CategoryMindshare

Exsting implementations which deserve a higher mindshare

Oh No Robot: Web Comic Transcription and Search

This one is unexpected. Oh No Robot is a search engine where users transcribe web comics, and you can then search the transcribed text. Surprise: It actually works. In less than 7 days, they have managed to add over 7000 transcriptions. This once again shows that users are willing to add value to a free service if they are provided with easy to use tools to do so.

For comics which have joined the program, you will occasionally see little buttons like this at the bottom:

If you click the button, you can enter a transcription:

The artist can moderate the incoming transcriptions. Oh No Robot then indexes these transcriptions and makes them searchable. A very cool idea; sadly, the added transcriptions are not available under a free license. What is it good for? Finding comics you vaguely remember is the most obvious application. But I think if this scales, it’s quite likely that commercial services for licensing comics on certain topics will be built on it; finding editorial cartoons on certain topics is one example.

Alas, it’s unlikely that Google Print will add anything similar to its scanned pages soon ;-). (Actually, Google Print is vulnerable to distributed transcription or OCR attacks, which is why Google does not index all pages of copyrighted books — a subset always remains unavailable to all users.)

I think these kinds of ideas could be collected on a new page called “[[edit:Distributed work|distributed work]]” in the wiki.

Wikis: Navigation popups

Wikipedia Popups

Wikipedia user Lupin has created a very cool JavaScript-based tool for MediaWiki called Navigation popups. It should work on any MediaWiki installation which allows users to configure their own JavaScripts (this, in itself, is a fairly cool feature that few other wiki engines have). Lupin’s navigation popups show you the first sentences of any Wikipedia article when you hover the mouse over links for a while. The script can operate without causing too much server strain because it only has to get the raw, unparsed wiki text of the pages it previews from the database. All rendering is done on the client side.

Besides giving a nice preview of articles, it gives immediate access to certain functions (such as edit, move, search, discuss, delete ..). It also has some built-in helpers for fixing links that point to disambiguation pages or redirects (these are Wikipedia-typical problems and not necessarily of interest for other wikis).

You can configure it in various ways; for example, you can define after how many seconds the little “popup” (not really a popup since it does not create an actual browser window) appears. Possible future coolness include support for getting pages from other wikis and cascading popups.

I think similar navigation popup functionality should be added to other wiki engines. That’s why this post is in the “Mindshare” category, a new category for cool implementations that are not widely known. Why not help identify and collect cool [[edit:Wikis|wiki ideas]] in the ID wiki itself?