Archive 5 ← Archive 10 Archive 11 Archive 12 Archive 13 Archive 14 Archive 15

Suggested updates template

Getting away from deletion debates for a moment I wanted to draw attention to Evad37's very useful template ((Portal suggestions)). It goes on the talk page of a portal, and based on pattern searching it suggests recent items that could be added to the portal: In the news, Did you know, high-quality pictures on Commons, and articles promoted to FA this year. You decide which suggested items do in fact belong on the portal. At present it is transcluded on only around 20 portals and I think it could be valuable on many more. Regularly checked, it could assist in keeping portals up to date and quality checked, and is an example of the blend of automation and human discretion I think we should be looking out for across the project: Bhunacat10 (talk), 17:17, 6 May 2019 (UTC)

@Bhunacat10: It was only posted on a few portals initially as a pilot to work out any issues before widespread use. If people think it's a good idea, we could scale up the testing, and maybe eventually just make it a submodule of ((WikiProject Portals)) to display on every portal talk page, like we do with ((Portal maintenance status)). @Evad37: What do you think of this approach? — AfroThundr (u · t · c) 04:38, 8 May 2019 (UTC)
I think it's an excellent idea, could be really helpful. WaggersTALK 10:24, 8 May 2019 (UTC)
I've tried it on a couple of my portals and not found it all that useful, to be honest. It also seems to make the page load extremely slowly. Espresso Addict (talk) 04:30, 9 May 2019 (UTC)
I've just created a page in my userspace User:Waggers/portal_suggestions, with the template for a few of the portals I contribute to. If adding the template to portal talk pages causes performance issues maybe this is a better approach.
What would be really cool is if any portal I add myself to as a contributor using ((Portal maintenance status)) could be added to that page automatically, but I suspect that's a bit of a tall order (and a bit lazy on my part, but laziness is the root of many good innovations!) WaggersTALK 09:14, 10 May 2019 (UTC)

Portals as vulnerability

See Wikipedia:Miscellany for deletion/Portal:Donald Trump, where the issue of possible POV-pushing is raised. --BrownHairedGirl (talk) • (contribs) 10:09, 4 May 2019 (UTC)

I agree this is an issue with portals related to any living person. UnitedStatesian (talk) 12:54, 4 May 2019 (UTC)
Thanks, BrownHairedGirl. I was coming here to advertise this as meriting wider discussion. @UnitedStatesian: I don't think it necessarily relates to all living people; only a handful of political figures at present are as polarising as Mr Trump. Espresso Addict (talk) 13:04, 4 May 2019 (UTC)
Certainly portal space needs to adhere to WP:BLP and WP:NPOV so we need to think about how we make sure that happens, just as we do everywhere else on Wikipedia. Do we need a tag or category for portals on potentially contentious subjects? (Perhaps one already exists?) WaggersTALK 10:35, 8 May 2019 (UTC)
Add a Sources section at the end of main biographical portal pages and use citations on the various subpages. The citations in the content that is transcluded from the subpages on the main portal page appear using the ((reflist)) template on the main portal page. Problem solved. North America1000 10:45, 8 May 2019 (UTC)
For this you'd have to ensure that the leads of all transcluded articles adhere to MOS:LEADCITE: Any statements about living persons that are challenged or likely to be challenged must have an inline citation every time they are mentioned, including within the lead: Bhunacat10 (talk), 11:55, 8 May 2019 (UTC)

Maybe I should have explained the problem in more detail.

The four POV issues raised there are:

  1. The vulnerability of the largely-unwatched subpages which contain content forks of the ledes of selected articles
  2. The stripping of references from excerpts, contrary to WP:BLP
  3. The risk of POV pushing in article selection
  4. The use in portals of recognized content lists such as WP:WikiProject Donald Trump/Recognized content raises POV issues, because those lists are are not drawn up with NPOV in mind. They record the progress made by editors in improving articles, but the choice of articles reflect editors interests and energies rather than NPOV policy, and the result may not reflect a balanced view of the topic — WP:WikiProject Donald Trump/Recognized content certainly doesn't meet NPOV, giving undue prominence to Impeachment March and Insane Clown President.

These issues apply to all portals, whatever the topic. However, the consequences are most severe where the portal topic is a BLP, and the risk probably greatest when the topic is polarising. --BrownHairedGirl (talk) • (contribs) 13:24, 8 May 2019 (UTC)

All valid issues; let's try to work on solutions.
(1) can be remedied by using transclusion instead of content forks, and we've been making steady progress on converting old-style portals to use transclusion templates. That work needs to continue, but this of course exacerbates issue (2).
(2) WP:POG says "It is common practice not to include references in portals. As on the Main Page, readers should be able to verify the portal content by following a prominent link to a relevant article, and checking the references there."
Personally I've never been entirely comfortable with that, even before WP:BLP came along - I don't think it's reasonable to assume that visitors will click through to see the full article (complete with references) so displaying references on the portal page itself makes sense. To do that, as User:Bhunacat10 says above, we'll need to ensure that the leads of transcluded articles adhere to MOS:LEADCITE and that the transclusion templates don't strip the references.
I think we need WP:POG to be changed to state that references should be included inline on the portal page, especially for BLPs. Then obviously there'll be quite a bit of work to do to bring portals up to speed as nearly all of them contain selected biographies - and it'll involve some changes in the transclusion template/module code presumably.
(3) This is indeed a risk but a manageable one, in much the same way as we manage it for articles. One solution is to use navigation templates or recognised content lists as the basis for article selection instead of manual curation, but of course that brings us on to (4).
(4) This is an issue for navigation templates too, because of WP:EXISTING, so not unique to portals, but that doesn't mean we shouldn't think about how to address it. Of course the best solution is to have a wide, balanced range of high quality content in article space on every subject, but that's not going to happen. Ultimately we do have to accept that Wikipedia is a work in progress and it's ok for portals to reflect that, while keeping as much of a neutral POV as we can. WaggersTALK 14:28, 8 May 2019 (UTC)
Good that we can find agreement!
  1. I think it's time to explicitly deprecate the use of subpages to store any text other than links. Can we agree to amend the guidelines accordingly?
  2. Is the solution simply to a) amend the Lua modules to stop them stripping refs, and b) add a refs section at the bottom of each portal?
  3. The risk is much higher in portals than in navboxes, because navboxes display all the links on the face of the navbox. In most portals, the list of selected topics is not displayed to readers, and readers become aware of the inclusion only when the selection rotates to that position.
    take e.g. the hypothetical Portal:Angela Merkel. If a miscreant decides to add to the selection the article on the (hypothetical) notable hate-book Why all Germans are really Nazis who should be exterminated, they could so a) creating a new subpage, if this is still a multi-subpage portal; b) adding it to page which lists the selection, as e.g. in Portal:Geophysics/Selected geophysicists; c) for portals which have the list encoded in the main portal page, add it to there (to see the point whee t could be added, see e.g. my edit[1] typo Portal:Donald Trump).
    In each of those 3 examples above, there is no overt notification to the reader, unless and until that article is selected. It seems to me that there is a fairly simple solution to all of this: require all portals to display on the face of the portal a list of all the article pages in each selection. This would be a huge boost to usability as well as to transparency, so it looks to me like a win-win.
  4. We disagree here. WP:EXISTING is a much broader issue, and the question of navboxes is much less severe issue because navboxes are an NPOV selection of existing articles. Recognized content lists have no NPOV basis at all, since they are internal tracking pages, and the Trump example shows how they make a very POV set which would not be tolerated in a navbox. Those lists should be retained for internal use only. --BrownHairedGirl (talk) • (contribs)
On a technical note, it's not as simple as amend the Lua modules to stop them stripping refs - references may be defined in the article body and inserted into the lead using <ref name=whatever />, which would break when transcluded. Similarly harvnb style refs would break by transcluding only the short footnote and not the full citation. Also, any portals transcluding multiple excerpts could have conflicts if different refs have the same value for name=. Such problems could potentially be worked out in Lua code, but it wouldn't be quick and simple. - Evad37 [talk] 00:56, 9 May 2019 (UTC)
OK, beginning to see red here. No, we don't all agree that subpages are unneeded. The current templates are ... how can I put this without being offensive... buggy, mediocre, and have some existing serious issues that I have banged on about to no avail in several forums for nearly a year without any sign of any improvement.
Where on earth are you planning to put the references? Even if the Lua problems could be fixed (no small issue), the design would need to include sources somewhere. Also the leads of the articles, if using the extract model, would need to include the sources too, which many projects don't agree with, where the material is cited in the body. Espresso Addict (talk) 04:43, 9 May 2019 (UTC)
Not to mention why are we bothering with this question at all when practically all portals on BLPs have been deleted for one reason or another over the past month or so, and I think practically all of those that are not already deleted are currently up at MfD. Espresso Addict (talk) 04:57, 9 May 2019 (UTC)
Here's another set of possible approaches then, all of which are far from ideal but might be worth considering. Bearing in mind that the only references we really need to worry about are those in BLPs that support "challenged or likely to be challenged" statements, we're actually talking about quite a small scale problem and managing it manually on a case-by-case basis may be more efficient than trying to get a technical solution in place. So if a BLP lead contains such a statement, we could either
  1. exclude it, either by using the paragraphs parameter in the transclusion template to exclude the containing paragraph or (even less ideal) not including that biography in the selected articles
  2. use the subpage method to include a copy of the text from the lead including the full reference (I know, we don't like this!). The references section could either go at the bottom of that subpage or at the bottom of the portal page itself. Obviously this approach would need regular reviewing/maintenance of that text
  3. the status quo, relying on visitors clicking through to the article to view the references. Considering the quote above is from WP:MOS (which only applies to articles) not from the actual WP:BLP policy (which covers the whole project) in addition the quote I provided above from WP:POG (the portal equivalent of WP:MOS), there's some justification for this approach.
To be clear, I'm not at all keen on option 3, I think it goes against the spirit of the policies and guidelines if not the letter, and the justification is tenuous. The other two could work for the small number of contentious statements we have to deal with. WaggersTALK 11:34, 9 May 2019 (UTC)
The main page, which quacks like a portal but is highly visible, regularly features excerpts from BLP leads without repeating their references. How many problems does that cause, and what steps have been taken to solve them? Certes (talk) 11:46, 9 May 2019 (UTC)
Portals should be neutral and certainly not be used for WP:POV-pushing. That said, let's keep this in perspective. Firstly, portals aren't great places to push POVs, which reduces the risk. Secondly, the number of portals where this has happened is low. Thirdly, hard cases make bad law so we don't want to create blanket rules for all portals because of a handful that editors might abuse - we can police that. I'd be against references in portals, but for guidelines e.g. that extracts of articles should either be a) a transclusion or (preferably IMHO) b) a faithful overview. HTH. Bermicourt (talk) 12:32, 9 May 2019 (UTC)
On reflection, I agree; there's a real risk of using a sledgehammer to crack a nut here; and if it's good enough for the main page, it's good enough for a portal. WaggersTALK 12:39, 9 May 2019 (UTC)
@Certes and Waggers: The comparison with the featured article on the Wikipedia Main Page is a very poor one. WP:Today's featured article has 4 active co-ordinators who scrutinise the pages , and with an average of over 16.6 million daily pageviews, any problems on the Main Page are spotted and fixed very quickly.
By contrast, there are no coordinators to monitor the sprawling sets of ten of thousands of portal subpages, and not even any list or category of them. So there is no option to use for example related changes, as can be done here for Wikipedia:Today's featured article/2019.
Suppose you were notified that some time in the past few months, someone unidentified had done a breaching experiment to one of the sub-pages. Something similar would be easily tracked if done to a TFA page, but how would you go about finding the michief in portal sub-pages? --BrownHairedGirl (talk) • (contribs) 17:39, 12 May 2019 (UTC)
I'm against the notion of the deprecation of portal WP:SUBPAGES. Non-automated portals are reliant upon the subpages, and the subpage system works just fine as-is. The subpage system has been a core part of portals from the start. Also, if a user want's to see a list of selected content, they can select, for example, a "More selected articles..." link on the main portal page and then see the whole batch. It's also too drastic of a change; too much too fast, and trying to change everything around all at once regarding portals in too rapid of a manner will likely lead to later unforseen problems. North America1000 04:41, 11 May 2019 (UTC)
I totally agree - subpages can be a vital part of a portal and address several of the issues raised by portal sceptics. They should not be deprecated. That said, I think there is a balance to be struck. Too many subpages can add an excessive maintenance burden. Bermicourt (talk) 08:32, 11 May 2019 (UTC)

Some interesting points above, but the fact remains that these subpages with excerpts of articles are WP:REDUNDANTFORKs. They are a long-established practice, but a bad practice. As @Northamerica1000 says, the subpage system has been a core part of portals from the start ... but many other long-established practices have been abolished.

I don't propose mass deletion of them, just deprecation, so that they can be phased out in an organised way. --BrownHairedGirl (talk) • (contribs) 13:27, 12 May 2019 (UTC)

I stress that those decisions were made by overwhelming consensus at two WP:CENT-advertised discussions, and that the turnout was exceptionally high. So your objections to reverting that clearly-deprecated format looks to me like a severe cases of WP:IDHT.
It's also a great pity to see that you continue the tired old game played by some portal fans of misrepresenting WP:ENDPORTALS. For the record, ENDPORTALS asked Should the system of portals be ended? This would include the deletion of all portal pages and the removal of the portal namespace ... and the close was There exists a strong consensus against deleting or even deprecating portals at this time.
So it was just a decision to reject a proposal to delete all portals, and nothing else. It was NOT a decision to keep all portals which have ever been crated. It was NOT a decision to convert all portals to automated forks. It was NOT a decision to retain content-forked sub-pages. Please stop this tedious practice of claiming that ENDPORTALS answered questions which were not asked, and were not mentioned in the close.
Yes, much of the restored content will be outdated and irrelevant and will decay over time. Where the reverted portals were abandoned as static pages, I have been nominating them for MFD. The others may be updated in various ways if editors choose to do so. --BrownHairedGirl (talk) • (contribs) 15:37, 12 May 2019 (UTC)
Obviously, many other subpages are lists etc, which are not forks. --BrownHairedGirl (talk) • (contribs) 15:55, 12 May 2019 (UTC)
Thank you, that's helpful. One positive outcome of this whole debate about portals is that it will definitely help me to sharpen my focus when improving and updating portals. Bermicourt (talk) 16:05, 12 May 2019 (UTC)

Project status update?

As The Transhumanist, for understandable reasons, appears to be no longer updating the newsletter, I think the project could do with taking stock.

Based on links provided by BrownHairedGirl there seem to be 1303 portals not currently tagged with a deletion request plus 228 portals currently at MfD. (The query seems to return in order of creation with oldest at the top, and I don't know how to sort.) So the project is roughly back where it started a year ago, at least in terms of number of portals.

Can anyone provide anything more useful? eg

WP Portals has got to start moving towards a realistic assessment of what type of portals are useful & to whom, how to maintain them, and what forms of portal & topics for portals are accepted by the community, if any, if it's not to be led by the editors at Miscellany for Deletion. Espresso Addict (talk) 03:26, 28 April 2019 (UTC)

Quick reply
  1. Glad the Petscan links were useful, but AFAIK Petscan won't sort. When I need to sort, I use its wiki output mode, save to a text file and load it into AWB.
  2. I'd also like a listing. I will write a bot request.
  3. Portals in a decent state? Probably no surprise that my answer is "very very few". I'm working on the bottom rung. i.e those which are just bloated rewraps of navboxes/lists or plain broken, but as I view nearly all of them along the way, it's v rare to find one to which I can actually say "yes, that's good". But if this portal wants to do more than triage the portals in big trouble, then it needs some actual criteria for what makes a good portal. And to avoid a rerun of 2018, those criteria need broad, RFC-based community support.
Observation: this project looks to me to be directionless. Most of what energies it has seem to e to be going into either resisting deletion of portals of at best marginal utility, and/or in restoring to an old multi-subpage manual format portals which had rotted so badly when in that shape that they were automated without protest. That;s too much energy focused on the least important pages, which is a sure path to decline.
Both seem to me to be backward-looking strategies. The community voted not to delete all portals, but has made no broad decisions on which topics it wants portals for. There is no broad consensus on what portals are actually for or how they should do it, and at least radically 3 difft models have support.
If those two big questions are answered with clear community support, then portals may have a future. Otherwise, sooner or later somebody will do an ENDPORTALS2 which may succeed. --BrownHairedGirl (talk) • (contribs) 03:55, 28 April 2019 (UTC)
(Request for clarification: BrownHairedGirl, which two big questions? I see the big question as What is the purpose of portals, and all the rest as relatively small questions relating to how to achieve the big answer.) · · · Peter Southwood (talk): 06:35, 13 May 2019 (UTC)
@Peter Southwood: Yes, the big question is purpose. The secondary question is which topics need to be enhanced for that purpose. Then how to do it.
What, which, and how. Probably in that order. --BrownHairedGirl (talk) • (contribs) 12:57, 13 May 2019 (UTC)
Thanks for offering to write a bot request, BrownHairedGirl; I have no idea how one goes about that.
I come at this from the other end of the quality scale, having mainly interacted with portal peer review and the featured portal process. When the latter was active, there were a handful of really good portals, plus perhaps 50–100 that were decent/mediocre and a substantial minority that should have been de-featured. But the process was last fully active back in ~March 2016, and all the former featured portals were of the static lead extract form. That was what Transhumanist was trying to fix, but the community has rejected most, if not all, of those methods. From the perspective of a portal-phile, I think we need to be more proactive in presenting the community with at least some good-quality, attractive, updated portals, plus a process to move a higher proportion of portals towards this state, and say, hey, look how great they are!
I don't think another portals RfC of any description will be of benefit to this Wikiproject at this time, and in particular I don't think asking the Usual Suspects (mainly long-term editors & admins) at an RfC is going to shed any light on what a reader might want from a portal. None of us here are (primarily) Wikipedia readers. I asked my social media network earlier this year about the main page, and not one person ever visited it, and more than one person expressed surprise that it so much as existed.
The project looks directionless to me, too, hence the unsubtle attempt to push it along, but I don't think I've ever been a member. Espresso Addict (talk) 05:02, 28 April 2019 (UTC)
PetScan's "Output" tab has limited sorting ability: by title, size, date or incoming link count. For example, here is the list of non-MfD portals sorted by title.
PetScan's "Page Properties" tab can filter by last edit before or after a given time. However, it can't filter out non-content edits such adding a tag.
Two open questions are whether every portal which was generated automatically is necessarily a bad portal which must be deleted, and whether whether every portal which is not maintained manually is necessarily a bad portal which must be deleted. Perhaps they are valid reasons for considering a portal for MfD but not valid rationales for deletion.
Automated portals could benefit from some semi-automated one-off improvements, such as customising the DYK and ITN lists. For example, I edited Portal:Bears to filter out people who "bear arms" etc. It may be worth resuming those tasks once it becomes clear which portals are likely to survive. Certes (talk) 11:24, 28 April 2019 (UTC)
@Certes, I agree that non all automated portals should be deleted. If I did think they should all go, I'd have saved the community a lot of time by 4 weeks just MFDing in one batch all of the ~4,500 portal which then transcluded Module:Excerpt slideshow. Instead I have been MFDing only those which a) replicate the set of links on one single page, and have no viable pre-automated version; or b) portals with too narrow a scope. That has emeant several hundred extra MFDs.
Thanks for the comments on Petscan. I have used it squillions of times, but had never found the handy output-sorting, even tho I use that tab lots to set wiki output. Duk.
I tried Petscan's "filter by last edit" functionality, but never got it work. The results on my test set were junk. --BrownHairedGirl (talk) • (contribs) 14:19, 28 April 2019 (UTC)
@BrownHairedGirl: From experimenting, I'm convinced that PetScan has an unhelpful and undocumented feature. After (or exactly) does what it says. In particular, it modifies its behaviour correctly if you tick Only pages created during the above time window (overrides "last revision"). However, Before (or exactly) always seems to select by creation date, i.e. it acts as if the box were ticked even when it's not, and checks the earliest edit rather than the last. I don't think we can work around it – hacking the URL to add &only_new=off doesn't help – but it sounds like a fairly easy code fix. I'll report it if you agree. Certes (talk) 15:28, 28 April 2019 (UTC)
Thanks, @Certes, but I'm a bit too busy now to set up my own tests again. However, if you can show me some testcases on small sets to illustrate the behaviour, then I can see whether I agree. Alternatively, just file a bug report yourself if you are happy to stand over your findings. You've done the spadework, so you should get the credit! --BrownHairedGirl (talk) • (contribs) 15:40, 28 April 2019 (UTC)
Reported. Certes (talk) 16:11, 28 April 2019 (UTC)
Thanks, @Certes. --BrownHairedGirl (talk) • (contribs) 17:03, 28 April 2019 (UTC)

Automated portals based on outlines

There is currently an ongoing discussion as to whether portals based on outlines, with no manual version worth saving, should be deleted on sight. As I said above, while I see consensus for navbox(es), and hoovering links from the head article is just a terrible idea, I'm not convinced that there's clear consensus at present regarding outlines.

There is some discussion of this at Wikipedia:Miscellany for deletion/Portal:Jordan (2nd nomination) and I'm also aware of Wikipedia:Miscellany for deletion/Portal:Classical architecture. Espresso Addict (talk) 03:51, 27 April 2019 (UTC)

@Espresso Addict, that's an outrageous act of WP:CANVASSing.
It is blatantly partisan, and misrepresents the nomination: there is no proposal to delete on sight. (That would be speedy deletion, which is not proposed).
The community consensus on forking is very clear: see WP:Content forking#Acceptable_types_of_forking, which provides no exceptions for portals for Outlines. --BrownHairedGirl (talk) • (contribs) 15:35, 27 April 2019 (UTC)
Calm down guys. BrownHairedGirl is right but sounding a bit like RedHairedGirl lol. :) Bermicourt (talk) 18:51, 27 April 2019 (UTC)
Sorry. I was very angry last night. Espresso Addict (talk) 05:09, 28 April 2019 (UTC)
TL;DR: Until we have one or more consensus purposes for portals, most of these discussions are irrelevant. When we have purpose, portals could use any method that legitimately achieves that purpose.
  1. The community consensus on forking is clear on some points - those that are mentioned in WP:Content forking. It is not clear on the points that are not mentioned in WP:Content forking, or if it is, it is not easily accessible, and should be mentioned in WP:Content forking.
  2. There are some things that are relevant to portals that are not mentioned in WP:Content forking at all. One of these is the word portal. It is not clear whether WP:Content forking applies to portals at all, of if it does, to what extent it applies. This is based on the assumption that portals are not articles because they are in a separate project space. It would be reasonable to assume that if portals are articles they would be in mainspace.
  3. It is reasonably clear that if portal subpages with independent content on the same topic as mainspace articles were in mainspace they would fall foul of WP:Content forking. However, these subpages have existed as a class for a long time in manually constructed portals and one of the biggest objections to portals is that these subpages get out of date and inaccurate, and require considerable detail maintenance, not that they exist. This suggests that there was a consensus that WP:REDUNDANTFORK did not apply to portals, but consensus can change, and may have done so in this case, so this point should be established and clarified, preferably by a neutral person or group. Otherwise it must be clarified by RfC.
  4. It is claimed that an automatically generated portal based on a single navigation template is a redundant fork of that template, and that adequate consensus for this opinion has been generated at deletion discussions. If this is so, this should be mentioned in WP:REDUNDANTFORK. This has not yet been done. Once it has been done and settled there, this can be linked to from portal guidelines. However the confusion as to the fundamental purpose of portals remains. They are in a distinct namespace which implies that they have a different purpose to articles, and that some rules that apply to portals may be different to those that apply to articles. If there are no differences to the rules, then portals are articles and there is no need for a separate space.
  5. The question arises of whether an automatically generated portal based on more than one navigation template, or parts of a single navigation template would be WP:REDUNDANTFORKs too, and by what reasoning. Similar reasoning could be applied to portals based on any external listing of articles, potentially requiring that the portal must have a sufficiently unique internal list of articles, either directly or implied by the actual articles displayed in the portal. This line of reasoning may or may not be legitimate depending on what the purposes and functions of portals are deemed by the community to be. Notice that I intentionally split purposes and functions, they are not necessarily the same thing. Ideally they would be, but in reality there may not even be an overlap. Without community accepted purposes, one cannot assess whether the functions match the purposes, and there is no future for portals and portal namespace except as a historical appendix. · · · Peter Southwood (talk): 06:19, 13 May 2019 (UTC)

@Peter Southwood, the community consensus on navbox-forked portals is exceptionally clear. Whether or not it is recorded in any particular document is just a matter of the organisation of policy/guideline documents.

I see no distinction of principle in the creation of portals forked off outline pages or list pages, and I see that nobody else raised one. If someone thinks that there is such a distinction, then please set it out.

As to multiple navboxes, the same principle of redundancy applies. Take the example of the imaginary city: XYZcity. It has a large, comprehensive navbox ((XYZcity)), with over 150 topics arranged in several group and sub-groups. A Portal:XYZcity which draws its selected article lists solely from that navbox would clearly be a bloated redundant fork of the head article, as agreed at the mass deletions one, and two.

But suppose instead that the same XYZcity had four smaller navboxes: ((XYZcity)) for geography, plus ((History of XYZcity)), ((Politics of XYZcity)), and ((Culture in XYZcity)). If Portal:XYZcity which draws its selected article lists solely from those 4 navboxes, it would duplicate the head article in exactly the same way.

In other words, what matters is not the number of navboxes, but whether the set of articles is already available on another page.

Please note that per WP:PORTAL, "Portals serve as enhanced 'Main Pages' for specific broad subjects". In other words, the portal has to add value to the head article.

These discussions would be much more fruitful if editors would keep a clear focus on that principle. Instead of arguing about the scope of particular documents, it would be far more productive to just ask whether a portal actually adds value.

The one point where I actually agree with Peter is that the underlying problem is lack of clarity about what is the purpose and function of portals.

The overall purpose is actually fairly clear — be an enhanced head article — but there is no community consensus on what portal can do to achieve that. I have repeatedly noted at MFD (and TTH picked it up in his latest "newsletter") that the two major functions of most portals have been showcasing articles and providing picture galleries, but that those two functions are largely redundant to two newish features of Wikipedia:

  1. mouseover: for ordinary readers who are not logged in, mouseover on any of the linked list items shows you the picture and the start of the lead. So the preview-selected page-function of portals is redundant: something almost as good is available automatically on any navbox or other set of links
  2. automatic imagery galleries: for ordinary readers who are not logged in, clinking on an image brings up an image gallery of all the images on that page. It's full-screen, so it's actually better than even a click-for-next image gallery on a portal

Similar features have been available since 2015 to users of Wikipedia's Android app.

Reading WP:POG#What_content_to_include in that light, I see nothing in the "Required" or "Recommended" sections which justifies the existence of a portal. Those are all wrapper items, except for the "Selected article" and "Selected picture" items, which as above are now redundant. The only added value comes in the "Optional" section: News, DYK, Selected anniversaries. It's not mentioned, but I would add "recognised content". Does that set justify creating a portal? I am not persuaded.

My experience of analysing hundreds of portals in the last few months is the News, DYK, and Selected anniversaries functions require a lot of editorial work, both in setup and in maintenance, but that the work is v rarely done. So e.g. we have hundreds of "news" pages which are ten years out of date, many of them not even clarifying that they are out-of-date (because they just list month/day, without stating the year). This is just one part of the much wider problem which TTH set out unsuccessfully to address last year: that models adopted for portals add value only if they get a huge amount of editorial work, but that work is not being done.

TTH's remedy of automation failed, because the community consensus is that it didn't add value. His latest "newsletter" basically concludes that portals are redundant, because they features are increasingly provided elsewhere.

So instead of fretting over whether a particular policy document explicitly deprecates a particular type of redundancy, it would be far more productive to go back to first principles and ask how can a portal actually add value? --BrownHairedGirl (talk) • (contribs) 14:21, 13 May 2019 (UTC)

"Portals serve as enhanced 'Main Pages' for specific broad subjects". - What does this actually mean?
I will set out a few questions here, if we can answer them we may be getting closer to an answer for the purpose of portals.
  1. What is a 'Main Page' for a broad subject?
  2. In what ways can this 'Main Page' be enhanced?
  3. How is this useful?
  4. To whom is it intended to be useful?
  5. How does one identify a subject sufficiently broad?
· · · Peter Southwood (talk): 18:14, 13 May 2019 (UTC)
Quick replies:
  1. In nearly all case, the main page for a broad subject is an eponymous article.
  2. how to enhance? As above, most of the content of current portals doesn't enhance.
  3. How is this useful? It mostly isn't. The portal concept has died in most other places on the internet, because good search and massively interlinked pages save made it redundant. So it's unsurprising that en.wp portals have such abysmal viewing figures.
  4. To whom is it intended to be useful? For now, sadly, primarily to the creators of portals. Readers are clearly uninterested, and most of the defences of portals have been along the lines of "we've always done it this way".
  5. My criteria are a combination of all of: covers enough broad sub-topics to attract interest beyond a single entity; wide rage of high-quality articles within scope, preferably at least 100 of C-class or higher; has a decent number of active editors creating or maintaining high-quality content. --BrownHairedGirl (talk) • (contribs) 19:12, 13 May 2019 (UTC)

Minor question

Can anyone determine what code is responsible for alphabetizing Portal:Nautical in the "S"s within Category:All portals? I don't see any DEFAULTSORT magic word. UnitedStatesian (talk)

The portal transcludes all or part of Gary Snyder and picks up its DEFAULTSORT of "Snyder, Gary". Module:Excerpt strips out DEFAULTSORT to prevent this sort of problem, but this is an older portal which does not use that code. Although Snyder is mentioned in the "May in nautical history" sidebar, I can't immediately see how the article gets transcluded. Certes (talk) 10:05, 9 May 2019 (UTC)
Redirect Portal:Nautical/May/08/Selected article may be relevant, and may mean that this problem only happens once a year. Certes (talk) 10:09, 9 May 2019 (UTC)
Thanks @Certes:, very helpful as always. I converted it from a redirect to a lead-only transclusion. UnitedStatesian (talk) 13:48, 14 May 2019 (UTC)
@UnitedStatesian: There must be plenty of similar redirects. Unfortunately, most of them have been blanked, including all the others for this portal such as Portal:Nautical/February/03/Selected article. Occasionally an article gains an image but the uploader is slow to add licensing information. This prompts an enthusiastic editor to blank the page to prevent a file which might possibly be non-free from appearing in a portal, and of course they don't bother to undo that damage when a licence appears or the image is removed. Module:Excerpt removes non-free images - a substantial part of its code and run time are spent examining file licences - so transclusions could be used to repair past blankings and avoid future ones. If any such portals survive then it might be worth checking and improving them in that way. Certes (talk) 14:36, 14 May 2019 (UTC)

Collaboration – Country and major geographical Portals

Most portals about countries of the world and about major geographical topics pass Wikipedia:Portal/Guidelines to qualify for standalone portals. Some of these portals need updating, checking for errors and correcting them when found, expansion with more content, and reversion to pre-automated versions when feasible. I have started a list below of portals I have recently worked on; feel free to further improve them if interested, and feel free to add to this list (along with your signature). When people work together, vast improvements are more easily realized.

The following topics meet WP:POG to qualify for standalone portals, but were recently deleted at MfD. I had reverted both to non-automated versions and was planning on expanding them, but both were nominated for deletion shortly thereafter, hours later. If anyone's interested in creating a new portal from scratch, below are two opportunities.

– North America1000 02:33, 11 May 2019 (UTC)

Northamerica1000, Thanks for this. I've done a bit of work on Portal:Jordan but there's more to be done. Very happy to help out with the others too.
More generally I think we need to move away from the model of individual named editors looking after individual portals, and instead have teams of editors looking after groups of portals. Maybe we should think about setting up some task forces. WaggersTALK 08:57, 14 May 2019 (UTC)
I don't agree with the first part. There's nothing wrong with individual editors looking after portals. I do agree, though, that - above the individuals - it would be to have a team of editors overseeing a group of portals. But IMHO rather than adding another layer, that role should be fulfilled by the WikiProject editors of the portals that fall within their purview. Those teams already exist and portal creation, oversight and maintenance is an obvious WikiProject function. So there would be up to 3 levels: an individual operating as part of a WikiProject team, guided by the portal project team. Bermicourt (talk) 12:21, 14 May 2019 (UTC)
That assumes that the editors in that WikiProject have any interest in maintaining a portal. As far as I can see, that interest is very rare. --BrownHairedGirl (talk) • (contribs) 19:24, 14 May 2019 (UTC)
That may be the status quo, but it doesn't need to stay that way. I think if we encourage project editors to accept responsibility for portals in their area of interest, one of two things will happen. Either they will embrace their portals, in which case, problem solved. Or they don't, in which case the portal project editors take control and decide whether to keep and oversee the portal themselves or put it up for deletion. What I think is highly unlikely is that we could somehow find a random bunch of editors to oversee multiple portals. Bermicourt (talk) 19:53, 14 May 2019 (UTC)
Bermicourt, the problem is that this basically amounts to asking WikiProjects to take responsibility for:
  1. Portals which in most cases were created without discussion at the project
  2. Portals which use a very complex and unfamiliar set of code, so a large investment of time is needed just to get started
  3. Portals which in most cases have a collection of outdated content forks
  4. Portals which hardly anyone views. With most portals, the head article is viewed between 100 and 2,000 times more often.
  5. Portals which a large chunk of the community thinks should not exist at all, including most of the more experienced Wikipedians (WP:ENDPORTALS swung towards rejection only after TTH tagged all the portals, bringing in the less experienced editors who don't watch VPP or WP:CENT)
  6. Portals whose ability to add value has nearly vanished, thanks to technical changes
That's a hard sell. See e.g. the discussion at WP:Miscellany for deletion/Portal:Ireland, where there was not a single Irish editor interested in maintaining the portal or advocating keeping it. --BrownHairedGirl (talk) • (contribs) 14:10, 15 May 2019 (UTC)
  • @BrownHairedGirl:. Yes it is a hard sell. But not as hard as finding a random group of editors to look after portals in a subject area they're not really interested in, which seems to be the suggestion. At least the projects have an interest in their areas. That said, again we have an education process. If the project editors don't understand how portals can be used by them to improve coverage, they'll show little interest and you'll be arguing "well there you go, let's delete them". It comes down to a point which I've made before, but nobody's listening: portals are very useful as a tool to improve project coverage, balance and quality if set up and used well. They also appear to have more reader-facing value if linked from categories and not from articles (except a couple of main topic articles). That needs time to set up and test, but German Wikipedia shows that it works (they also, you'll be delighted to know, have far fewer portals anyway). Bermicourt (talk) 18:17, 15 May 2019 (UTC)
  • @Bermicourt, I don't mind at all where the maintainers come from. I just don't want anyone to expect that project notification will be the magic door-opening.
I take your point that portals are very useful as a tool to improve project coverage, balance and quality if set up and used well. But that is a very big "IF" ... because as we both know, v v few en.wp portals are set up in any such way. We'll only get there if about 90% of the existing portals are rebuilt. --BrownHairedGirl (talk) • (contribs) 18:27, 15 May 2019 (UTC)
That's obviously a lot of work. But I'm sorry to say that it still looks to me like sticking plaster on the core problem: too many labour-intensive portals, too few maintainers, and high-maintenance portals which add little value to readers.
  1. Only one the three has a maintainer (@UnitedStatesian on Portal:Nigeria). The others will rot, just as they have rotted in the past
  2. The model used continues the old practice of multiple subpages with content forks of the led of the the articles. That creates a maintenance headache for the future, with each of those sub-pages needing to be updated as the head article develops. If you really do want to maintain a curated list of topics, why not use ((Transclude random excerpt)) with the topics listed as parameters? That eliminates the sub-pages with all their maintenance problems.
  3. I examined Portal:Dominican Republic, where the list of topics chosen is 80% redundant. The 9 topics are: Dominican peso, Hospital San Nicolás de Bari, Culture of the Dominican Republic, Economy of the Dominican Republic, Santo Domingo, Danilo Medina, People of the Dominican Republic, Pedro Santana, Leonel Fernández ... all except the two italicised ones are already linked in the navboxes at the bottom of the page. That means that a not=logged-in reader (i.e the ordinary reader for whom Wikpedia is written) gets an automatic preview of any of them simply by mouseover on the link. You you can test this without logging out by right-clicking on Portal:Dominican Republic, and the select "open in private window" (in Firefox) or "open in incognito window" (Chrome). Similar features have been available since 2015 to users of Wikipedia's Android app.
  4. The addition of navboxes to the portals was a good idea, but e.g. all the navboxes in Portal:Dominican Republic/Topics are eligible for addition to the head article, so I added them there. That means that readers of the head article get a built-in preview of each of the 385 topics listed in those nacvobes without needing to leave the head article. Try it yourself, by opening a private/incognito view of the head article Dominican Republic.
  5. The high-maintenance slideshow in the portal contains 9 pictures. But the built-in slideshow on the head article contains 60 pictures, complete with caption. Any not-logged-in-reader can launch the slideshow just by clicking on any image, but here's a direct link which works even when you are logged in: Dominican Republic#/media/File:Dominican_Republic_(orthographic_projection).svg. Note how it's full-screen, so as well as much bigger scope, it's a massively better display than the tiny image box on the portal.
So the result of all your work is little benefit to the reader, and nightmare to maintain in the future. It raises these portals just above the abandoned-and-rotted state they were in before, but still leaves them pointless for the reader.
I am sure that your intentions are good, I can see you have worked hard ... but the effect of this is just to make the case that the portals project is still clinging to a ten-year-old high-maintenance model which has been proven to be unsustainable, and now adds almost no value for readers.
The key principle of WP:PORTAL is that "Portals serve as enhanced 'Main Pages' for specific broad subjects". Even after all your hard work, current version of Portal:Dominican Republic is in no way an enhanced version of the head article; on the contrary it as massively-degraded version of the head article.
I urge you and other stop putting your energies into this polishing of a broken model, and start trying to answer the crucial question: what could portals provide which actually adds value for readers? --BrownHairedGirl (talk) • (contribs) 22:00, 16 May 2019 (UTC)
I disagree with much of this. I'll consider using Template:Transclude random excerpt, however, those that dislike portals may very well then complain that this method uses automation, which people have been frowning upon. The template also appears to omit references, which people have been complaining about in some MfD discussions, a lack of sources in portals. Furthermore, in the "Portals as vulnerability" thread above, everyone is not for deprecating portal subpages. North America1000 04:06, 17 May 2019 (UTC)
@Northamerica1000, rather than just saying that you disagree and ploughing on creating yet more redundant content forks, please can you explain what you disagree with, and why? --BrownHairedGirl (talk) • (contribs) 14:24, 17 May 2019 (UTC)
I've been editing for some time in this session, and I'm tired. I may respond more later. I'm concerned, though, because you are now the top nominator of portals for deletion at Miscellany for deletion, and while you're involved in all of that, you're discouraging editors from improving portals on this talk page. Comes across as a potential conflict of interest; unimproved portals are more susceptible to deletion. I started this thread to attempt to solicit collaboration, rather than sidetracking about other matters. North America1000 14:55, 17 May 2019 (UTC)
@Northamerica1000, I think I need to be direct. My concern is that what you are engaging is not improvement, just busywork which adds little or no value to readers while expanding a Rube Goldberg machine which creates a maintenance headache for the future.
I just looked at your recent page creations in portalspace, and I just see a swathe of content-forked subpages being created.
For example, dozens of content-forked DYK subpages for Poland (Special:PrefixIndex/Portal:Poland). Who on earth is going to watchlist all these things, let alone maintain them? What value is added by by these dozens of extra pages whose topics are chosen solely on the basis that they were new articles many years ago?
Or Portal:Nigeria/Selected article/10, a standalone page which transcludes the lead of Nigerian cuisine. This is folly on multiple levels:
  • It's good that you have started to use ((Transclude random excerpt)). But it doesn't need subpages. Just edit the portal and replace the subpage selection with ((Transclude random excerpt)), listing the topics as parameters.
  • The choice of Nigerian cuisine as a topic is daft, because it is already linked from the navbox ((Nigeria topics)), and hence available to preview via mouseover on any page which transcludes the navbox, including the portal and head article.
If you are genuinely interested in collaboration, then please take a beak from making these vast farms of redundant subpages, and take time to discuss how you could actually add value to portals. --BrownHairedGirl (talk) • (contribs) 15:35, 17 May 2019 (UTC)

Abandoned country portals

The list which @Northamerica1000 created above is only the tip of a large iceberg of abandoned portals, many of them country portals.

In the a last few days I did a lot of reverting automated portals, and I have almost cleared Category:Automated portals with article list built solely from one template. The remaining 23 pages in that category are all at MFD.

A significant number are country portals, and near the end of my work I began categorising the abandoned-with-almost-no-content portals in Category:Abandoned country portals. Some stage in the next few days I will go back and check the others.

My guess is that there are probably over 60 country portals in a long-term abysmal condition. They won't be fixed soon, and it's not fair to waste readers's time with these outdated pages which add nothing to the head article.

My preferred solution would be one mass MFD to WP:TNT the lot, and leave editors to create new, low-maintenance portals in their place as and when they want to.

The alternative is to draftify them. I have done some checking, and I am pretty sure that the plan I discussed above works: to move them without redirect to a new title with a prefix. Some minor tweaks may be needed to some subpages (such as the archives), but the core functions work without adjustment. I suggested "Incubator", so that e.g. Portal:Ruritania be moved to Portal:Incubator — Ruritania. Other editors preferred to use the letter Omega (Ω) as a prefix, e.g. Portal:Ω Incubator — Ruritania so that the draft would sort to the end of categories. I'd be fine with that too.

Any thoughts? --BrownHairedGirl (talk) • (contribs) 16:10, 15 May 2019 (UTC)

What could portals provide which actually adds value for readers?

I think this very succinct question posed above by User:BrownHairedGirl is a crucial one; there are other questions which follow and these must be discussed by a much wider audience. This Wikiproject is largely populated with users who tend to believe something affirmative in portals sufficient to deserve each of our efforts. There's some wisdom in having the initial discussion of those questions here. It is very easy to destroy something; it takes enormous personal energy to create something meaningful and useful. Many of those here made investments in effort and creativity to build and maintain portals; I'll confess defending this investment is why I generally oppose deletion of portals as a class of content on Wikipedia.

I keep coming back to the mainpage portal. If consensus were measured on this subject I believe it would hold strongly that Wikipedia needs to have an interesting landing page. I'll postpone for now fully acknowledging the enormous amount of energy many users exert in order to make the mainpage continue to exist, but it's correct to examine what the mainpage brings to the pedia as a starting place to both new and existing readers. It's pertinent to ask: What does the mainpage provide which actually adds value for readers? I have some answers to this but would rather hear what others think. BusterD (talk) 22:40, 16 May 2019 (UTC)

The main page offers a intensively-curated set of 5 things:
  1. A sample of the best writing WP:TFA
  2. A sample of the best pictures WP:TFP
  3. A sample of recently-created or expanded articles WP:DYK (NB the hooks are merely a way of introducing the new articles. The value lies in the newness, not the hooks, so the portals retaining ancient DYKs are missing the point of the exercise)
  4. A sample of articles on topics are in the news today WP:ITN (NB this is again a showcase for articles. Portals which grab items off Wikinews are farming external links, not showcasing Wikipedia articles)_
  5. Selected anniversaries. WP:OTD
Each is the result of a highly-intensive selection process, backed up in each case by a highly-intensive scrutiny process, and supported by a huge number of watchers who check for vandalism etc.
In other words, the main page is the product of a massive editorial effort, just like the front page of the New York Times.
No topic-specific portal can possibly hope to apply anything remotely resembling the intensity of effort scrutiny applied to the main page. Portals should abandon the pretence that is possible to do it, just as most editors have abandoned last year's pretence that something like this can be automated. --BrownHairedGirl (talk) • (contribs) 23:07, 16 May 2019 (UTC)
Sorry, that brevity came across as a bit gruff, which I didn't intend. Many thanks to @BusterD for the kind words about my question, and for opening this discussion.
I think it's very important to ask this question, because as BusterD acknowledges, much of the meta-discussion divides between those who have worked on portals defending this investment, and uninvolved editors saying the result is of little value. It seems to me that this is an example of what sociologists call escalation of commitment, and businesspeople call "sunk cost fallacy". I hope that this discussion will be a step on the path towards identifying whether that sunk cost has actually produced any asset worth retaining. --BrownHairedGirl (talk) • (contribs) 23:57, 16 May 2019 (UTC)
That's only part of the problem. But there are two other key aspects that are not well understood. First, we're all working from different assumptions. Each party, each editor even, has made assumptions about what portals are for, some with an actual knowledge of using them - but usually only for one of their possible uses - and others with only perceived knowledge - they've never actually used them, hence "they're no use to anyone" = "they've been no use to me". There's been little attempt to try and understand the position of other editors, even those on the same side. Second, we're all looking at portals as they are now, not as they could, or should, be. "It's a rubbish portal - it should be deleted" is an understandable reaction to some portals and I supported the deletion of mass-created autoportals. But there's been no serious attempt to even try and understand a) that portals have several potential purposes, b) there aren't many good examples because even portal editors are often only working from one perspective and hence c) no serious attempt to collaborate to create even a few high quality examples that serve those purposes. Instead, the battle continues in a costly and unhelpful way, each side being wary of the other because everyone thinks only they are right and so we aren't able to focus on a common goal. Oh and BTW, forget the main page - it's the wrong target. We're never going to be able to make subject portals rival the main page, nor do we want to. The main page serves a unique purpose; it's never going to show balanced coverage, be an aid to article navigation or a tool for project editors. So forget it as a model. Bermicourt (talk) 02:50, 17 May 2019 (UTC)
@Bermicourt, I agree that the mainpage is an unhelpful comparator.
I agree that there has been little attempt to understand the possible purposes of portals, let alone how actually existing portals meet them. But there simply is not going to be any funding for proper research into needs, so we have to go on for now is the massview data, which shows that existing portals are almost unused.
Beyond that, all we can do is to try to draft one or more definitions of what a portal can and should do or not do, and put them to an RFC.
For example, I have been repeatedly making the point that, the image slideshows and article slideshows are now redundant to features built into the Wikimedia software. Can we at least agree that that stuff no longer adds any value? --BrownHairedGirl (talk) • (contribs) 03:07, 17 May 2019 (UTC)
Can you point me to an example of a slideshow of the sort you think are redundant? I'm not 100% sure what I have in mind is what you're describing, so can't answer your question. Bermicourt (talk) 03:48, 17 May 2019 (UTC)
@Bermicourt, see my reply above[2] to @Northamerica1000, where I analyse Portal:Dominican Republic. That uses the fossilised purge-to-refresh model rather than a slideshow, but the redundancy would be the same if it was a slideshow. --BrownHairedGirl (talk) • (contribs) 14:20, 17 May 2019 (UTC)

I like where this discussion is going. To clarify, I didn't raise the mainpage as a direct comparator. It's clear that no other pagespace on the pedia draws the resources and labor made available to this essential structure. I so stipulated. I also agree that viewing current portal investment is similar to the "sunk cost fallacy" mentioned above. Further, I agree wholeheartedly (and perhaps we all agree) that participants in these meta-portal discussions come at each other with vastly varying worldviews and basic assumptions. I also agree fully that we don't have any successful models to which we can point which demonstrate the usefulness and PURPOSE of portals. If we did we'd all be following those models.

Where I believe the mainpage portal can offer us some modelling is not in what it literally does, but in what resources it draws upon and why those resources are freely provided by the community. I'll suggest a very different list than U:BHG offers. These are the values I see the mainpage offering the wider contributor community, and ultimately the reader:

  1. The mainpage provides a starting place, a launching point, for readers. It is a simple thing to make the mainpage the homepage in your browser or bookmark it for frequent use. As a longtime editor, I use my watchlist as my starting point, but a non-editor is likely to utilize over and over the mainpage. This translates to pageviews. There's a reason why Portal:Current Events is the second most viewed portal; It's a great homepage.
  2. The mainpage supplies a platform for displaying the very best content on the pedia, regardless of subject matter. This keeps the mainpage a valuable place to use as a homepage. Again, pageviews.
  3. This need for high-quality mainpage content has evolved an organized structure, a somewhat fair and impartial system, in order to display a rotation of content, ultimately driven by pageviews; this rotation system has evolved so fully that it creates an extrinsic incentive for continual improvement of ALL CONTENT, from start-class all the way up to featured work. This extrinsic incentive, a natural human competitiveness, and a limited mainspace availability resource each have had positive effects on the normal intrinsic motivations for writers to improve their work here to see them displayed on the mainpage portal. As demonstration of that evolving system, I'd assert that the move adding GA-class content to the DYK section has made a significant difference in populating the section and provided additional extrinsic incentive to pursue GA-class improvement in pagespace.
  4. This extrinsic drive towards improvement draws the most motivated sorts of editors to participate in mainspace content creation. Low edit-count users don't as often apply their content creation to mainspace forums. A newbie editor might create an excellent start-class page, but few of those are offered to DYK; it is more likely that an experienced editor might rack up many DYKs. On the other hand, links to articles which MIGHT appear on the mainpage are sometimes rejected because the page is shepherded by a less motivated or aware editor (ex. a few weeks ago Broadway star Carol Channing died, yet the link to her article was omitted from the mainpage because there were questions about lack of sourcing).
  5. Finally, the mainpage construction culture requires a number of community endorsed maintainer/administrators in order to perform tasks which an editor without the bit cannot do. These admins, almost always short-handed, manage to keep the balls in the air. These positions are filled with some of the most active contributors to the pedia.

I could go on, but it's a bit late here and I believe I have made a case that at least in the arena of the mainpage portal, pageviews drive content creation and especially content improvement. More eyes means more edits, even if merely to fend off vandalism. Tomorrow I'd like to compare/contrast navigation tools: portals and content area templates (and why TTH might have been on the right track with trying automation). BusterD (talk) 05:11, 17 May 2019 (UTC)

@BusterD, all this comparison with the mainpage is pointless. The mainpage can do those things because vast numbers of editors collaborate to run these processes across the whole range of en.wp content. Not even the 8 broad-topic portals linked from the mainpage could hope to attract even a small fraction of those resources.
It's like comparing a small town family-run corner store with Walmart. --BrownHairedGirl (talk) • (contribs) 14:15, 17 May 2019 (UTC)

I appreciate your helping me make MY point. We ARE comparing apples to applesauce factories. That's what is central to all these MfD nominations. To find all but the most popular portals, you have to know where to look. An ordinary reader won't do that. That's a failure of Wikipedia's navigation, not a failure in performance. Without page views, a navigation tool is largely useless. If an ordinary reader can't find a page via clinking a link on the navigation section of every page (like these) or by searching in the search window, all this improvement effort is useless. So in a portal deletion discussion, pageviews are a circular argument. The portal gets no pageviews so nobody sees the need to make updates, corrections or additions. Because the portal is as invisible as a small town family run store, nobody visits it. There are many items available only in small stores which an ordinary buyer would value. Unlike the geographic world, on the internet we are not limited to the main highways. A portal is a boutique, displaying curiosities which may interest the customer. The answer shouldn't be to eliminate all stores except the Walmarts. We have lots of server space. The internet supports boutiques (see Etsy). The answer should be to mechanically make portals less difficult to see. A random stub is treated better by the pedia than a featured portal. Why not a button for random portal, or at least include portals in the random selection? Why not make portals available as content choices in the search window results? An unintentional inherent bias against the portal structure. A mistake of navigation. BusterD (talk) 17:57, 17 May 2019 (UTC)

@BusterD, the boutique analogy doesn't works. A boutique stocks a curated selection of items carefully-chosen for the target market, but most portals are abandoned relics of decades-old content forks. Far from being a boutique, most portals are more like a small version of the "house clearance" shops in the poor corners of city, where a random selection of discarded items is offered for sale.
Sadly, that includes even portals many of the few portals which are being updated. See my ongoing discussion above with NA1K about the pointlessness of their busywork at Portal:Dominican Republic.
So there's no way that the community would support a higher profile for portals. Why promote a set of pages where are mostly useless?
I am disappointed. I thought that you might be interested in a discussion about how portals could actually add value, but instead there's just more of the why-isn't-the-junk-promoted.
Why not ask what might make them worth promoting? --BrownHairedGirl (talk) • (contribs) 20:24, 17 May 2019 (UTC)
Agree with Buster. Editor interest and reader interest drive each other. We have to explode one of the big fallacies underpinning the current deletion campaign: that lots of readers are looking at and rejecting portals. Even the briefest inspection and wrinkling of nose would count as a pageview. The public are not looking at portals because when searching Wikipedia they are barely offered them or made aware of them.
I believe portals could become a major and valued feature of WP: but it's a big ask, requiring as it would:
  • Agreed purpose of portals (I'd say "to provide an attractive overview of Wikipedia coverage of a broad topic area for the benefit of readers new to Wikipedia")
  • Drafting of a best-practice guide for creating and maintaining portals
  • Agreed criteria for deletion or retention of portals, to instil some confidence in those who may wish to invest effort in improving them
  • Then, enlist the collaboration of WMF (the horror!) in amending WP search so as to bring portals to the attention of new readers (identified by cookie?) as an alternative to articles or their embedded navboxes
  • Devising a quality assurance scheme which portals would need to pass before being offered to readers.
Only thus, it seems to me, could the "no pageviews", "redundant fork" and "junk" concerns be laid to rest. Can any such project be got off the ground before current trends extinguish what remains of interest in the portal namespace?: Bhunacat10 (talk), 09:48, 18 May 2019 (UTC)
Totally agree with all of that. For some time BrownHairedGirl and I have been trying to get that sort of intelligent and productive discussion going even though we sit at different points on the spectrum between "delete all portals" and "have thousands of portals" (just to be clear we're not at the extremes!). But we haven't gained much traction. Bermicourt (talk) 18:45, 18 May 2019 (UTC)
Hmm. I think that a lot of @Bhunacat10's proposal are problematic.
After a broad purpose, the crucial thing is agreeing what a portal can do which actually adds value beyond what the head article offers. (My view: 80% of what most portals currently do is redundant)
Good practice guides, deletion criteria etc, quality assurance, can only be considered only when the crucial question of "what can portals do" is answered.
There is just no way that any sort of promotion is a runner while so many portals are in such appalling condition. For every abandoned portal I nominate, I have found at at least two which aren't quite abandoned, but are so poor that they should be draftified pending a radical overhaul. Yet this project still has no structure for grading portals, and no appetite or deleting the junk. Anyone who wants portals to actually be promoted has a very steep uphill battle to climb, because so much of portalspace has for so long been a combination of abandoned junk, automated spam, and neglected dross which is just above WP:TNT levels. If this project wants any support for promoting portals, it has to be proactive in triaging. That will be the only way to build faith that promotion won't be leading readers to waste their time on stuff like Portal:Dominican Republic.
So the focus really has to be on "what can portals do to add value". The rest is dreaming. --BrownHairedGirl (talk) • (contribs) 19:13, 18 May 2019 (UTC)

1000 views per day as the basic threshold

from Wikipedia:Miscellany_for_deletion/Portal:Donald_Trump

back alley graffiti.

If 59 is "very high for a portal", then the typical portal is pretty sad. In my perusing of portals and their page views, I think a suitable threshold is more like 1000 views per day. --SmokeyJoe (talk) 08:03, 17 May 2019 (UTC)
  • Per What links here, there are presently 214 main namespace articles that have a link to the portal. Adding more links to articles tends to increase page views, whereas lesser visible links equates to lesser page views. North America1000 09:42, 17 May 2019 (UTC)
Only 54 of the current 1214 portals exceed 100 view per day. --BrownHairedGirl (talk) • (contribs) 23:35, 18 May 2019 (UTC)
User:BrownHairedGirl, yes. That’s right. Thanks for the link, is it exactly what I am speaking to. The top portal, Main Page, is not listed, the top three are funny Portals but ok fine, the next seven are exactly the broad topic inherently NPOV suitable Portals that are great for navigation. After #10 there is a big step to the lesser portals. Keep Portal:Geography, Delete Portal:Society and all lesser portals as failing their basic reason for existence. —SmokeyJoe (talk) 23:45, 18 May 2019 (UTC)
@SmokeyJoe, I think we should continue this discussion, but take it somewhere else. Is WT:WPPORT okay with you? --BrownHairedGirl (talk) • (contribs) 23:58, 18 May 2019 (UTC)

Continue here. —SmokeyJoe (talk) 02:10, 19 May 2019 (UTC)

  • Pointless comparison. A stub is content in the early stages of development. A portal is a tool to showcase and navigate content.
It's perfectly legitimate to consider whether a tool which not is doing its job should be deleted. --BrownHairedGirl (talk) • (contribs) 13:20, 19 May 2019 (UTC)
  • Stubs are content-proper. They present a stub of content. What is a portal stub? A small bit of connectivity? It doesn’t work.
Note that many object to “permastubs”, and usually the anti-permastub argument is that they are due to an over-inclusive SNG. —SmokeyJoe (talk) 14:36, 19 May 2019 (UTC)

For now I think it is better to focus on the quality of a given portal and the suitability of its scope, rather than viewing statistics, which drags discoverability into the discussion. isaacl (talk) 05:06, 19 May 2019 (UTC)

Low pageview numbers are a navigation and discoverability problem rather than a portal problem. Given last year's RfC decision, deleting portals which fail to jump some arbitrary new hurdle would require community consensus. Certes (talk) 09:08, 19 May 2019 (UTC)
Quite. We don't delete categories and they also have low page views e.g. even Category:Germany averages only about 23 views per day. Like portals they are not in mainspace and they're a navigation aid, although unlike portals that is their only purpose. Bermicourt (talk) 11:30, 19 May 2019 (UTC)
Comparisons of portal pageviews with the category pageviews are misplaced, because the categories are a distributed system. To get meaningful figures you need to include subcats. --BrownHairedGirl (talk) • (contribs) 13:06, 19 May 2019 (UTC)
While many have criticised categories as also failures for reader navigation, with pageviews showing low levels of use, they have a purpose in organising content for editors. More importantly, categories don't fork content. —SmokeyJoe (talk) 14:39, 19 May 2019 (UTC)
Portals require little maintenance once they are established, but categories require ongoing maintenance that places demands on our diminishing manpower. How can we be sure that the effort is worthwhile? Hawkeye7 (discuss) 21:24, 20 May 2019 (UTC)
"Portals require little maintenance once they are established,[disputed] — Arthur Rubin (talk) 00:08, 21 May 2019 (UTC)
I think there is a very good basic case for a massive cull of portals. The vast majority of portals are in poor condition, offering long-outmoded and largely redundant ways of displaying ossifying content forks. Their abysmal page views reflect their very poor levels of utility, and I would vigorously oppose measures to increase their discoverability unless the fundamental quality problems have been tackled.
The current cleanup process is basically removing three classes of portals: a) redundant automated portals; b) abandoned portals with little content; c) very narrow scope portals. The portals count grew from ~1500 at the time of WP:ENDPORTALS, and grew to 5,705 at the end of automated-portal-spamming. It is now down to 1221, with 188 at MFD, leaving 1058 which extant portals not currently at RFC. The final tally of the cleanup now looks likely to leave about 900–1000 portals, of which only about 300 really deserve even a poor pass grade.
We still have the same fundamental problem which TTH identified a year ago: the resources which the community is wiling to devote to portals cannot sustain anywhere near the current number of portals built on the current model. Too little editorial energy is being spread far too thinly over too big a set, using outdated techniques. TTH's solution of total automation has been roundly rejected, and even TTH acknowledges that it is largely redundant.
Unfortunately, some of the most prolific portal contributors are still engaged in many of the worst practices of the old days, spewing out farms of content forks which they cannot hope to maintain. The old chestnut of "promote portals" gets far more attention here than attempts to star a serious analysis of what portals can do actually add value for readers, and the assinine whining of "war on portals" etc continues to be deployed against the removal of automated junk and decade-old abandoned debris. (Seriously, folks, get a grip on reality. This sort of cleanup should have been routine and ongoing. Unless and until this project can proactively initiate and sustain such monitoring and cleanup itself, it has no future).
I have not entirely given up hope that the project may change course, but the chances seem slim.
There are now basically three ways ahead:
  1. Develop a model of how portals can add value for readers whilst be maintainable, get community support at RFC, and ruthlessly mothball all portals which don't meet those new standards, bringing them back online only if and when they pass assessment. That's what I would like, but the chances of this project actually applying quality control seem to me to be vanishingly slim.
  2. Have some sort of crude criteria-based massive cull, reducing the number of live portals to the low hundreds in the hope of concentrating resources
  3. Go for something like Joe's idea, and cull everything except the portals currently on the front page, which also roughly correspond to Vital Article Level 1.
Note that even Joe's extreme proposal could be seen as modest. The front page portals have the absolute prime position on one of the world's 5th-most-visited website, but they can still manage daily pageviews only around a thousand. Anyone who thinks that portals are unused because of lack of promotion is ignoring that evidence.
So unless this project fairly rapidly snaps out of its lament-the-old-days-of-extensive-rot mode, then I think that the likely next step is something like Joe's idea, which I would modify in on crucial way: don't immediately delete the other portals, but mothball them by tagging them as historical, unlinking them from articles and categories and navboxes, whole allowing a year for the portals project to win community support at RFC for a) standards which portals must meet; b) a process for ongoing assessment of all live portals.
Joe, what do you think of that? -BrownHairedGirl (talk) • (contribs) 14:28, 19 May 2019 (UTC)
The things that have changed are: (1) The auto-Portals idea, which I liked and encouraged, proved in practice to be a massive failure, and it led to (2) a massive backlash against low quality portals.
ENDPORTALS asked for overreach. I think the top ten portals are quite ok. They would be much better if it wasn’t for the rest of them. I think an ambition to have all articles linked in few steps from a minor portal is a poor idea compared to having a few major portals that link to all articles. I think the main page is the ideal of a portal, and a portal that no reader lands on is a waste of time. If it is editors you mean to create for, then use WikiProjects. WikiProjects are mostly moribund too, not all, just most. —SmokeyJoe (talk) 14:50, 19 May 2019 (UTC)
@SmokeyJoe, i agree with most of that. ENDPORTALS was over-reach.
It seems to me that we have the makings here of a new RFC, which could be called TOPPORTALS or MOTHBALLPORTALS. It would do three things:
  1. Keep all the frontpage-linked portals (Current events, Contents, Featured content, Biography, History, Arts, Mathematics, Technology, Science, Geography, Society)
  2. Mark all other portals as historical, and unlink them from articles, categories and navboxes
  3. Allow unlimited time for the portals project to achieve community support at one or more WP:CENT-advertised WP:RFCs for 1) a set of criteria for a type and set of portals which adds value for readers, minimises content-forking and redundancy, and is sustainable with available editorial commitment; 2) an ongoing review process for assessing portals, to ensure that only portals which have recently passed a review are made available to readers
Whaddaya think? --BrownHairedGirl (talk) • (contribs) 16:06, 19 May 2019 (UTC)
Just call it ENDPORTALS2. As you say, consensus can change, and this sounds like a good way to check whether it has changed in the ways that continuing the deletion process would assume. Certes (talk) 16:14, 19 May 2019 (UTC)
Sigh. @Certes:
  1. It wouldn't be ENDPORTALS2, because it would keep the top portals and because it would explicitly allow a way for other portals to return if and when this project stops whining and actually builds consensus support for a way in which a wider set of portals actually adds value to Wikipedia, rather than just being the unread plaything of editors who like making portals.
  2. The current deletion process does not assume any change in consensus since ENDPORTALS. It applies long-standing portal guidelines to remove portals which don't fit those guidelines: a) narrow topic portals which should never have been created; b) automated spam which should never have been created; c) the abandoned (often still-born) junk which should have been removed years ago.
    It is very sad that some prominent portal fans continue to try to misrepresent the process of cleaning up the debris, and to regard it as some sort of assault on every portal. But if portal fans continue to oppose cleaning up portalspace in accordance with their own long-standing guidelines, then that is a pretty clear indication that the project has no interest in maintaining even a minimum quality threshold. Similarly, the lack of interest in trying to engage the wider community in defining what might make a portal actually add value for readers is another symptom of a project burying its head in the sand.
    So unless the project actively starts to take a grip on portal quality, I think there is a good case for the community saying "take it offline until you get your act in gear". --BrownHairedGirl (talk) • (contribs) 17:08, 19 May 2019 (UTC)
I'm afraid that sounds as if you're asking the community to stand over portal editors like a school inspector wagging his finger. And if it's okay to for portals, well much of the rest of Wikipedia needs to get its act together too. There are shockingly bad articles out there and there are thousands that are works in progress. I'm with you that we should be improving portals to a decent standard, but being threatened with OFSTED and school closure is not constructive. Editors are not going to work on improving portals with a sword of Damocles hanging over them which will kill them if they don't meet some, as yet to be agreed, standard. Please, can we end the threats and get back to cooperation that recognises the community backed the retention of portals even if they didn't set standards. Bermicourt (talk) 18:50, 19 May 2019 (UTC)
@Bermicourt, the point I am repeatedly making is that any sword of Damocles exists solely by the choice of the portals project not to define portal standards which can gain community support, and uphold them.
Comparisons with articles are misplaced, because articles are the content of the encyclopedia. If we deleted all portals and all categories instantaneously, right now, we'd still have an encyclopedia ... but without articles we'd have nothing but a shell.
Categories are a navigational tool, albeit a flawed one ... and they are regularly deleted, merged, or renamed as part of ongoing maintenance. There is even a very busy discussion area devoted just to that task: WP:CFD, where many of the editors most actively involved in creating and populating categories are also actively involved in deletion and merger.
Portals are a navigational tool and showcase. If they aren't doing that job effectively, they should be deleted, merged, or renamed as a routine part of keeping the overall system functional.
However, the portal fans are not just failing to do that maintenance. Many have even resisted that cleanup when outsiders come to do it. So the result is that the progress gets initiated by outsiders.
What you call "threats" is just the sound of a vacuum being filled. --BrownHairedGirl (talk) • (contribs) 20:19, 19 May 2019 (UTC)
I agree there needs to be co-operation: with the editors who are interested in maintaining the subject area in question. To take another example than categories, editors work on navigation boxes without fearing someone else is going to raise a consensus to delete one, because they are the interested parties that are going to reach a consensus. If portals are going to be developed more fully, they need to be built in conjunction with the same editors who maintain articles, navigation boxes, infoboxes, and so forth that are related to the topic. These are the groups that need to be engaged. isaacl (talk) 22:54, 19 May 2019 (UTC)
That leaves us with a minimum of 100 pageviews/day, which is risibly low for the WP:POG requirement that portals should be about "broad subject areas, which are likely to attract large numbers of interested readers and portal maintainers".
But even a minimum of 100 view leaves us with only 58 portals passing the threshold.
Some sort of cull that will be needed, because it is very clear that there simply are not editors working on portals to sustain anywhere near the current number of portals. Whatever basis is chosen for the cull, it needs to happen. --BrownHairedGirl (talk) • (contribs) 21:29, 22 May 2019 (UTC)
@BrownHairedGirl: I'm not against a cull, just against one that is based on page views. Your concern is with maintenance, so why not look at that directly? Perhaps something similar to the treatment of wikiprojects can be used: classified as inactive if there has been no maintenance over some period of time, deleted or archived if it hasn't been maintained in a few years. RockMagnetist(talk) 01:02, 23 May 2019 (UTC)
My comments were not based simply on pageviews. I looked at the portals, and I noted a correlation with that threshold, and an inflection point in the curve below it (Portal:Society appears below the boundary). Above Portal:Society, the portals are obvious suitable entry points for top-down browsing and insusceptible to POV issues. At Portal:Society, the very definition of the portal name becomes a debate-worthy question, it is subjective, it is not universally agreed. Below Portal:Society, there quickly comes a list of topics that I think are obvious POV/Advocacy/promotion targets, whether conscious or unconscious, and if it is not happening is it because they are effectively dead, you can get more traction writing graffiti on a laneway. Potential POV/Advocacy/promotion targets, including for sure all living person portal topics, need to be anchored by core policy citation expectations, and this makes them unsuitable for portals. Even Portal:United States is screaming "product of unconscious biases". --SmokeyJoe (talk) 01:53, 23 May 2019 (UTC)
@SmokeyJoe: It took me just a few minutes to compile a list of several portals that have fewer than 50 views per day and were nominated for deletion but kept: Cetaceans, Ireland, Architecture, Robotics, Apple Inc., Geophysics, and Harry Potter. There are probably about 35 portals in this situation, and those are just the ones that have been nominated. Are you denying the validity of these decisions? RockMagnetist(talk) 04:12, 23 May 2019 (UTC)
No, I don’t deny them. I think it is unfortunate that these things are going to the artificially rushed and critical forum of MfD. I think the questions should go to RfCs. These portals that you list are not terrible, I don’t think their problems rise to the level justifying deletion. Deletions means that they should never have been started. I think fewer portals should cover more content. Nation portals like Ireland and the United States I think would be better managed if covered by Portal:Nations. Exactly how to restructure them all, I’m not sure, I think it should be talked about. I prefer to say that the top Portals are excellent, but could still be improved. The long tail of weak portals needs to be cut. In the middle, I am not sure, except that in the current state of affairs, they are not working. —SmokeyJoe (talk) 04:28, 23 May 2019 (UTC)

Maintainers, or lack thereof

Portal maintainer stats Pages %
Maintainer 113 21.24%
No maintainer 419 78.76%
Total 532
Purge this page to update the totals.

Yesterday I created two tracking categories:

They are automatically populated by ((Portal maintenance status)). If any of maintainer, maintainer2 etc parameters is non-blank, then it's a yes; otherwise no.

And per the table on the right, the stats are grim. Currently less than 8.5% of portals have one or more named maintainers.

It may be that some portals are maintained, but the editors who maintain them are not aware of how to identify themselves as maintainers. How about mass-posting a message to the talk page of each portal in Category:Portals with no named maintainer, to any maintainers to identify themselves? --BrownHairedGirl (talk) • (contribs) 14:23, 15 May 2019 (UTC)

This is useful. There are several portals that I've contributed to and even keep an eye on on a regular basis, but I haven't added myself as "the" maintainer. Having named maintainers seems a bit WP:OWNy to me, so I've tended to be reluctant to add my name. WaggersTALK 15:07, 15 May 2019 (UTC)
@Northamerica1000, I dunno. Maybe it was a new idea then?
But now that the concept is a year old, and seems to be accepted, shouldn't it be promoted? --BrownHairedGirl (talk) • (contribs) 16:24, 15 May 2019 (UTC)
A concern is that at MfD, people are demanding maintainers, but it appears to be a relatively new concept that has not been advertised or promoted until very recently. Doesn't seem right: people demanding a maintainer while users were not made aware of the possibility of being able to do so from the start. Amounts to changing the goalposts when the goalposts were not made visible in the first place. North America1000 16:27, 15 May 2019 (UTC)
@Northamerica1000, I just looked at WP:POG's history. Here's the lead of the January 2007 version: Please bear in mind that portals should be about broad subject areas, which are likely to attract large numbers of interested readers and portal maintainers. Do not create a portal if you do not intend to assist in its regular maintenance.
The wording has been strengthened, but the goalposts have been visible for over 12 years.
Since last year, we have a way for maintainers to register their interest. So I'l ask again: what do you think of a notification reminding maintainers to add their names? --BrownHairedGirl (talk) • (contribs) 17:23, 15 May 2019 (UTC)
A note in the next project newsletter about the option to add oneself as a maintainer would be a good thing. North America1000 17:24, 15 May 2019 (UTC)
@BrownHairedGirl: we've worked cooperatively together and, I think, have built a degree of trust. But I'm concerned that every time the dust settles on a round of portal deletions, you find another statistic to propose yet more deletions and, as others have pointed out here, the need for a named maintainer has never before been a requirement. I only put my name down for the portals I was overseeing because I feared TTH replacing them with those awful auto-portals. So I'm beginning to lose my faith in the cooperative process we were engaged in. Please can we stop finding reasons for more deletions and focus on the aims, the processes and the standards needed for decent-quality portals? Once that's done, we'll all be on the same side, promoting some portals, improving other portals and, yes, culling others that don't meet our jointly agreed criteria. Bermicourt (talk) 18:11, 15 May 2019 (UTC)
@Bermicourt, I think you misunderstand my purpose. We have indeed built up some trust, which I value, but you don't seem to be fully AGFing here. This isn't about seeking a reason to delete.
I have been working fairly consistently on abandoned portals for several weeks. I have MFDed those which just have a static set of abandoned sub-pages. I have never cited the lack of a named maintainer as a reason for deletion, and I don't intend to start doing so. My interest is in whether a portal is maintained, and the lack of a label doesn't turn a good portal into a bad one.
For me, the purpose of identifying a maintainer is wholly positive. If an apparently-abandoned portal actually has a maintainer who who has actual plans to improve it, then that for me would be a reason not to take it MFD. --BrownHairedGirl (talk) • (contribs) 18:46, 15 May 2019 (UTC)
You're right and I probably have misread your intent. I'm sure there will be maintainers out there like Waggers who haven't put themselves forward or whom we haven't yet identified. So plugging that gap would be a useful exercise. Bermicourt (talk) 06:29, 16 May 2019 (UTC)

Putting this in perspective

A lot of the contributors to this page feel that the number of portals, currently about 1200, is much too large. But let's put it in perspective. According to Size of Wikipedia, about 50,000 new articles are added to Wikipedia per month. All of the portals combined take up a negligible amount of space in Wikipedia.

Compare portals to navboxes. I'm not sure how I'd go about determining how many there are; you can search for templates, but that includes a variety of other kinds of template. But clearly, there are a lot more navboxes than portals. They tend to be somewhat smaller than portals, but each is transcluded in several pages. Template creep is almost inevitable and can result in monstrosities. And more get added all the time. The problem is particularly egregious in articles on medicine. As an example, Pulmonary insufficiency is a short article that gets about 50 views a day; it has two giant navboxes at the bottom with a couple of hundred links. These make it seem that about 200 articles link to this one when in fact only 12 would without the templates.

As far as I can tell, navboxes rarely provide much navigational help, but they create clutter and overlinking (which slows down page loading), and they obscure the true relationships between articles. But a lot of Wikipedians are convinced that they are valuable, and once a navbox is created it's really hard to get rid of it. Appeals to WP:NAVBOX often fall on deaf ears.

I'm not denying that bad portals should be removed, but if a portal that is reasonably well designed attracts only 50 views per day, what harm is it doing? Certainly much less than bad navboxes. Some editors have been chiding portal maintainers for wasting their time; I ask those editors - are you making the best use of your time? RockMagnetist(talk) 05:46, 21 May 2019 (UTC)

Wise words, indeed @RockMagnetist. Something I've been intending to raise (particularly the utter fixation by the anti-portal squadron about pageviews), but have been too busy IRL to do so far. More later, hopefully. -Cactus.man ✍ 19:32, 21 May 2019 (UTC)
I think that's a helpful perspective. Pageviews are being used, uniquely, as a stick to beat portals when, as I and others have pointed out, they're not even in mainspace so why would we expect hundreds of views. RockMagnetist's point about bad portals is entirely right and I have voted to remove portals myself that I don't think cut it. His other valuable point is about navboxes that grow too large for the average article size. But navigation is a role that well-designed portals could handle well. Bermicourt (talk) 20:41, 21 May 2019 (UTC)
With the current "random display" format of portals, navigation is not a role a portal could handle well. An oversized navbox might be replaced by an Outline, but not by a Portal. But the first step is to establish what a Portal is supposed to do. So far, there is no consensus. — Arthur Rubin (talk) 21:04, 21 May 2019 (UTC)
I disagree: there is consensus around the existing WP:POG guideline, and all the WP:IDONTLIKEIT won't change that. While some of its terms are not specific, there is NO consensus to change it, and no open proposals to change it. WP:POG tells us all we need to know. UnitedStatesian (talk) 21:54, 21 May 2019 (UTC)
There is a WikiProject consensus for WP:POG, just as there was WikiProject consensus for WP:RY. I have to say that there doesn't seem to be much support for WP:POG recently. — Arthur Rubin (talk) 05:31, 22 May 2019 (UTC)

RockMagnetist, see WP:OTHERSTUFFEXISTS. --BrownHairedGirl (talk) • (contribs) 21:02, 21 May 2019 (UTC)

Yes would be best to read WP:Some stuff exists for a reason. --Moxy 🍁 21:13, 21 May 2019 (UTC)
Yes, especially: "This essay is not a standard reply that can be hurled against anyone you disagree with ..." My discussion above is intended to stimulate some thinking. A lot of good has come out of this re-examination of portals, but when people are reduced to talking about judging portals by pageviews, I suspect that they're at the point of diminishing returns. RockMagnetist(talk) 21:52, 21 May 2019 (UTC)
WP:POG makes no mention of required pageviews, nor indeed does any other policy or guideline on WP. Just click on "Random Article" on the main page. My last 4 results were:
  • Jan Henne Started Nov 2007, Average of 4 pageviews per day this year, 1 Editor.
  • Butter Lamp Started Jan 2015, Average of 6 pageviews per day this year, 0 Editors.
  • Pang Qing Started Feb 2006, Average of 17 pageviews per day this year, 1 Editor.
  • Haynes, Hanson and Clark Conditions Stakes Started June 2007, Average of 2 pageviews per day this year, 1 Editor.
Hardly a stellar result, and it's probably symptomatic of a massive percentage of our content, suggesting "readers are not interested" or there is "a lack of maintainer". But there is no policy reason to delete, For GOOD REASON - this project was promoted from the beginning as being "the sum of all human knowledge", not that of "the sum of all human knowledge, that some like or approve of". Or do you want to spend the next 10 years at AfD?
Limited pageviews are just a weak argument, not based in policy, promoted by the "Anti-Portal Squadron" (as opposed to the "Portal platoon") to attempt to bolster their deletion efforts (see WP:AADD). Along with other spurious arguments like "no named maintainer" (see WP:NOTCOMPULSORY), "unlikely to attract readers" and other tenuous justifications for deletion of stuff the nominators don't like, I see these as desperate attempts to railroad through a non-policy compliant deletion spree on the back of a massively ill-considered Pile-On. Who on earth had the time or inclination to properly review over a thousand Portals nominated for MfD, other than to just default to "most are !voting delete, seems like the right thing to do". A good example of massively misguided Groupthink and not really an "overwhelming consensus".
Having said all that, there are many Portals that have been ill conceived and quite justifiably deleted, just on the whole on a knee jerk response rather than sound policy reasons. - Cactus.man ✍ 23:13, 21 May 2019 (UTC)
Wikipedia pageviews have a distribution with a long tail (see this article), and that may be a feature, not a bug; long tails are an integral part of the strategy of companies like Amazon. And if it's acceptable for articles, why not for other kinds of page? RockMagnetist(talk) 00:09, 22 May 2019 (UTC)
WP:POG has simply been used when it appears to support an argument for deletion and ignored when it doesn't. Frankly, it's vague, it's way out of date and largely irrelevant because, as editors on both sides of the debate have conceded, we need to agree new standards for portals and new processes for creating and maintaining them. But at the moment the effort continues to be to reduce them to the lowest possible number without violating the outcome of WP:ENDPORTALS. The irony is that, if we agreed the standards first, there would be much greater consensus on whether to keep, improve or delete individual portals and the eventual number would sort itself out. Bermicourt (talk) 06:23, 22 May 2019 (UTC)
@Bermicourt, the requirement for pageviews, maintainers, and broad topics has been part of WP:POG since 2006[3], with he wording almost unchanged since then. If editors believe that portals should be about narrow topics, and/or be unmaintained and/or have no regard to pageviews, then they are free to open an RFC to change the guidelines.
I agree that new guidance is needed about what portals should do and how they should be constructed, but I see little chance of those requirements being changed. But if you disagreement, then why not an open an RFC calling for more unmaintained micro-portals which nobody reads? --BrownHairedGirl (talk) • (contribs) 19:49, 22 May 2019 (UTC)

@Cactus.man, we have been over this a million times before, so I will try to be brief:

The cleanup process has gotten rid of nearly of all of last year's tsunami of spam, and has made good progress through the enormous pile of detritus which has accumulated through a decade of neglect by this project.

As I have said before, if there is going to be a future for portals, then this project needs to get a grip on the core questions of how portals can actually add value to the reader and thereby meet the WP:PORTAL principle that "Portals serve as enhanced 'Main Pages' for specific broad subjects". But if it continues to be dominated by moans that others are clearing the project's decade-old housekeeping backlog, then sooner or later there will be further efforts to just mass-cull much of the rest. --BrownHairedGirl (talk) • (contribs) 08:00, 22 May 2019 (UTC)

@ BrownHairedGirl I was wondering when you'd show up to trot out the same old mantra once more. Oh, and congratulations on keeping it "brief", perhaps it's you who is in need of a new dictionary. So, here's my "brief" reply:
Yes, Portals can rightly be deleted, and many have been. But you presented two massive batches, of around 1400 and 1150 in just two listings. I grant you that the deletion rationale was well researched and presented, but my comments relate to the logistics of expecting !voters to realistically assess such a number of Portals and that's what, in my view, calls into question the validity of "true consensus". Hence the Groupthink comment. I never claimed that WP:NOTCOMPULSORY prohibits deletion of Portals, just that it supports the view that frequently asserted argument that a "lack of maintainers" is the TRUE "red herring" repeated ad nauseam in Groupthink manner in the various deletion arguments (much like quoting low pageview numbers). You usually take 30 minutes on each one in preparation, so you were a very busy bunny indeed preparing those two nominations for 2550 portals - that's around 53 days of solid work (without sleep) or, more realistically, over three months of steady work, allowing for downtime, rest and sleep - IMPRESSIVE. So, I'll leave the comedy to you, because you seem to be rather good at it. I don't believe the moon is made of blue cheese, just as I don't believe that Ronald Regan was born in Ballyporeen. So we can both agree on that and I suggest we leave our disagreement there unless, of course, you absolutely insist on having the last word. -Cactus.man ✍ 11:02, 22 May 2019 (UTC)
@Cactus.man: are you actively trying to miss the point?
The mass nominations grouped pages with a shared characteristic. The issue for !voters to decide was a question of principle: whether that shared characteristic merited deletion. That decision of principle would have been the same whether the set of pages involved was one or one million. Your choice to label a well-attended discussion on a single issue of principle as "groupthink" says very little about the discussion, but at lot about your style of rhetoric.
No, as you well know, I didn't spend 30 minutes on each of those mass-nominated portals; and as you well know, that comment refers to my individual nominations. For the mass nominations, I spent over 20 hours spread over two days experimenting with methodologies to make the list, and I fully documented the process used, and invited editors to check and verify it.
Anyway, if you believe that the consensus was invalid, WP:DRV is thataway. Unless and until you succeed in getting those discussions overturned at DRV, please accept the consensus.
And kindly desist from misrepresenting what I say about maintenance. I repeatedly nominate portals at MFD because they have not been maintained, with detailed analysis of how the pages have stagnated, e.g. at WP:Miscellany for deletion/Portal:Sindh.
Anyway, you don't seem to have moved on from the mode you were in at WP:Miscellany for deletion/Mass-created portals based on a single navbox, when you had a long rant about my alleged sloppiness ... and cited in evidence a portal which had not been nominated.
Anyway, this is just another thread lamenting the cleanup of spam and abandoned dross. If the portal fans put a fraction as much effort into establishing a broad consensus on how portals can actually add value for readers, then portals might have future. --BrownHairedGirl (talk) • (contribs) 12:28, 22 May 2019 (UTC)
  • I'm not ready to wash my hands of the portalspace yet. Unlike the Book: space, it is, FBOFW, highly linked, and not just the 11 portals that are linked directly from the Main Page. Portal:United States has over HALF A MILLION inbound links from article talkpages. And since the portals that will be left will remain so highly accessible, it remains in everyone's interest to make sure they are of as high quality as possible. UnitedStatesian (talk) 19:55, 22 May 2019 (UTC)
  • @UnitedStatesian, yes, I agree that the quality of portalpsace as a whole will be improved by continuing to remove the dross.
But please note that what we are doing now is simply removing the very worst. Beyond that there are hundreds of portals with less than ten selected articles which adds little to the topic ... and then there is the big issue: unsourced content forks.
I just checked Portal:United States. It has 1202 subpages, so I ran then through AWB to look for "<ref". Only one of the 1202 (Portal:United States/Selected article/7) has even a single "<ref" tag. This is a blatant violation of the core policy WP:V, and it will have to be dealt with. --BrownHairedGirl (talk) • (contribs) 21:17, 22 May 2019 (UTC)
@BrownHairedGirl: that's why in rescuing Portal:Nigeria I used ((Transclude lead excerpt)): since the resulting subpage always matches the article, any info in the subpage can be verified at the article, and WP:V concerns disappear. Now we just have to edit every other subpage . . .UnitedStatesian (talk) 22:02, 22 May 2019 (UTC)
I know how you feel.--Auric talk 22:23, 22 May 2019 (UTC)
  • @UnitedStatesian, unfortunately the way it has been done with Portal:Nigeria is to keep the forest of subpages, but use ((Transclude lead excerpt)) in each of them. I count 61 subpages in all, which is still a maintenance nightmare, because a) there is no external indication of whether they are forks; b) even those which still use transclusions can be rewritten with any other text. The only way to solve this is to use ((Transclude random excerpt)) in the main portal page, and dump the subpages. --BrownHairedGirl (talk) • (contribs) 22:24, 22 May 2019 (UTC)
@BrownHairedGirl: I disagree; I make sure I have every subpage of Portal:Nigeria on my watchlist (as I think every maintainer should), and so catch any rewritings each time I use the Portal+all changes filters to view my watchlist. Very easy to monitor. The beauty of subpages is they can be shared across portals using redirects, so Hakeem Olajuwon could be excerpted in the bios of both Portal:Nigeria and Portal:Basketball, for example. But should a portal-by-portal decision whether to use ((Transclude lead excerpt)), or ((Transclude random excerpt)), or a mix of both; I wouldn't want to mandate, UnitedStatesian (talk) 14:48, 23 May 2019 (UTC)
  • @UnitedStatesian, once transclusions are being used, there is no need to share sub-pages. That was a useful exercise when the sub-pages were content forks, but it is pointless complexity when the shared data is just an article title.
And while I am sure that you have watchlisted all 61 subpages, the problem is that nobody else is going to do that unless they are also a dedicated maintainer. The advantage of listing all the topics in the main portal page is that it then takes only watchlist entry to watch the portal, so any editor can do it in the usual way: just watchlist the portal page.
The whole business of keeping subpage farms is outdated, and a huge vulnerability. --BrownHairedGirl (talk) • (contribs) 16:48, 23 May 2019 (UTC)
That's one more "<ref" than the main page, then. As UnitedStatesian points out, claims in the excerpt can be verified in the body of the article, just as the information in a lead can. References have never been routinely copied to portals (nor the main page). Indeed, many of the changes requested to Module:Excerpt (which does the heavy lifting for ((Transclude lead excerpt))) were to remove a reference which had slipped through into the excerpt due to being in some obscure format not recognised by the filter. The clear convention is to omit references in such cases. Consistently following the precedent set by the main page and almost every portal for a decade does not need to be "dealt with" and is not a deletion rationale. Certes (talk) 00:20, 23 May 2019 (UTC)
@Certes, the precedent set by the main page is of refs being omitted for a set of excerpts closely monitored by an extensive and highly active team which provides several layers of assessment before anything is published.
The reality with portals is of sprawling sets of content forks up to 15 years old which are randomly chosen by software. Only yesterday I encountered one which had been hijacked two years by a completely different topic, and nobody had fixed it.
If you really want to insist that the main page is the model, then we need to take offline any portal which doesn't have at least one team of 4 named active maintainers like WP:TFA. --BrownHairedGirl (talk) • (contribs) 06:12, 23 May 2019 (UTC)
I think that replacing outdated forks (text pasted from old versions of articles) is achievable. We should check that the text actually is an outdated fork rather than a lovingly hand-crafted summary but that's tedious; we might boldly go ahead anyway with an edit summary of "please revert if you intend to maintain the previous version". Clearly we need to limit the process to portals which no one intends to nominate for deletion. (I have stopped wasting my time improving threatened pages and can't expect others to do so.) Therefore, identifying such portals has to be the first step. We also need to check that no one feels that transcluding a template written by a supporter turns a portal into "automated junk created by a portalspammer" and triggers an MfD. Certes (talk) 15:51, 23 May 2019 (UTC)
Folks, we all need to stop throwing teddies around and tarring all the opposition with the same brush or leaning on a single pass/fail criterion for portals. Portal-keepers do not like rubbishy, un-maintained portals and most portal-deleters are willing to accept that some decent and maintained portals have a place. What we have yet to agree on is a) the purposes of portals, b) the standard of acceptability (what does a good portal look like - it falls out of (a)) and c) what sensible processes we should adopt to ensure we get there. If we can agree all those, we won't need to find numerical criteria carefully chosen to produce the keep or delete result we want and there will be far fewer arguments. Bermicourt (talk) 20:59, 22 May 2019 (UTC)
If the lack of viewers is a question, then we can try solutions.
If many portals are poor, then we can fix them or delete them.
If we don't agree what's the purpose of a portal, then let's discuss it.
--NaBUru38 (talk) 17:01, 25 May 2019 (UTC)

MfD nomination of Portal:Kiribati

Portal:Kiribati, a page which you created or substantially contributed to, has been nominated for deletion. Your opinions on the matter are welcome; you may participate in the discussion by adding your comments at Wikipedia:Miscellany for deletion/Portal:Kiribati and please be sure to sign your comments with four tildes (~~~~). You are free to edit the content of Portal:Kiribati during the discussion but should not remove the miscellany for deletion template from the top of the page; such a removal will not end the deletion discussion. Thank you. Robert McClenon (talk) 00:27, 6 June 2019 (UTC)

More