin on the testing framework project.

Roadblocks:

  • None

Oloruntobi Ogunbiyi

Status:

  • Integrated the role switching code developed by people at UW

Next steps:

  • Figuring out the bug that causes the expired session message.

Roadblocks:

  • Get an expired session message every time I try to login as an admin or student or grader

Yansong Zang

Status:

  • Didn’t work on MarkUs this week because of midterms.

Next step:

  • Return back for the dashboard data task next week.

Roadblocks:

  • My last midterm this Friday 😉

Ibrahim Shahin

Status:

  • Drag and drop ordering almost done, saving the order, and updating the unit tests are left

Next Steps:

  • Testing Framework with Karel

Roadblocks:

  • Can’t run unit tests due to rake complaining about missing sham

Bertan Guven

Status:

  • Working on evan’s branch for yml_upload.
  • Testing how things work so far

Danesh Dadachanji

Status:

  • Understanding how to create an HTTP request.
  • Working on creating a controller for the API.

Next Steps:

  • Create the controller and start with a SHOW request to display a user

Roadblocks:

  • It’s hard to understand what’s happening with rake tasks and HTTP requests. I don’t have too much background info on it.

Written by Ibrahim

February 24th, 2011 at 2:40 pm

Posted in Uncategorized

Meeting Minutes for February 17th, 2011

without comments

Misa and Viv are working on passing tests that fail due to a missing due date and trying to avoid having a default date.

Yansong brought up mark distribution graphs which lead to a discussion about where the graphs should go.
There was some talk about tabs between Karen and Nelle. Look for a blog post in the near future regarding the creation of a new tab in MarkUs.

Karel will be getting Mike to look at an issue of his, mark decays not appearing in the list of submission rules.

Danesh wondered about commits for format; Code formatting commits should be separate from coding commits.

Toronto Markus devs will be having a nice lunch together sometime during Reading Week 🙂
There was also a discussion going on about MarkUs becoming a Google Summer of Code project. Very exciting!

And don’t forget, there is still a Markus meeting next week on IRC.

Have a nice Reading Week everyone!

Written by Ibrahim

February 17th, 2011 at 5:17 pm

Posted in Uncategorized

Week 5: Markus Plagiarism and Markus Research

without comments

Markus Plagiarism

Benjamin Thorent

  • Status
    • Tests on PMD performed
    • Working on Ant and how to use it with Plaggie
    • Specification Book
  • Road Blocks
    • Hard to make the link between Ant and Plaggie
  • Next Steps
    • Write a report about the tests on PMD
    • Keep working on Ant

Shion Kashimura

  • Status
    • Specification book (interfaces part + general description etc)
    • Added things in the state of art (found another Open Source text plagiarism
      detection tool)
  • Road Blocks
    • None
  • Next Steps
    • See more about Plaggie’s bugs in order to take a decision between Plaggie
      and PMD
    • Specification book (modifications)
    • See more about the integration part

Markus Research

Anthony Le Jallé & Michael Lumbroso

  • Status
    • Work on some use case.
    • Begin the scope statement.
  • Roadblocks
    • None
  • Next steps
    • Continue the scope statement.
    • Discuss with Benjamin Vialle and Nelle Varoquaux (Markus experts) about our
      specs.
    • Discuss with our teachers about the choices we took in the specs.
    • Keep, modify and find missing functionnalities they want to include.

Written by nvaroqua

February 17th, 2011 at 3:52 pm

Posted in Uncategorized

Punchlines for February 17th, 2011

without comments

Danesh Dadachanji

Status

  • Have not made any progress in the last week because of road blocks.
  • Will catch up for missed time during reading week as I have lots of free time

Next Steps

  • To catch up on the lost week during reading week do what I was meant to do last week.
  • Finish off issue 6 by adding the suggestions made on review board

Roadblocks

  • Many, many assignments due this week and job interviews

Bertan Guven

Status

  • Couldnt get anything done this week due to assignments. Therefore will work on the same stuff I said I was going to work on last week. Last weeks tasks were containing the following points:
  • Figuring out how code works with Rubrics and CSV
  • Will get a feel of YML by writing a standalone ruby program

Karel Kahula

Status:
  • Still working on mark decay feature.
  • Fixed that pesky CRLF nonsense in my repo.
Next Steps:
  • Getting this feature in order and ready for review.
  • Working with Ibrahim on the Test Infastructure.
Roadblocks:
  • Having some trouble tracking down a bug in my feature.

Oloruntobi Ogunbiyi

Status

  • Submitted a pull request for the fix for issue 169

Roadblocks

  • Understanding the workings of the code to integrate role switching
  • Problems with git
  • Home work for other courses

Next Steps

  • Integrating role switching

Misa Sakamoto

Status:

  • making changes according to review board
  • 4 hours debugging mystery bug… it’s fixed but it’s still a mystery
  • other debugging – all tests pass other than those that have to do with the nil date on Assignment remark due dates

Next steps:

  • discuss with Vivien about the nil date on Assignment remark due dates
  • more test debugging (the ones that have to do with nil dates)
  • merge with Vivien if there is time

Road Blocks:

  • mystery bugs suck

Ibrahim Shahin

Status
  • Finished writing tests
Next Steps
  • Drag and drop for grade entry items
  • Put up another review
  • Test Infrastructure with Karel

Roadblocks

  • Exams 🙁

Yansong Zang

Status:

  • Post a blog about the Dashboard Dada task plan
  • Post a blog about the Bluff Graph showing each TA’s grade distribution
  • Try to understand the part of our code which is related to the database

Next Step:

  • Make our dashboard show the TA’s grade distribution graph.

Roadblock:

  • midterms are coming

Written by Danesh

February 17th, 2011 at 12:02 am

Posted in Uncategorized

Dashboard Data — Show TA’s Distribution

with 3 comments

As I have mentioned in my blog “Dashboard Data — Plan” the first thing I want to do is “Show each TA’s grade distribution vs. the whole distribution in the graph.” To do this I should see how to collect (maybe calculate) the data to be shown, namely the mark given by each TA to a submission. I have sent emails to Severin and Mike asking this question and thanks to their help I got some basic ideas about where to seek these data. If I have more questions about that, I will post another blog later as I start do the implementation.

But for now in this blog I want to discuss several ways to present the TA’s grade distribution. Please response in the comment which way do you prefer 🙂

* Use Bluff.SideBar

Pros

  • Show all the TAs’ distribution in the same graph which makes it very convenient to compare them

Cons

  • According to my test things going to mess up when there are more than 3 TAs. The best case is when there are only 2 TAs and 3 is tolerable, but it will be very hard for the user to see the distribution when there are 4 TAs or more
  • Cannot compare each TA’s distribution vs. the whole distribution

* Use Bluff.SideStackedBar I

Pros

  • Show all the TAs’ distribution and the whole distribution in the same graph
  • Performance does not change when there are more TAs. According to my test the graph still looks good when there are 5 or 6 TAs (although I don’t think there will be that many TAs in an assignment ;))

Cons

  • Although all TAs distributions and the whole distribution are shown in the same graph, it seems harder for the user to compare each of the TA and the whole distribution

* Use Bluff.SideStackedBar II

Pros

  • Compare each TA vs. the whole distribution, thus making it’s easier to see each TA’s grade distribution out of the whole distribution

Cons

  • In fact what I expect in the above graph is yellow bars can be showed at he left side and the blue bars on the right side, but after I tried several times I cannot fine a way to change the order of these bars 🙁
  • We can turn on the tooltips of Bluff, which allows you to hover the mouse over a data bar to see the name and value of the bar in a popup box next to the mouse. But if we hover the mouse over the blue bars the popup box will give us the sum number of other TAs’ other than the total number of students in this interval. (If we store the total number of the students in the blue bars the total length of the bar will not represent the actual number of the students, in fact TA1’s number will be counted twice)

* Use Bluff.Net

Pros

  • Show all the TAs’ distribution and the whole distribution in the same graph and it seems the best way for the user to compare each TA’s distribution as well as with the whole distribution

Cons

  • tooltips of Bluff can not be used in this kind of graph 🙁

* Final Solution
Render each of the TA’s distribution and the whole distribution separately in different graphs.

I just omit the other two TAs’ graph but it’s just the same thing 🙂 Following is the whole distribution.

Pros

  • Neatest graph and easiest to see

Cons

  • Hard for comparison
  • More graph’s needed

Written by Yansong

February 13th, 2011 at 7:23 pm

Posted in Uncategorized

Dashborad Data — Plan

without comments

This term I (Yansong) will improve the dashboard feature. Currently, the Bluff Graph on Dashboard can only show the whole grade distribution of an assignment but we also want to see each TA’s distribution. Also the table on the right of the graph only shows some basic information of the assignment and simply lists all the TAs’ names who mark this assignment, but we want to show some detail information about the assignment and each TA.

It seems there will be several missions to to. Basically, this blog lists what I think so far should be added to our dashboard in order of priority.

* Show each TA’s grade distribution vs. the whole distribution in the graph.
* Show each TA’s marking progress in terms of “total # of submissions/# of completed” in the table.
* Show each TA’s average mark in the table.
* Show both the total and average number of annotations given by each TA.
* Show more statistical information about an assignment and each TA such as MIN-MAX marks, Standard Deviation and Median

So above are all my ideas so far any other comments or suggestions? 🙂

Written by Yansong

February 13th, 2011 at 4:58 pm

Posted in Uncategorized

MarkUs 0.9.4 is out!

without comments

The team is pleased to announce MarkUs 0.9.4!

MarkUs 0.9.4 fixes a bug that prevents admins and graders to release and unrealeasing students using the select-all-across-pages feature in the submissions table.

Or, if you’re running 0.9.3 and want to patch up – download the patch here (you’ll want to apply sequentially). It’s short, sweet, and shouldn’t require any special configuration. All wins.

Great job team – keep up the good work!

Written by Benjamin Vialle

February 13th, 2011 at 4:12 am

Posted in Uncategorized

Meeting Minutes for February 10, 2011

without comments

Major topics:

  • Discussed and reviewed the major projects the team members were working on
  • Some people are still finishing off their bug fixes
  • There will be a meeting in reading week on the 24th of Feb
  • Research in Action day on April 8th in Toronto, there will be MarkUs and UCOSP booths
  • France is making excellent progress!

Yansong:

  • Created a proposal for some dashboard elements.
  • There will be a blog post by the end of the week brainstorming his ideas

Ibrahim:

  • Coming along fine, finishing off some code reviews for his bug
  • Currently trying to understand the drag ad drop features

Misa:

  • Close to fully implementing her feature
  • Working on merging her code with Vivien’s

Tobi:

  • His birthday!
  • Encountered some major issues with line endings, however it has been resolved
  • Has resolved the DB issue for which Severin posted a review

Karel:

  • Has written his logic for his implementation
  • Will write a blog post for others to review and discuss it

Bertan:

  • Figuring out how code works with Rubrics and CSV
  • Will get a feel of YML by writing a standalone ruby program

Danesh:

  • Had the same issue as Tobi, it has also been resolved
  • Finishing off the code review before moving on to the next issue

Written by Danesh

February 12th, 2011 at 5:36 pm

Posted in Uncategorized

Week 4: Markus Plagiarism and Markus Reasearch

without comments

Markus Plagiarism

Benjamin Thorent

  • Status
    • Working on how to integrate a tool like Plaggie in MarkUs (and the licence
      problem related to this)
    • Specification book
  • Road Blocks
    • Hard sometimes to understand how MarkUs works
  • Next Steps
    • Test PMD (an other possibly interesting tool)
    • Specification book (mainly specific points such as workflow, memory
      issues, …)
    • Start to work on Ant and PMD

Shion Kashimura

  • Done
    • Completed state of the art for both text and source code plagiarism
      detection tools
    • Specification book (interfaces part)
    • Continued tests on Plaggie
  • Road blocks:
    • Sometimes hard to think from a grader’s point of view
  • Next steps:
    • Tests on PMD
    • Continue writing specifications

Markus Research

Anthony Le Jallé & Michael Lumbroso

  • Status
    • Work on a draft of workflow with Anthony.
  • Next steps
    • Write use cases, and try to go into more detailed specification.
  • Roadblocks
    • Meet some researchers to better define their expectations

Written by nvaroqua

February 10th, 2011 at 4:47 pm

Posted in Uncategorized