mike.lifeguard+bugs wrote:

I don't think this is possible with the current schema which tracks links between pages. Furthermore, I don't see the utility in changing things.

• Bug 21266 has been marked as a duplicate of this bug. ***

hardy wrote:

Oh, great. The fact that there is one person somewhere who FAILS to see the utility is now considered a cogent point.

When I move an article and then click on "what links here", I sometimes see a
couple of hundred links nearly all of which result from pages using a template.
After editing the template to link to the new title, I want to see the
(usually few) links from articles themselves, rather than those resulting from
a template. Currently the only way to do this is to wait 24 hours so that when
you view "what links here", the template links will no longer link to the old
title. Currently one can click on "hide redirects"; there should be a similar
thing to hide links resulting from articles' use of templates.

I reported the above as a bug, and it got marked as a duplicate of this even though it proposes something DIFFERENT.

It's a bit different, but the problem is the same: currently templates are expanded and only after that links are parsed and replaced. To fix this, we should change parsing older to parse links in the same time as we expand templates and change the pagelinks database table to know if the link is directly inserted in the page or comes through a template. Also a problem would be parameter substitution: {{Some template|link=[[Some page]]}} or {{Link|page=Some page}} where Template:Link contains something like "[[{{{page}}}]]", for those two cases, should the link be considered as comming from the page, or the template? (first example might be easier than the second one)

sorry: order, not older.

• Bug 28855 has been marked as a duplicate of this bug. ***

ekonomiska wrote:

I posted that duplicate bug, and I agree that it's the same problem as already given here.

I don't know which of the proposed solutions should be used to fix this bug, but it desperately needs to be fixed.

rd232 wrote:

@Alexandre Emsenhuber:

Also a problem wouldbe parameter substitution: {{Some template|link=[[Some page]]}} or {{Link|page=Some page}} where Template:Link contains something like
"[[{{{page}}}]]", for those two cases, should the link be considered as comming
from the page, or the template? (first example might be easier than the second

one)

Those cases should clearly be considered as coming from the page calling Template:Link, since "some page" is linking data stored in the calling page, not the template.

(In reply to comment #8)

@Alexandre Emsenhuber:

---

Also a problem wouldbe parameter substitution: {{Some template|link=[[Some
page]]}} or {{Link|page=Some page}} where Template:Link contains something like
"[[{{{page}}}]]", for those two cases, should the link be considered as comming
from the page, or the template? (first example might be easier than the second

one)

Those cases should clearly be considered as coming from the page calling
Template:Link, since "some page" is linking data stored in the calling page,
not the template.

Yeah, that's easy to say for a human. But a piece of software is likely not going to able to distiguish that.

for example:

Template:Foo contains:

This is a [[department]] [[portal]].

{{#ifeq: {{{type|normal}}} | special | Welcome to the [[special]]
department. | Welcome to the department page of the [[normal people]].}}

Department:Special contains:

{{Foo|type=special}}
Blablablabla.
[[Category:Departments|Special]]

Does the link to [[special]] come from the template or from the article ?
The answer to this doesn't matter and likely not obvious, but untill the parser has been rewritten and world piece has been made I dont think such advanced extractions can be made from a simple page with templates.

That doesn't mean we can't do any extractions at all. A simple way to approach this bug's goal would be to compare links generated from {{Foo}} to whatever the template call was (eg. {{Foo}}), and whatever was added consider that caused by the article, and the rest to be coming from the template and thus further down the tree on WhatLinksHere.

(things to keep in mind though are constructions with <includeonly> and the like, and comparisons to {{PAGENAME}} etc.)

rd232 wrote:

@Krinkle

I'm not sure it has to be that complicated. Any time a template wraps a parameter in [[ ]], it's link data the calling page is responsible for. Otherwise, it's link data stored in the template. Though comparing links from the template and the template call, as you suggest, may be helpful.

I'm sure there are hard cases which may be generated what might be considered categorisation errors, but this system doesn't have to be perfect in correctly including/excluding templates, as long it categorises links one way or the other. (I'm imagining there would be a switch so you could choose to look at page links, templates links, or both together.)

• Bug 17728 has been marked as a duplicate of this bug. ***

• Bug 21266 has been marked as a duplicate of this bug. ***

Sounds like bug 1392

Or like Bug 3241 - Transclusion should produce tree-like results in "What links here"

I made this comment at T5241#3246126, because I'm not entirely sure which is the interesting ticket: I think Special:Search is sufficiently powerful at this time to take care of curating "manual" links--reference this search using linksto and insource to isolate links to the article on Issik Qaghan.

We won't be able to get to this any time soon, based on our quarterly and annual goals; moving to later on our board.

I think this will require digging into the parser. What is stored in mysql, and elasticsearch, is based on what the wikitext parser says was used in the page. Currently the wikitext parser says that page was linked so that's what gets stored in mysql and indexed into elasticsearch.

In some cases it could be desirable to link back, so if this is going to be implemented it should be in a manner that is editor controlled as to the link back target.

https://en.wikipedia.org/wiki/User:PrimeHunter/Source_links.js seems to be a workaround.

@PrimeHunter has also made https://en.wikipedia.org/wiki/Template:Source_links

+1 to implementing this functionality.

Example d'jour, https://en.wikipedia.org/w/index.php?title=Special:WhatLinksHere/Vinmont_Veteran_Park&hidetrans=1 shows me a ton of inbound links, all of which are because my page is listed in https://en.wikipedia.org/wiki/Template:Protected_areas_of_New_York_City, and that's transcluded all over the place. What I really want to see is the pages that actually link directly here, not because of a transcluded navigation template.

Putting a +1 here because someone on zh.wp telegram group asked for a way to not list navbox-introduced links. I agree that these links don't make sense to be counted and add unnecessary noise to "most wanted" and similar stats.

I would just like to mention that being able to hide navbox links or just see the template would be very useful for us mobile editors who can't see navboxes.

Like, please. It makes 'What links here' basically useless for any article linked in a template.