All The Tropes Wiki Project, Part XXI

[robkelk]

Important heads-up:

I'm trying to prune subpages for stuff like Funny/Film and other things we have dedicated categories for on the main page (to eliminate repeated content and to cross out red links for wanted pages), and in the process, I found a lot of works do not have {{worksubpagefooter}} on other subpages. While all new subpages will include it by default, you can easily tell if the source page and related existing subpages do not by seeing if they have a blue link for the category or if the category itself links all available subpages. If it doesn't, they all need fixed to replace all categories save the subpage category to just have {{worksubpagefooter}}, which automatically assigns the base page name of the page as the category and will do so in case the page needs a rename in the future, as it will auto-redirect to the new page name.

If anyone wants to double-check my work, please do. Note, this is not something easily fixed by bot, so it needs done by hand.

One note on this: {{worksubpagefooter}} and {{tropesubpagefooter}} (identical except that {{worksubpagefooter}} puts the name of the top-level page in italics to match our Style Guide) only work properly on a first-sublevel page ("FOO/subpage"). They break on a second-sublevel page ("FOO/subpage/subsubpage"), and there's nothing that I can do about that with MediaWiki's current collection of magic words.

[GethN7]

Important heads-up:

I'm trying to prune subpages for stuff like Funny/Film and other things we have dedicated categories for on the main page (to eliminate repeated content and to cross out red links for wanted pages), and in the process, I found a lot of works do not have {{worksubpagefooter}} on other subpages. While all new subpages will include it by default, you can easily tell if the source page and related existing subpages do not by seeing if they have a blue link for the category or if the category itself links all available subpages. If it doesn't, they all need fixed to replace all categories save the subpage category to just have {{worksubpagefooter}}, which automatically assigns the base page name of the page as the category and will do so in case the page needs a rename in the future, as it will auto-redirect to the new page name.

If anyone wants to double-check my work, please do. Note, this is not something easily fixed by bot, so it needs done by hand.

One note on this: {{worksubpagefooter}} and {{tropesubpagefooter}} (identical except that {{worksubpagefooter}} puts the name of the top-level page in italics to match our Style Guide) only work properly on a first-sublevel page ("FOO/subpage"). They break on a second-sublevel page ("FOO/subpage/subsubpage"), and there's nothing that I can do about that with MediaWiki's current collection of magic words.

I assume that mostly hits recaps/source pages, which tend to be second-level subpage heavy (and some character pages). That can be worked around with a navbox template on all secondary subpages that lists them all and links back to their parent primary subpage, with the navbox itself axing the busted function and merely linking back to the proper primary subpage to avoid that issue.

The only downside to this is that each work where we need a navbox for a group of subpages will need its own custom navbox as a link repository for all the secondary subpages, but it's something Wikipedia does all the time to deal with this exact issue and we can just modify each one to work around that issue Rob noted as appropriate.

[Black Aeronaut]

Guys.

I can't believe we don't have this listed as a trope.

Get It Off Me!

As in the event whereupon some hapless fool makes the horrible blunder of provoking something small and seemingly harmless (and sometimes even downright adorable looking) into attacking them. Often with shocking effectiveness, causing the victim to run around in circles screaming "GET IT OFF ME! GET IT OFF ME!"

Or is this just filed under a different name and I'm just not creative enough with my google-fu to find it?

[robkelk]

We appear to have almost nothing in the way of "panic" tropes. We should fix that.

[Bob Schroeck]

First start a category, and put "Eek! A Mouse" in it...

[GethN7]

We appear to have almost nothing in the way of "panic" tropes. We should fix that.

Good idea.

Meanwhile, while combing through the extensions list for things that could be useful for better searches, Miraheze has this:

https://www.mediawiki.org/wiki/Extension:AdvancedSearch

It enables finer-grained searches, like specifying all pages in the categories that both in "Comedy Tropes" and "Smile Tropes", adding all subpages of a specific page to the search request and filtering out anything else, and other more refined search features. Claims to require CirrusSearch, but that is a "soft" dependency, it simply can use that as the search engine backend if it's available, but will default to the core search backend if it's not available, which Miraheze has defaulted it to. (the CirrusSearch extensions are buggy anyway even if it is enabled)

Note: This is stable and actively supported by the WMF, so its compatibility is very high.

All it simply does is add an optional "advanced search" to narrow down search options more, regular search can be used if you have no need of it.

You can see how it works here:

https://www.mediawiki.org/w/index.php?se...arch&go=Go

Note: Best used for very specific searches, overly broad searches will simply fail and report an error, but this would be useful for needing to find specific pages without needing wade through as much clutter. This would make our search, even without CirrusSearch as the backend, far more versatile than the TV Tropes version, which is dependent on Google and can be overly broad, while this would be far more specific and limited to our internal database.

[Bob Schroeck]

Huh. That sounds familiar, I wonder if we were using it years ago and just stopped.

In other news, regarding KF and its users, we may have less to worry about than we initially thought. Matomo Analytics shows only four links from there in the last month, which I'm going to take as encouraging.

[GethN7]

Huh.  That sounds familiar, I wonder if we were using it years ago and just stopped.

In other news, regarding KF and its users, we may have less to worry about than we initially thought.  Matomo Analytics shows only four links from there in the last month, which I'm going to take as encouraging.

Most of the time we only get linked in their TV Tropes thread whenever they want to slag off on TVT or compare and contrast something boneheaded they do to something sensible we do (or at least less boneheaded).

Given their super high SEO, I'd prefer not to be the major target of their mockery, but being linked to so they can snark at something TVT does in comparison is cool.

Also, they tend to like using archive links quite a bit, which can skew the Matomo tally. One thing they do for stuff they fear someone will try to delete evidence of is to immediately make an archive link of it, and many of them use VPNs and proxies which further skews the analytics. We typically don't have a habit of burning evidence of anything embarrassing, we tend to own it and be transparent, so we rarely if ever have had them seriously worry we'll do what TV Tropes does and try to delete everything that could embarrass them. In short, acting like adults has redounded to our credit on that score.

As for the search thing, not to my knowledge, but enabling it should be harmless and easily removed if it causes issues since all it does is add another search mode to the Special:Search page, but does not change the function of the default search at all.

[Bob Schroeck]

Well, I'm for trying it. Rob, what do you think?

We typically don't have a habit of burning evidence of anything embarrassing

I don't think we've ever had anything embarrassing that wasn't from TV Tropes to start with...

[robkelk]

Well, I'm for trying it.  Rob, what do you think?

I so very want AdvancedSearch for the freebie wiki - so much that I can taste it.

Oh, and we can add it to ATT, too.

We'll have to ask for it, though. Let me know the Phabricator ticket number and I'll piggyback on it for my request to add it to the freebie wiki, too.

We typically don't have a habit of burning evidence of anything embarrassing

I don't think we've ever had anything embarrassing that wasn't from TV Tropes to start with...

Hey, if somebody wants to help archive our pages in the Wayback Machine, more power to them. That's an extra level of backups, however spotty, in case of catastrophic deletion.

[GethN7]

Well, I'm for trying it.  Rob, what do you think?

I so very want AdvancedSearch for the freebie wiki - so much that I can taste it.

Oh, and we can add it to ATT, too.

We'll have to ask for it, though. Let me know the Phabricator ticket number and I'll piggyback on it for my request to add it to the freebie wiki, too.

We typically don't have a habit of burning evidence of anything embarrassing

I don't think we've ever had anything embarrassing that wasn't from TV Tropes to start with...

Hey, if somebody wants to help archive our pages in the Wayback Machine, more power to them. That's an extra level of backups, however spotty, in case of catastrophic deletion.

Advanced Search can be added by any bureaucrat, we'd need to request the CirrusSearch thing if we want that, and it will work without that, so we can just enable the extension in the ManageWiki settings since it's provided as an extension we can enable and disable at will.

CirrusSearch would be nice, but I doubt Miraheze can get that going at present, it requires a TON of resources that would require offloading it to its own dedicated server. We don't need it for the functionality of anything AdvancedSearch provides save the "deep category" search feature, which is a super fine-grained drill down that searches through categories and associated subcategories. For our purposes, that would not be overly important or essential at present.

Besides, the "deep category" feature is a known buggy feature that can and will die on WMF servers, so it's not very reliable to begin with (and would need manually enabled only after CirrusSearch was added), so basic AdvancedSearch will be more than enough.

Update: Both Moderation and the Edit Profile feature are not easily accessible, with the Moderation feature generally being visible on the main site only when we see the notification (if we check the log and don't do anything, it disappears until we get more edits in the queue). The EditProfile feature is not even visible on Special:SpecialPages by default (useful for removing information by request or spam). Given these are useful admin tools, I added them to an "admin only" section on the sidebar for easy access, only those with proper rights can use them.

[robkelk]

Advanced Search can be added by any bureaucrat, we'd need to request the CirrusSearch thing if we want that, and it will work without that, so we can just enable the extension in the ManageWiki settings since it's provided as an extension we can enable and disable at will.

I didn't see it when I went looking - which tab is it under? EDIT: Never mind - I just found it, and activated it on the freebie wiki. Looks like it takes a while to load, at least for the first time. RE-EDIT: More than just "a while". at this rate. Logging of for the evening now; let me know whether the freebie wiki's search page gets the extension any time soon.

[GethN7]

Advanced Search can be added by any bureaucrat, we'd need to request the CirrusSearch thing if we want that, and it will work without that, so we can just enable the extension in the ManageWiki settings since it's provided as an extension we can enable and disable at will.

I didn't see it when I went looking - which tab is it under? EDIT: Never mind - I just found it, and activated it on the freebie wiki. Looks like it takes a while to load, at least for the first time. RE-EDIT: More than just "a while". at this rate. Logging of for the evening now; let me know whether the freebie wiki's search page gets the extension any time soon.

I just enabled it for ATT as well, and according to the documentation, this is normal, it essentially needs to have the backend servers update to accommodate for it. If it's not working after 24 hours, I'll give the Miraheze staff a heads up.

Got it working on ATT, try typing something in the search bar, it should flush the loading bar and it works just fine.

NOTES: Due to the fact we do NOT have Cirrus Search, the "Structure" Search settings in Advanced Search are kinda useless, but the other advanced search options work just fine. Considering this lets our searches for certain page text be finer-grained than before, that's still an improvement. We also can search for specific file types by size (width and height), which regular search did not let us specify.

[Bob Schroeck]

With luck this will also minimize the redundant results. I'm so tired of doing a search and seeing the same pages show up multiple times.

[Bob Schroeck]

Get It Off Me!

Um, BA...  you need to create new tropes in the Trope Workshop, with the proper template.  We're going to have to reject your proposal as it's currently made.

To use the Trope Workshop, go to the menu on the left edge of every page.  In the section "Troping Utilities" you'll find "Trope Workshop" as the second entry.  Click that.

Where it says "Enter new trope name", put the name.  Remember we can use punctuation in page names, so you can add the exclamation point you left out of your original entry's name.

Click the button that turned blue when you entered a name.  The edit window will open with the page template in place.  Paste your description in over the line that says "Describe the trope here, using more than one paragraph. Remove this line when you place a description on the page.", along with its markup.

Please remember that we need a minimum of three examples along with a good description.

You might want to read our Trope Workshop Guidelines before you try again.

[GethN7]

With luck this will also minimize the redundant results.  I'm so tired of doing a search and seeing the same pages show up multiple times.

https://www.mediawiki.org/wiki/Help:Searching


Here's how default search works (copied from above link, it's CC0 Public domain licensed), and in bold is my notes on what Advanced Search does:

• It searches all pages on the wiki with some restrictions.
  

Advanced Search enhances this to allow searching any namespace. Default search was limited to content pages, multimedia pages, or everything. There was a limited advanced searching feature with the default search, but it's less precise with exact text searches.

• The article content is searched in its raw (wikitext) form - i.e., it searches the text that appears in the edit box when you click "edit", not the rendered page. This means that content coming from an included template will not be picked up, but the target of piped links will be.
  

Mostly unchanged behavior, but this time we can be more precise in nailing down where the text appears for specific text strings

• The search functionality can be considered to operate on whole words, separated by spaces or other punctuation marks. So if your search term includes the word 'book', the results will not include pages that only have the word 'books' or 'booklet'. And if your search term includes the term 'inter', the results will not include pages that only have the word 'international', but they may include pages that have the term 'inter-national'.
  

Advanced search lets us be more precise by excluding certain terms.

• The results will only include pages that contain all the words in your search.
  

Advanced Search allows exclusion of some words and specific combinations of words to be included and excluded.

• You can search for parts of words using a wildcard (
  
  
  
  Code:

  *
  ), so 
  
  
  
  Code:

  book*
   will return results for 
  
  
  
  Code:

  book
   and 
  
  
  
  Code:

  books
  . This operation is more expensive.
  

We have slightly finer-grained searching now, we can exclude "book" and just pick up "books".

• You can search for a phrase using double quotes. A phrase can be considered to consist of whole words, so the phrase 'Prime Minister' will not be found by a search for "ime Min", but it will be found by a search for "Prime Minister".
  

Mostly unchanged.


FINAL NOTES:

Without CirrusSearch, some options like the "Structure" Search features and any other page sorting feature will not work aside from the "Relevance" option. If we ever can get CirrusSearch enabled, we can do even finer-grained searching by categories and subcategories (up to a limit of 255 pages and 5 categories) and make use of the other Structure-based search options.

By default, we should be getting all search results in the order of most relevant search results with the optional better exclusion of terms to whittle down some of the clutter if nothing else.

[Black Aeronaut]

Sorry about that, Bob. There's a reason why I don't wiki very much.

[GethN7]

I have written up a tech blog for August 2021, recommending some further tweaks to our AdvancedSearch usage (Rob, you may wish to follow this for your Freebie wiki as well) that require Miraheze staff intervention, and noted some further tweaks we can do on our end to provide better data on page changes and technical analytics.

https://allthetropes.org/wiki/Blog:Tech_...ugust_2021

I look forward to any comments on these suggestions.

Update: Added some extension recommendations and a possible fix for a longstanding page bug.

[Bob Schroeck]

'Sokay, BA. Sorry if I came off snarky, I didn't mean to. I just kinda have a mental boilerplate for that kind of thing after so many years.

[robkelk]

With luck this will also minimize the redundant results. I'm so tired of doing a search and seeing the same pages show up multiple times.

https://www.mediawiki.org/wiki/Help:Searching


Here's how default search works (copied from above link, it's CC0 Public domain licensed), and in bold is my notes on what Advanced Search does:

• It searches all pages on the wiki with some restrictions.
  

Advanced Search enhances this to allow searching any namespace. Default search was limited to content pages, multimedia pages, or everything. There was a limited advanced searching feature with the default search, but it's less precise with exact text searches.

• The article content is searched in its raw (wikitext) form - i.e., it searches the text that appears in the edit box when you click "edit", not the rendered page. This means that content coming from an included template will not be picked up, but the target of piped links will be.
  

Mostly unchanged behavior, but this time we can be more precise in nailing down where the text appears for specific text strings

• The search functionality can be considered to operate on whole words, separated by spaces or other punctuation marks. So if your search term includes the word 'book', the results will not include pages that only have the word 'books' or 'booklet'. And if your search term includes the term 'inter', the results will not include pages that only have the word 'international', but they may include pages that have the term 'inter-national'.
  

Advanced search lets us be more precise by excluding certain terms.

• The results will only include pages that contain all the words in your search.
  

Advanced Search allows exclusion of some words and specific combinations of words to be included and excluded.

• You can search for parts of words using a wildcard (
  
  
  
  Code:

  *
  ), so
  
  
  
  Code:

  book*
  will return results for
  
  
  
  Code:

  book
  and
  
  
  
  Code:

  books
  . This operation is more expensive.
  

We have slightly finer-grained searching now, we can exclude "book" and just pick up "books".

• You can search for a phrase using double quotes. A phrase can be considered to consist of whole words, so the phrase 'Prime Minister' will not be found by a search for "ime Min", but it will be found by a search for "Prime Minister".
  

Mostly unchanged.


FINAL NOTES:

Without CirrusSearch, some options like the "Structure" Search features and any other page sorting feature will not work aside from the "Relevance" option. If we ever can get CirrusSearch enabled, we can do even finer-grained searching by categories and subcategories (up to a limit of 255 pages and 5 categories) and make use of the other Structure-based search options.

By default, we should be getting all search results in the order of most relevant search results with the optional better exclusion of terms to whittle down some of the clutter if nothing else.

Search by category doesn't work, either... which is why I wanted this for the Freebie Wiki in the first place. I have filed Phabricator ticket T7740 about this.

[GethN7]

With luck this will also minimize the redundant results.  I'm so tired of doing a search and seeing the same pages show up multiple times.

https://www.mediawiki.org/wiki/Help:Searching


Here's how default search works (copied from above link, it's CC0 Public domain licensed), and in bold is my notes on what Advanced Search does:

• It searches all pages on the wiki with some restrictions.
  

Advanced Search enhances this to allow searching any namespace. Default search was limited to content pages, multimedia pages, or everything. There was a limited advanced searching feature with the default search, but it's less precise with exact text searches.

• The article content is searched in its raw (wikitext) form - i.e., it searches the text that appears in the edit box when you click "edit", not the rendered page. This means that content coming from an included template will not be picked up, but the target of piped links will be.
  

Mostly unchanged behavior, but this time we can be more precise in nailing down where the text appears for specific text strings

• The search functionality can be considered to operate on whole words, separated by spaces or other punctuation marks. So if your search term includes the word 'book', the results will not include pages that only have the word 'books' or 'booklet'. And if your search term includes the term 'inter', the results will not include pages that only have the word 'international', but they may include pages that have the term 'inter-national'.
  

Advanced search lets us be more precise by excluding certain terms.

• The results will only include pages that contain all the words in your search.
  

Advanced Search allows exclusion of some words and specific combinations of words to be included and excluded.

• You can search for parts of words using a wildcard (
  
  
  
  Code:

  *
  ), so
  
  
  
  Code:

  book*
  will return results for
  
  
  
  Code:

  book
  and
  
  
  
  Code:

  books
  . This operation is more expensive.
  

We have slightly finer-grained searching now, we can exclude "book" and just pick up "books".

• You can search for a phrase using double quotes. A phrase can be considered to consist of whole words, so the phrase 'Prime Minister' will not be found by a search for "ime Min", but it will be found by a search for "Prime Minister".
  

Mostly unchanged.


FINAL NOTES:

Without CirrusSearch, some options like the "Structure" Search features and any other page sorting feature will not work aside from the "Relevance" option. If we ever can get CirrusSearch enabled, we can do even finer-grained searching by categories and subcategories (up to a limit of 255 pages and 5 categories) and make use of the other Structure-based search options.

By default, we should be getting all search results in the order of most relevant search results with the optional better exclusion of terms to whittle down some of the clutter if nothing else.

Search by category doesn't work, either... which is why I wanted this for the Freebie Wiki in the first place. I have filed Phabricator ticket T7740 about this.

Added some notes to your ticket since ATT has similar issues and needs.

[Bob Schroeck]

On another topic entirely, I just noticed that Hollywood Apocrypha cleverly uses tabber to put the self-demonstrating and plain English versions both on the same page. Anyone have objections to me doing the same for Finnegans Wake?

[robkelk]

On another topic entirely, I just noticed that Hollywood Apocrypha cleverly uses tabber to put the self-demonstrating and plain English versions both on the same page.  Anyone have objections to me doing the same for Finnegans Wake?

Feel free!

And "Canada, eh?" should get the same treatment.

[GethN7]

Since Miraheze is looking into the big item on the list, just ignore the first item on this:

https://allthetropes.org/wiki/Blog:Tech_...ugust_2021

Aside from that, the other items I'd like a vote/comments on, especially since one of them has two different ways to solve the same problem but I'd like to know what you guys think would be the less complicated solution long-term.

[Bob Schroeck]

On another topic entirely, I just noticed that Hollywood Apocrypha cleverly uses tabber to put the self-demonstrating and plain English versions both on the same page.  Anyone have objections to me doing the same for Finnegans Wake?

Feel free!

And "Canada, eh?" should get the same treatment.

The French version of Canada, eh? is handled as a page in our French language version... I'm not sure I want to break the organization there.