Wikisource:Requests for comment/Redirect policy
Discussion
[edit]- Discussion started on Wikisource Scriptorium
I think we should amend our policy on redirects. Currently we delete almost all redirects, especially in the main namespace, as unneeded. This generally occurs when we move pages. However, this may be causing a lot of broken links and general linkrot; potentially causing a lot of reputational damage to Wikisource. People following a link from a forum, blog post, reddit, tweet, or whatever should reasonably expect to find content at the other end. What they will often get if they use a pre-move URL is an error message. Most people arriving from off site wouldn't think to search, they just assume it was deleted or that our system is falling apart. We are making ourselves look bad for no good reason. I thought about this when we started migrating works to the Translation namespace (Catullus 16 is one of our more popular works) and I've read up on this a bit. Tim Berners-Lee himself has something to say about this. There is no need to ever delete any page. We are not going to run out of space (if anything, deletion uses more memory as the deletion action needs to be recorded too) and it doesn't obscure searches (redirects are generally omitted as standard). The redirects can happily sit in the background, allowing the project to run smoothly without ever getting in anyone's way. I think we should change our policy and stop deleting redirects, unless there is a strong reason to do so (for instance, it would still remain acceptable for very new pages or inaccurate/bad faith redirects). - AdamBMorgan (talk) 12:24, 15 November 2013 (UTC)
- Comment - With all due respect, the proposal makes claims of one thing or another with no factual basis given to support such claim(s). I've seen no evidence of 'lost traffic' due to broken-links, link-rot or anything else link-related - but that claim is just as unsupported as the proposal's. Yet it is one based on my limited personal experience in investigating WS 'traffic' to date (page views in particular) so I'm comfortable in making it here. Regardless, without some actual statistical evidence of the actual number of first click-ins landing on now deleted redirect pages or similar, I challenge the proposal's assertion that the current practice is indeed adversely impacting the total possible number of en.WS visitors as a entirely valid one.
The claim users landing on an once-existing re-direct only to find an "error message" is also not entirely the case. Even when I'm logged out and manually input an URL to a 2011 deleted redirect ( https://en.wikisource.org/wiki/Statement_by_Prime_Minister_Tomiichi_Murayama ) I still get both the log entry for the target of the now deleted re-direct as well as the deletion log entry for the removal of said re-direct. So again, the assertion readers need to initiate a separate action (a search) to locate the desired content isn't exactly a full picture of the realities at hand. A careful reading of the given log entries frequently provides the target URL of the now deleted redirect entry.
In closing, I'm opposed to the proposal in its current form without tangible evidence to support its assertion(s). I am open, however, to narrowing the scope and/or instances where deletions of redirects are no longer recommended practices if that helps any. -- George Orwell III (talk) 05:21, 16 November 2013 (UTC)
- I don't think it's a good idea to expect visitors to be able to understand the deletion log message. Web design frequently emphases the skittishness of your average internet user--if they can't get what they came for immediately, they'll back out and go somewhere else. Prosody (talk) 05:45, 16 November 2013 (UTC)
- Point taken but I can't fix stupid nor force anybody to read carefully - the message further down from the log entries, however, does a fairly good job of explaining things. The part about "Was the text deleted? Occasionally, texts have to be deleted. This usually happens because they infringe copyright, or violate our inclusion guidelines. Check the deletion log (emphasis mine) for deleted pages. " seems applicable in the case of a deleted redirect. I'm all for Users landing where they expect to but I just don't believe deleted re-directs are the primary factor in those cases (at least not without some 'proof'). -- George Orwell III (talk) 06:03, 16 November 2013 (UTC)
- I have not seen any evidence of this in action; I only anticipated it from the Translation migration. After looking into it, keeping redirects is best practice in the internet and doesn't cost us much, if anything. I think we can better accommodate casual readers who are not familiar with our policies. - AdamBMorgan (talk) 16:05, 17 November 2013 (UTC)
- Point taken but I can't fix stupid nor force anybody to read carefully - the message further down from the log entries, however, does a fairly good job of explaining things. The part about "Was the text deleted? Occasionally, texts have to be deleted. This usually happens because they infringe copyright, or violate our inclusion guidelines. Check the deletion log (emphasis mine) for deleted pages. " seems applicable in the case of a deleted redirect. I'm all for Users landing where they expect to but I just don't believe deleted re-directs are the primary factor in those cases (at least not without some 'proof'). -- George Orwell III (talk) 06:03, 16 November 2013 (UTC)
- I don't think it's a good idea to expect visitors to be able to understand the deletion log message. Web design frequently emphases the skittishness of your average internet user--if they can't get what they came for immediately, they'll back out and go somewhere else. Prosody (talk) 05:45, 16 November 2013 (UTC)
- Sound argument, support JeepdaySock (AKA, Jeepday) 15:26, 15 November 2013 (UTC)
I am the one deleting most of these redirects. The issue is, when we disambiguate a work with twenty chapters, we move the entire subpage structure — thus creating numerous redirects from subpage to subpage; and then we turn the base page into a disambiguation page. Thus, for example, The Europeans becomes a disambiguation page, but The Europeans/Volume 1/Chapter 1 becomes a redirect to a particular chapter of a particular volume of a particular version or disambiguated work. If that subpage has any right to exist at all, it should be a disambiguation page, not a redirect. After all, "The Europeans/Volume 1/Chapter 1" is an ambiguous title. Would anyone like to volunteer to create disambiguation pages for every single chapter of every single disambiguation page? No? I didn't think so. Therefore they should be deleted.
I am of the view that there are no valid reasons for us to retain our massive mess of invalid redirects in subpage space. Thus I go around deleting them. I never delete one of these redirects if it has internal incoming links. But I don't think it is feasible to extend that policy to external incoming links — which we are powerless to fix. I do feel very strongly that this is an awful mess that should be routinely cleaned up, and I don't have a problem with that cleanup occasionally breaking the rare deeplink from elsewhere in the internet. Hesperian 07:40, 16 November 2013 (UTC)
- I share Hesperian's view. Let me add that useless redirects are also a burden when searching for "start with prefix" pages, for a clean-up work of a bot.--Mpaa (talk) 09:04, 16 November 2013 (UTC)
- My principle thought is with the root pages. I think we probably can keep subpage redirects without much impact but the root pages were my main concern. I'm not really concerned with talk pages at all, although technically the practice would apply to them as well. - AdamBMorgan (talk) 16:05, 17 November 2013 (UTC)
- If you're talking about root pages then I think I would have to dispute the original premise that we delete nearly all redirects as unneeded. We have tens of thousands of root page redirects! Hesperian 22:07, 19 November 2013 (UTC)
- That is also inline with my observations. I don't believe there is any substantial practice of frivolous deletions affecting root-page, mainspace, redirected title - nor to any number of existing redirects that might already be pointing to a target title. Off the top of my head, the only case where I'd delete a mainspace redirect is when its an obvious mis-spelling that was moved at some point just to correct it (& not the kind where its American spelling vs. Traditional English spelling either). The other case would be for DAB purposes - after that, I just don't see any evidence that this proposal's claim is actually an issue; possible or otherwise. Garbage mainspace, root redirects aren't that prevalent in my travels as well. -- George Orwell III (talk) 00:54, 20 November 2013 (UTC)
- If you're talking about root pages then I think I would have to dispute the original premise that we delete nearly all redirects as unneeded. We have tens of thousands of root page redirects! Hesperian 22:07, 19 November 2013 (UTC)
- My principle thought is with the root pages. I think we probably can keep subpage redirects without much impact but the root pages were my main concern. I'm not really concerned with talk pages at all, although technically the practice would apply to them as well. - AdamBMorgan (talk) 16:05, 17 November 2013 (UTC)
This is already getting to element of TL;DR, on an otherwise long page.
There is already evidence and strong opinions presented on what should stay and what should go. I would prefer that we split this discussion to a separate page (as close as we get to an RFC), structure the argument so that comment can be made and determinations can be made contextually.
There are three decision groups when faced with the decision on a redirect 1) clearly delete, 2) clearly keep, and 3) let us think about it, so let us separate to groups, and then argue over the middle group, and give some guidance. To me half the argument is not about the redirects but what we provide to people who land on a 404 page to give them some sensible result that helps them find the page of interest. To me, presenting a short search result, and the ability for the user to get a larger search result seems to cover those that land on mislinks or deleted redirects. — billinghurst sDrewth 23:43, 16 November 2013 (UTC)
- If there is a better way to solve the same potential problem then I would like to hear about it. Can we change that page into something more useful (to, I stress, a casual user, possibly one following an interesting link from a years-old forum post or whatever). - AdamBMorgan (talk) 16:05, 17 November 2013 (UTC)