Summary

Alternate taxonomy for tasks

Discussion

We currently have two types of "live" tasks, Task and Template:Puzzle. I've since learned that some of the tasks under the "Puzzle" banner are classic computing questions, while others are not. I'd like to suggest an alternate taxonomy:

Practice Task -- Something which has regular importance in "real-world" programming, but may not be particularly interesting from a language guts perspective. User Output - text and Creating a Window would fall under this classification.
Theory Task -- Something which may not be practical, but is descriptive of a question programming theory and how that question maps into aa particular language. Lucas-Lehmer test and Sieve of Eratosthenes would fall under this classification.
Comprehensive Task -- (Needs a better name) A task which goes beyond basic understanding of most languages and may be a composite application of multiple concepts. RCHQ9+, 24 game and 24 game Player would fall under this classification.

I would like to deprecate Template:Puzzle, and gradually shuffle the members of Template:Task to using different types. Also, our current task categorization tree is a mess, and these could serve as a new first round of that category.

Thoughts? Alternatives? --Michael Mol 13:27, 20 January 2010 (UTC)

Yeah I thought the task tree would work out better. I'm not sure how these categories will make tasks easier to find. A bad sorting algorithm may not be practical, but if I'm new to programming I may think it is. Most likely I think a user in that situation would try to find "sorting algorithms" if they're looking for a category. Basically I'm concerned that people who don't necessarily understand the usefulness or frequency of use of a programming problem they're working on won't be able to use a system like the one you suggested. As if they could figure out the current one...--Mwn3d 14:46, 20 January 2010 (UTC)

(Geh. I need to force preview on myself.) I may be able to alleviate some of that by adding the category tree extension. However, most of RC's repeat contributors fall under the 'theoretical interest' area, while most looking for practical information appear to be coming in directly via search engines. People specifically looking for practical code to deal with their immediate interests aren't currently browsing through the SbPT category tree. --Michael Mol 14:56, 20 January 2010 (UTC)

The category tree extension probably would help a lot as long as it doesn't take up too many system resources. How the current users act is important, but also we need to design for how we want new users to use the site. If you want the users to decide that (which I guess is the wiki way), then I guess we can go with whatever seems to fit the way they use it now and see how they react. --Mwn3d 15:21, 20 January 2010 (UTC)