FANDOM


  • I was looking through a wikis Special:DeadendPages, Special:LonelyPages, and Special:UncategorizedPages. I then thought, wouldn't be great if there were a way to get a report of the three in combination?

    I would call the page Special:IslandPages, meaning the items on the list returned would be complete islands on the wiki (and may not even belong on it).

    Quick reference of what each does:

    • DeadendPages means the page has no links to any other page.
    • LonelyPages means no page links to the page.
    • UncategorizedPages should be self explanatory.

    There could be true islands which are in all three groups, uncategorized dead ends, uncategorized lonely, and lonely dead ends. {An uncategorized lonely dead end is a very sad page [or very unwanted (or plain old vandalism)].}

    This is just a thought, a musing, a whimsical notion. Take it as you will.

    Have a nice day!

      Loading editor
    • It is an interesting idea and the general idea of combined reports is something I've suggested to staff in the past. Special:insights partly does this, since the point of these reports is to keep bring them down to 0 and insights works reasonably well for this. 

      The only page missing from insights is lonelypages which I probably suggested in the past.

        Loading editor
    • Dessamator, I just looked at Special:Insights and am not impressed by what is being stressed on the list of insights.

      One item in particular seems to emphasize the wrong thing: Pages without an infobox. How about pages with only an infobox or pages without prose? Why are infoboxes so stressed, yet prose is not stressed at all?

      Also, for some wikis, images are few and far between. Why stress them?

      Dessamator the annoyance is not aimed at you, sorry if it seems that way. I needed to vent.

        Loading editor
    • Well, different users have different needs. Insights serves as a source of general maintenance reports. Just as uncategorized pages might not necessarily be a bad thing in places like w:c:dev: where CSS is sometimes stored in a subpage, and css doesn't need a category at all.

      A page without sufficient prose is quite difficult to define because some pages might only need something like 3 sentences, or some pages might only transclude a template. Community central for example is notoriously bad because of their many uncategorized pages.

      Both maintenance reports and insights are  reports that don't necessarily need to be acted upon. In fact, the background color of the numbers in insights is indicative of their perceived importance as judged by Wikia staff.

      Personally, I like insights because it provides a simple summary without jumping to several  report pages.

        Loading editor
    • I still wish I could get a list of the uncategorized lonely dead ends or any combination there of.

        Loading editor
    • This is quite an interesting idea - essentially, being able to mix different special page lists to see what pages exist on more than one. I'll bring this up with the team.

      If you have other ideas for cunning ways of combining existing reports, do let us know!

        Loading editor
    • Kirkburn,

      There could be other combinations of special pages where the items which match both criteria are listed first:

      Uncategorized categories mixed with Unused categories.

      Uncategorized templates mixed with Unused templates.

      Uncategorized images mixed with Unused images and Unused videos.

      Then there is the possible wanted items with the same name in different namespaces. Let's say there is a group of articles about a smallish group of people who are barefoot. So editors put articles in [[Category:Barefoot people]] and also thought [[Template:Barefoot people]] would be nice too, but there isn't even an article on barefoot people. Oh, and they want [[File:Barefoot people]]. (This is for a highly organized wiki where naming conventions are pretty tight.)

      As for the uncategorized, lonely, dead end listing, it might be a good idea to sort the list of pages in this order:

      Uncategorized lonely dead end pages

      (...list...)

      Uncategorized lonely pages

      (...list...)

      Uncategorized dead end pages

      (...list...)

      Lonely dead end pages

      (...list...)

      Uncategorized pages

      (...list...)

      Lonely pages

      (...list...)

      Dead end pages

      (...list...)

      However, I know that is a lot of work.

        Loading editor
    • Thanks - I can't make any promises, but we'll certainly give this some thought.

        Loading editor
    • Kirkburn, I could write a dummy article on a barefoot person for you to test out the same name over various namespaces special page. 8)

        Loading editor
    • 452

      Since I've already written a script which retrieves various special pages and maintenance categories and displays them on a single page - long before insights - I've repurposed it to create a mix-and-match maintenance list.

      I've been using it for several days, and I haven't come across any major issues.

      I'm following this thread, so if you find any bugs, or have any suggestions, leave a comment below and I'll look into it.

      There are 3 different ways to use it, and 2 options, as explained below.

      You can edit the lists of special pages and categories so they are more relevant to you.

      edit: Script removed due to question-asker not accepting it as a solution.

        Loading editor
    • 452 wrote:
      Since I've already written a script which retrieves various special pages and maintenance categories and displays them on a single page - long before insights - I've repurposed it to create a mix-and-match maintenance list.

      I've been using it for several days, and I haven't come across any major issues.

      I'm following this thread, so if you find any bugs, or have any suggestions, leave a comment below and I'll look into it.

      Indeed, the original was very nice, and I still use it from time to time.

      This looks good overall. A minor optimization tweak would be to use the categoryinfo api to just fetch the number of pages in a category, rather than all its contents until there is a need to do the intersection. Unfortunately, there isn't a similar api call for Special reports probably because this would be a very expensive operation, unless it was stored and cached separately.

      Another useful tweak would be to clearly separate categories from special pages because it becomes kind of confusing to see it all mixed up like that, and there's a difference between tracking categories (generated automatically) and categories populated by users.

        Loading editor
    • 452

      The original and continued purpose is having all the reports immediately available, so I won't be adopting your first suggestion.

      I had considered having them labelled with "Special:" and "Category:", but chose not to due to space concerns, and the fact that I could easily tell the difference.

      Unlike the previous version of MaintenanceReport, the list is the same as defined in the Arrays, with Specials first and Categories second. The previous version ended up mixed, due to empty reports always being listed below populated reports, as well as race conditions.

      Anyone who can't tell the difference between "Wantedpages" and "Cleanup", and want to add "Special:" and "Category:" labels for themselves can easily do so. And, as I said, anyone who uses the script should customise the lists to make them relevant to themselves anyway.


      I've updated the script with a few minor changes, as well as fixing a small logic problem:

      When performing an intersection, comparing 2 worked fine, but when comparing 3 or more lists, any time the previous comparisons resulted in an empty array, the next loop filled the array with the entire next list. But that is fixed now.

      I could have caught that bug earlier by performing more rigorous tests, but if a multi-million dollar corporation can get away with insufficient testing, I can too.


      Edit: I've also just added another user-defined array "defaults", which should contain items from the previous arrays which should be listed by default. This makes the default results more similar to the previous incarnation of MaintenanceReport, and several fewer clicks before relevant information is visible. As with the other Arrays, it should be customised by whoever uses the script to make it more relevant to themselves.

        Loading editor
    • 452 wrote: There are 3 different ways to use it, and 2 options, as explained below.

      What are the 3 ways and 2 options? You went right into code without explaining.

        Loading editor
    • 452

      I said it was explained below.

      What's below? The code.

      So where is it explained? In the code.

      452 wrote:

      There are 3 different ways to use it, and 2 options, as explained below.

      452 wrote:

      There are two ways to view MaintenanceReport: a div or a modal.
      
       *Special:MaintenanceReport to auto-start in a div
       *Add <div id='MaintenanceReport'></div> to any page to auto-start in a div
       *Or, add <div id='MaintenanceReport' class="button">Maintenance Report</div>
          to any page for a button which launches a modal popup.
      

      452 wrote:

       window.maintenanceReport = {
        "options": {
          removeZero: false,  //set to true to remove pages with zero results
          addButton: false    //set to true to add a button instead of autostarting.
      
        Loading editor
    • It still has the old issue with redirects showing up as "wantedfiles". I had long fixed it for myself, and forgot that quirk. Something like this should fix it

      $(querypage.results).each(function(index, value) {
      if (value.value > 0 && querypage.name == "Wantedfiles") {
               return;
      }
      }

      Also, when a category name is added in full it spits out errors probably because the api rejects pages such as "category:category:test" because they don't exist.

      Anyway, Lady Aleena is probably not familiar with javascript tools, and would benefit from reading this:

      Help:JavaScript and CSS Cheatsheet

        Loading editor
    • 452

      Dessamator wrote: It still has the old issue with redirects showing up as "wantedfiles". I had long fixed it for myself, and forgot that quirk. Something like this should fix it

      That's not an issue for me, as it's functioning as I intend it to. If someone leaves a redirect when moving a file, then I want to know about it because it needs to be fixed - and the person warned to pay attention to my instructions not to leave file redirects next time.

      Dessamator wrote:

      Also, when a category name is added in full it spits out errors probably because the api rejects pages such as "category:category:test" because they don't exist.
      Solution: Pay attention to the fact that the neither the "Special:" nor "Category:" prefix are specified in any of the lists.
        Loading editor
    • 452, do you have this working on a wiki where I can see it in action so I can see if it is what I wished for?

        Loading editor
    • 452

      I can tell you right now that it is not what you wished for. But it is another way of achieving the same result.

      Once you've added it to your global.js or common.js - as explained on the page Dessamator linked - you can go to Special:MaintenanceReport and see it right here.

        Loading editor
    • 452 wrote:
      I can tell you right now that it is not what you wished for. But it is another way of achieving the same result.

      Once you've added it to your global.js or common.js - as explained on the page Dessamator linked - you can go to Special:MaintenanceReport and see it right here.

      Can you add support for Special:ShortPages and Special:LongPages?

      Thank you for your great tool/script :)

        Loading editor
    • 452

      113.167.200.93 wrote: Can you add support for Special:ShortPages and Special:LongPages?

      That's very funny.

      For those who don't get the joke: those two pages list every article, either sorted by the shortest first or the longest first.

        Loading editor
    • AH, I know that, but I only use the information of Top 20 or 30 from these pages, so I can detect what pages are suddenly got blank (from short pages) or get ridiculous long because of inserting much more content than needed and therefore remove or seperate it to another page.

      So you can make scipt to get a list of 20 or 30 titles from Special Short and Long pages, i think it's enough to use. :)

        Loading editor
    • 452

      ShortPages/AncientPages/FewestRevisions are great to decide which articles to improve, I use templated versions of them to suggest things to do both in the welcome message, and on my to-do list page.

      But since they're not maintenance issues, and can't be zeroed, they're outside the scope of my script.

        Loading editor
    • Could you consider add support for CacheCheck script or something similar like CacheCheck do?

      http://dev.wikia.com/wiki/CacheCheck

        Loading editor
    • 452

      Okay, I've considered it, and my conclusion is that I'm happy to wait until the daily cache purge. The "Time left until cache update" is right there in case anyone is confused.

      I'm currently testing a change to store the cached special pages data in local storage to prevent repeatedly asking the server for the same information which is known to only change once a day.

      I'd already done the same thing to my fork of the standard edit summaries script, to cache the summaries instead of retrieving the template on every edit, so it was pretty easy to add, but I'm going to test it for a while to make sure there are no bugs.

        Loading editor
    • Dessamator wrote: benefit from reading this:

      Help:JavaScript and CSS Cheatsheet

      Nothing against cute JS hacks, but aren't we supposed to use scribuntoLua as next step after templates? For the original example it could suffice to allow transclusions of the special pages, {{:Special:foo}} instead of [[Special:foo]] links. My FANDOM for JS is limited to UTCclock() and SkinSwitchButton().;-) –Dunnoob 💩 21:12, October 23, 2017 (UTC)

        Loading editor
    • Dunnoob wrote:

      Dessamator wrote: benefit from reading this:

      Help:JavaScript and CSS Cheatsheet

      Nothing against cute JS hacks, but aren't we supposed to use scribuntoLua as next step after templates? For the original example it could suffice to allow transclusions of the special pages, {{:Special:foo}} instead of [[Special:foo]] links. My FANDOM for JS is limited to UTCclock() and SkinSwitchButton().;-) –Dunnoob 💩 21:12, October 23, 2017 (UTC)

      Lua is an alternative to wikitext templates, not an alternative to JavaScript. They have very different purposes.

        Loading editor
    • All this javascripting would be nice if javascript were anabled on my wiki. I was hoping for an official programmatic solution. I guess there isn't enough demand for it to be worth the time to write the code for it. Lady Aleena (talk) 02:19, October 24, 2017 (UTC)

        Loading editor
    • A FANDOM user
        Loading editor
Give Kudos to this message
You've given this message Kudos!
See who gave Kudos to this message