Our aim is to build a highly-structured database. Doing so will involve a lot more work initially (in comparison with a purely aesthetic redesign) but will make changing the website easier and more fool-proof. This is also in keeping with Drupal website best practices.
- Departments
- e.g., 'health', 'ddot', 'ocfo', 'dpw'
- May not always technically be departments - e.g., they may be agencies or authorities.
- Includes offices / agencies within departments (e.g., 'web team', 'iet', 'office of civil rights')
- Departments and sub-departments need to be associated with certain pieces of information, e.g.:
- department contacts
- Name
- Position
- Phone
- social media specific to the given department
- can be associated with a topic (e.g., 'police department employment opportunities', 'police department crime & saftety')
- media & files associated with the given department
- department contacts
- Initiatives
- e.g., SisterFriends, Adopt-a-Park, Motor City Makeover, Keep Detroit Beautiful, Detroit Youth Violence Prevention Initiative
- May be related to more than one department
- May be relatively (compared with departments) short-lived
- Content category (?)
- not sure of correct name for this taxonomy - perhaps 'vertical' or 'audience'? 'target audience'?
- e.g., 'government', 'residential', 'business'
- Data Type
- e.g., 'permit', 'license', 'entertainment', 'property tax', 'how do i', 'socrata', 'esri'
- question: should 'data source' be a separate taxonomy (especially for things like 'socrata', 'esri')?
- Interactions
- e.g., 'payment', 'service', 'appeal'
- Tags
- Tags are useful as the site grows, used as a filter for search results
- Media / Content type
- these will be handled by the drupal media module
- how do we handle events?
- would be nice to be able to link to socrata data better
- avoid stale data / be able to identify stale data easily
- avoid broken links
- ideas:
- instead of linking from detroitmi.gov directly to individual socrata datasets (which are sometimes liable to disappear or go stale), our links could essentially be keyword queries, which would bring back results from socrata related to the links (if this is doable on the socrata end of things)
- socrata links should live in content types which are labelled consistently by their data source or type of content, so that we can easily identify everywhere where socrata data is used in the site and quickly identify broken / stale links
- e.g., pdfs, etc.
- map to new source location: move everything from /portals/0 to a directory on the drupal web server
- add an htaccess rule to the drupal web server's apache server to point the name 'portals/' to the new directory
- how to export redirects
- have post_row_save create redirect (old url to new url) once the row item has been saved