Background
T244870 will enable us to test DiscussionTools on specific talk pages, prior to DiscussionTools being enabled for the entire wiki.
This task is about using DiscussionTools on a to-be-defined list of target wiki (T233627) talk pages to ensure the comment parser (T232780) is working reliably. [1]
Open questions
- What talk pages should we test DiscussionTools on? These should be talk pages that contributors use heavily so we can be relatively confident these pages contain a large percentage of the conventions contributors use across the wiki.
Done
- Document the talk pages DiscussionTools were evaluated on
- On each of the target Wikipedias (Arabic, Dutch, French and. Hungarian):
- Review talk page comments on talk pages from the following namespaces: article talk (namespace 1), user talk (namespace 3), and project/Wikipedia talk (namespace 5)
- Document the instances where the reply link is not appearing where and/or how it should be (e.g. multiple reply links are being appended to one comment, a reply link is appearing where it should not be, etc.)
- A minimum of 100 cumulative comments [2] should be reviewed.
- "Reliable": we have set the acceptable failure rate at 0.5%. Meaning, for the purposes of this check, we will consider the comment parser to be "reliable" if <1 in 100 talk page comments are being detected incorrectly.
- "Cumulative" = 100 total comments across the 4 target Wikipedias
