Right now, edits to templates are described per the same, high-level actions as other nodes -- i.e. insert / remove / change / move. Like Text, however, there are likely subcategories of Template edits that we care about -- e.g., how many parameters changed and which ones? We should consider defining a taxonomy of Template edit types and determining what it would mean to be able to distinguish between these different types.
The reason for distinguishing between these subtypes (beyond more info = better for some applications) is that certain actions have very different patrolling implications or difficulties depending on what was done. For example, changing a link destination could be a sneaky form of vandalism and editing captions is a specific (and difficult) need on the wikis around accessibility/understanding; there are lots of different table actions but changing the header row generally has no different implications/difficulty than changing a data row.
Other potential nodes with rich sub-types:
Wikilinks (title vs. anchor) -- seems important and relatively easy
Media (file vs. size/position parameters vs. caption) -- seems important and relatively easy
Tables (add/remove rows/columns? change header?) -- would be harder and maybe lower priority
Headings (title vs. level) -- seems easy but less clear about importance
Right now, edits to templates are described per the same, high-level actions as other nodes -- i.e. insert / remove / change / move. Like Text, however, there are likely subcategories of Template edits that we care about -- e.g., how many parameters changed and which ones? We should consider defining a taxonomy of Template edit types and determining what it would mean to be able to distinguish between these different types.
The reason for distinguishing between these subtypes (beyond more info = better for some applications) is that certain actions have very different patrolling implications or difficulties depending on what was done. For example, changing a link destination could be a sneaky form of vandalism and editing captions is a specific (and difficult) need on the wikis around accessibility/understanding; there are lots of different table actions but changing the header row generally has no different implications/difficulty than changing a data row.
Other potential nodes with rich sub-types: