THE CODE WE NEED

If CanTags becomes a standard, it will encourage the development of innerward and outwardly facing code, making it easier to use and employ CanTags.

GEOGRAPHIC SMARTS

Once you have tags that can identify cities — or even postal codes — it makes sense that you create an open source (say, Drupal-based) automatic geographic taxonomy.

If I want to pull in arts content from all over Ontario, I shouldn’t have to hand-aggregate (or even automatically aggregate) feeds from all the towns and cities in Ontario. Nor should users have to explicitly tag their content by province.

Rather, it's easy to construct or inhale a list of cities and towns in Ontario, and then allow the system to construct the aggregate of Ontario content based on that list.

The system would know that Toronto, Hamilton, Windsor, London, etc. are all part of the same province.

Geographic taxonomies are simple and reliable since places don’t move and few towns and cities change their names, or disappear.

GEOGRAPHIC VISUALIZATION

It is only natural that aggregated tags should be displayed on a zoomable, onion-skinned map.

You should be able to see a map of Ontario, zoom in on arts content in Hamilton or London and add and subtrack onion-skin layers representing individual or collective topic tags.

BACKEND CLOUD ASSISTANCE

Editors and other journalists (or citizen reporters) using content management or blog systems should be able to have appropriate tags suggested for them based on a content-analysis of the material they’re editing.

For example, for mentions of “Hamilton," “music” or “opera," suggest a #hamarts tag.

The system should list potential tags that can be editor approved or rejected.

FRONTEND EXPOSURE

Tags applied to stories or content should be visible to the content audience, and audience members should have the opportunity to add additional tags — which should be washed against a canonical list for consistency.

EASY TAG CLUSTERING

Users should be able to easily assemble tag clusters from a variety of related tags and/or sources.

For example, it should be easy for a user to build an RSS feed of content tagged #hamarts, #hamfoo, #hameats and #hampol.

And, users shouldn’t have to build separate feeds for Flickr, del.icio.us, Twitter, etc. and frakenfeed them together.

There should be a service that automatically polls selected content platforms and assembled the uberfeed automatically.

TEXTEXPANDER TAGS

A Type4Me, TextExpander list of tag macros should be created to make it simple to enter consistent tags on content.