The note states "Pages in the User, User talk, and Wikipedia talk namespaces are not included in the error tracking categories." Should other namespaces be excluded, such as Talk, Help talk, Module talk, Template talk? Thanks! GoingBatty (talk) 01:20, 3 May 2013 (UTC)[reply]
@Trappist the monk: Having them included in the category would make them easier to find and fix quickly. I will go fix these manually now. I trust you have a better sense of the impact of making this change when you are implementing other changes to the module. Thanks! GoingBatty (talk) 19:12, 21 February 2015 (UTC)[reply]
Right now, the tests are: (1) does the value assigned to |url= look like it might be a relative protocol uri? (2) are there zero or more characters that are not a forward slash preceding a colon? (^[^/]*:). There are at least 80 permanent uri schemes plus some number more that are provisional, historic, or other. To properly identify uri schemes that aren't correct, Module:Citation/CS1 must know which are correct. This of course, is doable. The question is: should we. This backwater is probably not the place to discuss that possibility.
Per the first two cites in this rev, it appears that protocol-relative URL now produces a CS1 error and places the article in this cat. Therefore it appears the comments on the Category page are incorrect, but I'm not confident enough to do a bold edit. Suggest changing:
The URL field is checked to ensure that it does not contain spaces. The URL may be protocol relative (begins with //). If there are no spaces and the URL is not protocol relative, then the scheme must comply with RFC 3986. Further validation is not performed.
To:
The URL field is checked to ensure that it does not contain spaces, and that the scheme complies with RFC 3986. Further validation is not performed.
Getting the url checking code right has proved to be more difficult that I thought it should be. There are other problems besides this one. I'll address this and the other problems at WT:CS1 in the coming days.
The current live version of the url checking code doesn't like fully qualified domain names. That has been fixed in the sandbox. You can probably remove the dot after http://www.afhra.af.mil and get a correctly working link.
After running my bot through this category, there are still over 4,000 articles with URL errors. Is there a way to generate a report with the most common errors, so we can see if we can fix them via bot and/or expanding AWB's general fixes? Thanks! GoingBatty (talk) 17:29, 10 January 2021 (UTC)[reply]
Yes you are right. Even so-called backlog categories are seeking attention from editors and not yet cleared after years and my point was absolutely absurd on that claim to adding Backlog template.––kemel49(connect)(contri)12:28, 1 January 2025 (UTC)[reply]
Who knows? Categorization is not instantaneous and mysteriously, sometimes lags for a long time. Perhaps these articles were added to the category as a result of this 2020-04-08 edit – the template edit dumped the 4500ish articles that use the template onto the job queue so those articles can be refreshed and that should cause them to be categorized if they weren't already.
