Wikipedia:Consultation on the future of portals

This is an essay.

It contains the advice or opinions of one or more Wikipedia contributors. This page is not an encyclopedia article, nor is it one of Wikipedia's policies or guidelines, as it has not been thoroughly vetted by the community. Some essays represent widespread norms; others only represent minority viewpoints.

This page in a nutshell: This is an overview of the key arguments, data and proposals in the rather lengthy discussion on the village pump proposal to remove portals. The proposal in question: Wikipedia:Village pump (proposals)/RfC: Ending the system of portals.

Shortcuts

Please feel free to contribute to this overview of the 2018 portal discussions.

Key arguments

The original proposal suggested deleting all portals.

The main arguments in support of the proposal were:

Very few Wikipedians use portals, and therefore very few Wikipedians maintain them.
As very few Wikipedians maintain portals, they often contain out-of-date and/or irrelevant information.
Many portals on certain subjects are less useful as an introduction to their subject than the main article(s).
The time Wikipedians do spend on maintaining portals would (thus) be better spent on maintaining articles.
Because they are poorly maintained and have few watchers they attract vandalism and POV pushers, which is not quickly reverted.
The only portals with significant traffic are the 8 linked from the top of the mainpage, and these get far less traffic than the corresponding articles

The main arguments in opposition to the proposal were:

Although few Wikipedians use portals, many visitors do. (See table below to that effect.)
Portals are a useful and innovative way to get information on a topic.
Blanket deletion of an entire namespace, containing a total of 185,578 pages, may have many unintended consequences – as mass deletions have done in the past.
Many articles are also poorly maintained and get few viewers and nobody is suggesting the discontinuation of the article-space.
Portals which nobody wants to improve can be deleted individually. The good portals should not all be deleted because of a minority of bad portals.
There are clearly some portals that are well-maintained, such as Portal:Current events, and there is little reason to delete them.
The deletion would discourage editors to spend their time and efforts on Wikipedia.
Some value would (thus) be lost by deleting all portals. An overhaul, by contrast, could potentially create value.
Portals which are in bad shape can be improved. The solution is improvement and not mass deletion.
Wikipedias in other languages want to stay in the Portal system. The English Wikipedia would be isolated without Portals.
New technology allows the building of Portals without subpages. The Portals need to be improved with this software and should not be deleted.
The WikiProject Portals has been restarted, overhauled, and is under new curation. In addition to working on portals directly, the team is also upgrading portal design in the form of new components that require less maintenance.

Data

Some users collated useful data; this is reproduced here. Please feel free to add any further relevant data to this section.

Number of pages in affected namespaces

Count of portal pages from Wikipedia:Database reports/Page count by namespace originally supplied by Renata3, supplemented with a total.

Namespace ID	Namespace	Total pages	Pages with redirects	Pages without redirects
100	Portal	148,868	13,188	135,680
101	Portal talk	36,710	2,701	34,009
TOTAL		185,578	15,889	169,689

2018 main page link traffic

Supplied by Andrew Davidson, this table, edited to show only links that are permanently featured on the main page, demonstrates that the main page linked portals receive a substantial amount of traffic.

Rank	Page	Readers in 2018
1	Deaths in 2018	10,440,731
2	Portal:Current events	4,387,154
4	Wikipedia:Featured articles	2,677,238
5	Wikipedia	1,795,210
8	Wikipedia:Community portal	1,089,287 (not a Portal)
9	English language	1,022,730
24	Portal:Contents/Portals	343,999
25	Portal:Arts	251,984
29	Wikipedia:In the news/Candidates	225,142
30	Wikipedia:Your first article	224,475
33	Portal:History	210,188
35	Encyclopedia	207,631
36	Portal:Biography	196,294
37	Portal:Geography	193,451
38	Portal:Technology	191,849
44	Portal:Mathematics	129,103
47	Portal:Science	122,774
49	Wikipedia:Help desk	119,773
51	Free content	104,088
52	List of historical anniversaries	100,432
53	Wikimedia Foundation	93,859
58	Wikipedia:Recent additions	78,397
62	Wikipedia:Featured pictures	70,178
63	Wikipedia:Introduction	67,917
65	Portal:Society	63,415
66	Wikipedia:Reference desk	55,095
67	Template talk:Did you know	52,856
73	Wikipedia:Village pump	38,022
83	Wikipedia:News	25,514
87	Wikipedia:Local Embassy	22,625

Proposals

Delete all portals (original proposal)

Should the system of portals be ended? This would include the deletion of all portal pages and the removal of the portal namespace. 14:23, 8 April 2018 (UTC)
Support Taking at look at one example of Portal:Cricket: it contains a summary of the lead of Cricket, which is out of date (there are now twelve full members); obscure random articles that is just something someone took the effort of making them good - like Yorkshire captaincy affair of 1927; out of date news, random anniversaries and other random stuff like that. Readers aren't looking for random cricket-related stuff - it is clear, from the extremely low page-views, that readers don't care about portals. The most-viewed portals are purely from being featured on the main page; but for example Portal:Science gets only 8 out of 100000 of the views of the main page, a few hundred people a day, and they are likely from random clicks - not from people interested; which would likely account for most views of other portals too. There have been suggestions of automated systems for helping to develop portals, which even if developed wouldn't help, because portals aren't useful in any way. Personally, I've never felt the desire to read, say, a random science article, which is what portals consist of (most portals indeed have literally randomly selected content from a list)

In essence, portals try to straddle reader-facing and editor-facing stuff, but are terrible at both. They aren't really part of the encyclopedia; nor do they help in the backend - they don't benefit the encyclopaedia in anyway (the main page, which could be called a portal of everything, in contrast, encourages people to improve articles). Any navigational purpose, which I don't think portals help with at all, is better served through outlines. Featured articles and other stuff in a topic are cared more by wikiprojects, which generally link them already. Implementation could be reasonably easily done, as nearly all, I reckon, portal links in mainspace and in all pages indeed are through templates like {{portal}} (in all pages I estimate 99% of links are from being linked in wikiproject banners), which can be blanked to remove links; once the links are gone from mainspace, the portal pages can be deleted. Galobtter (pingó mió) 14:29, 8 April 2018 (UTC)

Delete all portals except current events

Support – but what will happen to actually useful portals, such as Portal: Current events? Nixinova 04:33, 11 April 2018 (UTC)
@Nixinova It would be deleted per the argument below that this proposal does not go into detail. - Knowledgekid87 (talk) 13:29, 12 April 2018 (UTC)

No, any competent closer will see the lack of support for their deletion or marking historical - rfc outcomes don't have to match the original wording and can exclude those specific portals. Galobtter (pingó mió) 13:32, 12 April 2018 (UTC)

Delete all portals except the most significant

It seems that there's consensus or at least significant support to keep at least some portals: notably Portal:Current events and the like. Assuming the proposal to deprecate the Portal namespace passes, what will happen to these pages? Will the Portal namespace be kept for them alone, or will they be moved to another namespace like the Wikipedia namespace? Narutolovehinata5 13:39, 11 April 2018 (UTC)
This is the issue... nobody is thinking these things out. - Knowledgekid87 (talk) 13:43, 11 April 2018 (UTC)
Hmm? Like I suggested above, the wikipedia namespace seems fine for these, seeing as they are barely like other portals, and the moves shouldn't be too onerous to do Galobtter (pingó mió) 14:58, 11 April 2018 (UTC)
These are part of Wikipedia's core navigation system: Portal:Contents (see /URL for the list) Such as retain the Portal namespace just for these: Portal:Contents, Portal:Featured content, and Portal:Current events. Thank you. — The Transhumanist 10:36, 11 April 2018 (UTC)

Keep portals and only delete the portals which are in bad quality

Oppose - Why not nominate them on their individual merit rather deleting them in bulk? There are over 1 million WP articles which are in poor state that doesn't mean delete them in bulk. They will improve if we decrease their quantity. Störm (talk) 18:02, 16 April 2018 (UTC)

Overhaul through large-scale automation

Oppose in favor of technical overhaul. Specifically, I would like to see a portal system with:
The automatic production of article synopses of the appropriate length when articles relevant to the portal topic are accepted and the ability to edit these synopses if they need improvement.

The automatic addition of these synopses to the pool from which the portal draws its content selections.

The ability to sort or filter the article synopses on the "more articles" page (or "more pictures", "more DYKs").

The criteria portals use to select content should default to chronological rather than random (ie it shows the last article to be featured in that subject).

The automatic addition of DYK hooks after they've been displayed on the Main Page to the pool from which the portal draws its DYK selections.

The automatic addition to the portal's content pool of featured and quality images when they get promoted at Wikimedia Commons.

The automatic generation of an image summary for the featured pictures based on their synopsis at the Commons, but with the ability to edit and improve it if needed.

The ability to automatically pull pictures from DYK articles to be associated with their hooks on the portal.

The ability to randomize all of the individual DYK hooks instead of manually devising "blocks" of hooks.

An automatically generated list of new and recently expanded articles relevant to the subject.

Foundation sanction for direct outreach by Wikiprojects to portal-goers like offering topical reference desks, advertising within-project contests, user adoption drives, etc.
Sorry for the textwall, I just thought it was worth noting that the pro-portal camp has put forth a concrete plan for reform and I think implementing these changes would address most of the complaints people have about our currently busted system. Abyssal (talk) 01:35, 10 April 2018 (UTC)

Keep portals, and upgrade them

Oppose. Keep portals, and upgrade them – The initial design concept for portals was that they would each be a main page for a subject. The reality has been that most of them have become a snap shot (one day's version) of a main page for their respective subjects. Imagine if Wikipedia's main page never changed its content. That portals never became what they were envisioned to be is the crux of the matter here.
Looking over the problems of portals reported in the discussion, they boil down to 1) out of date / lack of maintenance (lack of volunteer labor) 2) useless (static / unchanging) and 3) low traffic (few repeat visits). These are problems we can solve. The support to do so is obvious from the above discussion.

The 3rd problem (relatively low traffic) is misleading, for two reasons. First, portals as a whole get more traffic now than ever before, with well over 20 million views per year. Second, portals get their traffic internally, rather than from external search engine results.

Please keep in mind that portals are an internal feature intended to enhance the user experience once the user is already here. Traffic is higher for those portals that provide ongoing services that users return to them for.

But, most portals do not have that level of volunteer labor available to them. Therefore, automatically-generated dynamic content, for example, in the form of randomly generated on-topic selections, automated news feeds, and so on, would be a valuable service, turning the portals into a form of periodical or newsletter.

Also, with such pages in place, who knows what enhancements could be made to them in the future. Technology is accelerating as we speak.

I believe the solution is automation, with configurability (to provide flexibility to portal designers). Refreshing the intro entry, using selective transclusion, so it doesn't go stale is one form of automation that can help.

Obviously, there is no consensus to delete. But, the message is loud and clear that the status quo is unacceptable. The portals need a lot of work. They need an upgrade, to turn them into what they were originally intended to be: main pages for their respective subjects. It's time to roll up our shirt sleeves, and get to it. I foresee a major and fun collaboration coming on. You can expect to see me there. Sincerely, The Transhumanist 22:17, 16 April 2018 (UTC)

Keep portals and make them more visible

Oppose There is a certain question about whether or not this entire RfC was poisoned based on the inadequate wording of the question. Obviously, all portals should not be deleted. There are portals that Wikipedia themselves maintain and rely on, and which are rather popular. We don't need to blow it all up just because some of the more obscure portals don't get much attention. Nonetheless, there is a question to be made about what could be done about portals that barely crack even 100 views a month (a very generous deadline imho). Portals seem to be a more niche thing that a normal viewer may not want to view but an experienced Wikipedian may visit frequently especially as a topic they're interested in. Though it seems that the very way that they're advertised is the issue. I went through some of the portals via "what links here" feature and find that they're most relegated to the last section of the page, after see also, where barely few readers make it through and it seems that by only luck you would find the link to X portal. This is unacceptable, and probably the main reason that portals are not utilized. They are but a scarce link near the last of a page, barely even a footnote, led alone enticing a reader to click them. I don't think all portals should be removed, but they should be in a more accessible, more GUI-friendly (click me to learn more about X topic) sort of thing. Tutelary (talk) 06:46, 14 April 2018 (UTC)