# Input

## Further Input for Discussion

Of course, the submissions are starting points for discussion and for implementation work. However, if you have further ideas and suggestions to make, please add them here. Also, please comment others' suggestions.

This section has three subsections, one for each of the main topics that appeared in the submissions.

### Collaboration / Content Sharing

#### Mathias Hiron: Collaborative Website

A first add-on comes from Mathias Hiron. After reading other submissions, he wrote a kind of requirement analysis for A collaborative website for the IOI community.

This page: Collaborative_website contains an updated list of requirements, and what various platform provide for each requirement.

Wolfgang, 2010/03/11: That's a pretty thorough analysis of what might be the content of a web-based collaboration tool for the IOI community. This Wiki is a "Semantic MediaWiki" and could be used for trying Mathias' proposals.

Mathias, 2010/05/07: I started to play with the features provided by the semantic extension, and it is more powerful than i expected. To get an idea of what's possible, start by looking at my user page User:Hiron. Then look at the source code (edit tab), and then look at the source code of the template Template:User that defines the way the data is presented, including queries on other pages (list of tasks, list of papers).

Misof 23:11, 7 May 2010 (UTC): From my experience with various wikis, I would say that if MediaWiki does most of what we need, it is probably the best choice -- due to its being used in Wikipedia it is probably the most robust wiki around, with much more stuff (plugins, support, etc.) than any other.

The Art of Problem Solving wiki does also run on MediaWiki, probably with some custom extensions -- for example, their wiki makes it possible to integrate Asymptote source code to draw figures for tasks that involve geometry.

To add to the list of features we need, LaTeX support seems pretty important to me. On other discussion boards I lack it way too often.

Mathias, 2010/05/07: I agree that MediaWiki itself is very robust. However, if we have robustness issues, it probably will be with extensions, not with the standard wiki part. So we really need to see which platform has robust extensions that do what we need.

About latex, i found that MediaWiki uses a subset of AMS-LaTeX markup, a superset of LaTeX markup". From your experience, is this good enough ? Test : $\sqrt{1-e^2}$

"To have math rendered in a particular MediaWiki installation, one has to set \$wgUseTeX = true; in LocalSettings.php.". Wolfgang, could you please check that?

On the moinmoin side, there are parsers like /mathtran or /inline_latex that claim to do the job.

Mathias 23:34, 8 May 2010 (UTC) After playing with mediaWiki + semantics, as well as foswiki (fork of twiki), my conclusion is that mediaWiki + semantic is very powerful, and foswiki quite disappointing.

the main issue with mediaWiki is that it's not designed to have fine grained access control, and to protect sensitive data. They admit that no matter which extension you use to handle access control, denying read access should be seen as a "nothing to see here, move along," sort of thing rather than a guarantee of secrecy.

that means that if we use mediaWiki, we can't put the "unseen tasks" repository in the same instance as data destined to be published, or anything that we wouldn't want everyone to see.

Mathias 10:54, 9 May 2010 (UTC) I think creating multiple mediaWiki instances might work, to avoid the access control issue. For example, a public instance with all the non-sensitive content, and a more private instance for the unseen tasks repository. The private instance could automatically import some data from the public instance, to avoid having to maintain redundant information. We could use extensions such as Data Transfer, Data Import Extension or External Data do transfer semantic data from the public instance to the private instance(s).

### Spectator Friendliness

#### IOI 2010

The GA minutes of IOI 2009 say: "... during IOI‘2010 advantage of the ‘experimental regulation’ regulation would be taken, and that scores would be more open." That is, organizers of IOI 2010 may introduce new contest procedures, aimed at higher accessibility of the contest.

At the workshop, Gordon Cormack, scientific chair of IOI 2010, will present (in a tele-presentation) the current status of his preparations in this direction. Read an early summary of Gordon's ideas, which he communicated in late February to this workshop's programme committee: I gladly publish it here, with his permission:

#### Report of the Visualization Group

At the workshop,  Jittat Fakcharoenphol and Willem van der Vegt formed a working group on aspects of visualizing contests. They produced a distinguished subgroup report.

### Research articles

In recent years, many useful articles on competitions were published in various journals and conferences. A central repository of (links to) such articles may be a useful thing to have. --Misof 12:02, 4 May 2010 (UTC)