Difference between revisions of "WorkingDPL"
(→What (summary)) |
(→Why this is important) |
||
| Line 9: | Line 9: | ||
== Why this is important == | == Why this is important == | ||
| − | + | In our effort to provide [[CommonAnswer]] we believe our strength is our ability to address long-tail queries like, Table Tennis Clubs London, not individual search terms like Table Tennis. We desire mechanisms that allow us to write '''and maintain''' useful articles quickly. | |
| − | |||
| − | |||
| − | |||
== [[DoneDone]] == | == [[DoneDone]] == | ||
Revision as of 18:30, 28 August 2007
What (summary)
Support the creation of category-like pages from combinations of large or specific categories. This is one step along the way to making categories behave more like tags in other sites.
By "working" we mean a mechanism that works efficiently enough to be deployed under high server load, unlike DPL that doesn't.
Why this is important
In our effort to provide CommonAnswer we believe our strength is our ability to address long-tail queries like, Table Tennis Clubs London, not individual search terms like Table Tennis. We desire mechanisms that allow us to write and maintain useful articles quickly.
DoneDone
- Works without taking our site down (less than 5 seconds per query for sysadmin use, less than 1 second per query for general use).
- List is generated in Wiki format.
- Syntax is
<page_list> Category:Portland Category:Car -Category:Honda - </page_list>
Retrieve a set of pages that matches some specific criteria
- Has Category: X
- Doesn't Have Category: Y
- Linked to from Page: Z
- Not Linked to from Page: Q
Set operations on page lists:
- Intersection
- Needs to give the ContentTeam the information they need to be able to create good pages CommonAnswers and CategoriesAsCommonAnswers.
- Needs to provide information about server costs so that we can extrapolate into the future.
