Wikipedia Portals
– About Portals – List of Portals – WikiProject Portals

In portal deletion discussions, reference is always made to the Portal Guidelines. The guidelines state: '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.'

Note that on 26 September 2019, per the discussion at Wikipedia:Village pump (policy)/Archive 153#RFC: Formalize Standing of Portal Guidelines as a Guideline (18 July 2019), it was determined that there is clear consensus that the "Portal guidelines" are not, in fact, official guidelines.

Portals are a feature of Wikipedia that can be used for showcasing, navigation, instruction or promotion, and for fun. They have always been a feature of Wikipedia, and, since 2006, have been in Portal space. Some editors are very enthusiastic about portals. It's not entirely clear why, but some people think they must be something that are liked because they are seen as technically neat, rather than as functionally valuable.

Portals are intended to be maintenance-intensive, but normally they are not maintained. The advocates of portals are often perceived as denying the need for maintenance.

Early in 2018, when there were just under a thousand portals in existence, there was an RFC to delete all portals. It was closed with a consensus not to delete all portals, but with no other specific conclusions. Then a task force decided that we (English Wikipedia) needed more portals, and decided to create thousands of more portals. They did this more or less quietly, and had created a total of 5700 portals, and most of the new portals were just automated crud. The thousands of portals were reported at WP:AN, and since then some have been bringing portals to MFD for deletion. The portalistas have been claiming that this is a "war on portals". (The opposite view is that they conducted a sneak attack by creating thousands of them, but that is only one opinion.) Most of the portals that were created in the wave of reckless portal creation were deleted in two bulk nominations to MFD. But since then, a few other editors who are skeptical about how much portals add have been working slowly to nominate some of the abandoned portals for deletion. As of October 2019, the number of portals is just below 600. Many of the portal deletion debates are bitter and unpleasant.

Any informed assistance that we can be given in checking the status of portals and deleting the cruddiest ones will be appreciated. Some of those who are trying to clean up portals would like to see a few hundred high-quality portals, and some would prefer to see more like twenty, or to delete all of them except for the main page, which is a super-portal and is labor-intensive.

It is commonly stated that a particular topic is a broad subject area, and that that is sufficient reason to have a portal. There are at least three problems with using the concept of "broad subject area" as the sole reason for considering whether a portal is in order.

The first is that the concept of "broad subject area" is often used in an absurdly expansive way that abuses the concept of breadth, to argue not only that a country is a broad subject area (it probably is), but that a capital city of a country is a broad subject area (it might be or it might not be), or a state of a country is a broad subject area (it might be or it might not be), or an area of knowledge is a broad subject area (it might be or it might not be).

Philosophers make a distinction between a priori and a posteriori knowledge, between knowledge that is available in advance and knowledge that must be based on observation. The advocates of portals frequently say that a particular topic is a broad subject area, and so the subject should have a portal. It is possible to decide a priori that particular types of subject areas, such as countries, or big cities, are broad subject areas. However, that is an incomplete quotation of the portal guidelines, and, because of its incompleteness, is misleading. The portal guidelines say that "portals should be about broad subject areas, which are likely to attract large numbers of interested readers and portal maintainers." It is not possible to decide a priori that a subject area will attract readers and portal maintainers. That must be observed, and assessed a posteriori.

The second problem is that experience has not always shown that topics that are thought to be broad subject areas attract large numbers of readers, let alone portal maintainers. Experience has been that the number of readers of a portal is usually less than one-hundredth as many as the number of readers of the head article. Most readers who choose to explore a topic area that has a portal do so not by using the portal but by following the links or sometimes the categories. One reason is that search engines provide access to articles, bypassing any portals, and many readers who come into Wikipedia directly do so by looking for broad-subject articles, specific articles, or categories. Broad subject areas do attract readers, but not always to portals.

Third, however, in deciding whether a portal should be created or whether an existing portal should be kept, is the original definition of a portal. 'Portals serve as enhanced "Main Pages" for specific broad subjects.' The Main Page is a labor-intensive effort. Many Wikipedia volunteer editors spend considerable time contributing to the Main Page. Portal maintenance is likewise a labor-intensive effort. A portal should not be created unless someone is willing to provide the labor of maintaining the portal. Experience has shown that automated portals provide only a pretense of maintenance, and that only human effort can provide a miniature Main Page. A portal without a maintainer is a gateway (a portal) to nowhere. The answer isn't to create portals and wait for portal maintainers. (That is like Waiting for Godot.)

Part of the answer is not to create portals that do not already have portal maintainers. At least, creating portals that don't have maintainers is part of the problem.

Two newish features of the Wikimedia software mean that the article and navbox offer all the functionality which portals like this set out to offer. Both features are available only to ordinary readers who are not logged in, but can be tested without logging out by right-clicking on a link, and the select "open in private window" (in Firefox) or "open in incognito window" (Chrome).

1. mouseover: on any link, mouseover 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. Try it on a template.
2. automatic imagery galleries: clicking on an image brings up an image gallery of all the images on that page. It's full-screen, so it's actually much better than a click-for-next image gallery on a portal. Try it by right-clicking on an article.

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

That sets a high bar for any would-be-portal-builder to vault if they try to satisfy the WP:PORTAL principle that "Portals serve as enhanced 'Main Pages' for specific broad subjects". It would take a lot of work to make a portal which genuinely offers more than a head article that has a navbox.

But maybe someone will find a way to make such a better portal, and a team of editors to maintain it.

One common method for the design of portals, in use at least since 2005, has involved sometimes large numbers of subpages of the portal, one for each selected article and picture, and sometimes for news items and Do You Know (DYK) items. Often the subpages for selected articles consist of a copy of the original article, or a copy of the first part of the original article. The subpages for In The News (ITN) and DYK items may also be copies of the lead paragraph or a portion of the article page. This approach to design of portals is sufficiently commonly used that it can be considered standard. However, it is an honorable experiment that has failed, and should be abandoned. In numerous cases, it has been found that portals have displayed outdated and incorrect information to the reader. These discrepancies have been especially common with, but not limited to, political leadership. These discrepancies are a serious problem because they cannot be readily corrected. Wikipedia is the encyclopedia that anyone can edit, meaning that any reasonably computer-literate person can edit an article; but editing the displayed information in a portal requires specialized technical knowledge of how portals are implemented, which is presumably why errors persist, sometimes for years. An editor who has Twinkle installed can tag articles in need of editing if they do not have the time or knowledge to fix them; but tagging via Twinkle is not available for portals. Experience has shown that the use of portal subpages that copy portions of articles results in outdated information being displayed, sometimes for years, because it is difficult to correct. This design technique, partial article copies, has been an honorable experiment over the course of more than a decade, but the experiment should be assessed to have been a failure. Some other design approach for portals should be used in the future.

Some editors have stated that particular levels of regions should have portals. One editor wrote (in April, in an MFD): "all countries should have portals". An administrator wrote: "I consider U.S. states to meet WP:POG guidelines in terms of being broad enough in topical scope to qualify for a portal." Such statements raise a two-part question, having to do with people, and with policies.

The first part of the question is: Who is expected to do what in order to provide the portal? Should Wikipedia provide and maintain a portal? Should Wikipedia provide a portal without maintaining it? Should Wikipedia provide a portal, contingent on having a portal maintainer and a portal maintenance plan? Should the Internet, of which Wikipedia is a prominent site, provide a portal? If only that, the government of the nation, state, or province can and almost certainly does provide and maintain a portal in the form of its web site, as do lesser regions such as counties, cantons, districts, communes, cities, towns, townships, boroughs, and villages.

If Wikipedia is expected to provide and maintain a portal, how can that obligation be reconciled with Wikipedia is not compulsory? If Wikipedia is only expected to provide an empty portal, what good is that? It appears that the implication is that Wikipedia is expected to provide a portal to a maintainer and to continue to keep the portal facing outward toward the readers whether or not it is being maintained. Portal advocates who think that particular levels of regions "should have" portals should clarify what obligation they are implying and on whom.

The second part of the question is how the idea that countries or states "should have" portals should be reflected in Wikipedia's policies and guidelines. At present, the page that is designated as the Portal Guidelines states that portals should be about broad subject areas that will attract readers and portal maintainers. Portal advocates have focused on the reference to "broad subject areas" and have disregarded the two-part reference to readers and portal maintainers. At present, the status of the portal guidelines is in dispute. Those who would like to retain existing regional portals, and possibly create more regional portals, may either deal with the existing guidelines or propose to revise them. If they prefer to deal with the existing guidelines, there are two issues. The first is that the status of the guidelines is in doubt, appearing to have been a failed proposal. The second is that the existing document refers to readers and maintainers, who cannot simply be assumed or willed into existence. Since the present (contested) guidelines refer not simply to broad subject areas, but to broad subject areas that will attract readers and portal maintainers, any specific portal can be shown by observation not to be attracting readers or maintainers.

The other option for the advocates of regional portals, or for anyone who wants to provide better guidelines with regard to portals, would be to publish a Request for Comments to implement new portal guidelines, either the old guidelines, or a slightly revised version of the old guidelines, or an entirely new set of guidelines. In that case, advocates of regional portals should be on notice that the new guidelines either should explicitly identify certain subjects that are considered portal-worthy even without maintenance, or it can be understood that regional portals, like other subject areas, are only considered to be broad subject areas if they demonstrate that they attract readers and maintainers.

If an editor really wants to maintain a portal, it is a hobby, and it should not be discouraged. A portal can be kept if it has an active portal maintainer, regardless of whether it is a broad subject area. Broad subject areas will not necessarily attract portal maintainers (and will not necessarily even attract readers to the portal), but portal maintainers should be encouraged or at least not discouraged if they want to make the investment in time.

• User:DexDor/Portals (user essay)
• Wikipedia:Consultation on the future of portals
• Wikipedia:Village pump (proposals)/RfC: Ending the system of portals