Wiktionary:Votes/2020-09/Removing Old English entries with wynns

Removing Old English entries with wynns
Voting on: Except for the entries for the character wynn itself (ƿ, Ƿ), all Old English entries spelled with wynn shall be
 * deleted, with users automatically redirected to the corresponding entries with w using the software (Option 1), or
 * converted to hard redirects (using #REDIRECT Target page name ) (Option 2).
 * If option 1 passes, the software will convert searches with wynn to w automatically (as is done with long s), and the linking templates will convert wynn to w as well.
 * If both options pass, the option with a greater support:oppose ratio will be implemented.

Schedule:
 * Vote starts: 00:00, 22 September 2020 (UTC)
 * Vote ends: 23:59, 21 October 2020 (UTC)
 * Vote created: - -sche (discuss) 10:59, 13 September 2020 (UTC)

Discussion:
 * [[Image:Wikt rei-artur3.svg|20px]] Beer parlour/2019/November
 * [[Image:Wikt rei-artur3.svg|20px]] Wiktionary talk:Votes/2020-09/Removing Old English entries with wynns

Support option 1 (Delete and automatically redirect)

 * 1)  —Μετάknowledge discuss/deeds 20:11, 23 September 2020 (UTC)
 * 2)  Having separate wynn pages basically doubles the amount of maintenance that needs to be done (e.g. if vowel length on a OE word needs to be rectified). There's not any benefit to having them either, given that, as far as I'm aware, very few current editions of OE texts feature wynns. Hazarasp (parlement · werkis) 23:34, 23 September 2020 (UTC)
 * 3)  --  07:31, 24 September 2020 (UTC)
 * 4) . HeliosX (talk) 21:09, 28 September 2020 (UTC)
 * 5)  - -sche (discuss) 21:16, 28 September 2020 (UTC)
 * 6)  Benwing2 (talk) 05:24, 29 September 2020 (UTC)
 * 7)  — Mnemosientje (t · c) 13:49, 4 October 2020 (UTC)
 * 8)  PUC – 18:21, 15 October 2020 (UTC)
 * 9)  J3133 (talk) 03:28, 21 October 2020 (UTC)

Oppose option 1 (Delete and automatically redirect)

 * 1)  As per support for Op2 below.  – Guitarmankev1 (talk) 13:21, 26 September 2020 (UTC)
 * 2) . May be I should have abstained, but the wynn spellings that are attested in manuscripts should not be deleted.. Delete only the unattested. Not a good vote anyway. —  inqilābī  [ inqilāb   zindabād  ] 17:45, 26 September 2020 (UTC)
 * 3) . I'm worried that deleting the pages will mean we lose information in cases where there's content on pages with wynn that isn't on the pages without wynn. See my comment in support of option 2. —Globins (yo) 04:51, 1 October 2020 (UTC)
 * This won't happen. If this option passes, I'll probably be the one implementing it, and I will make sure no information is lost, e.g. by checking the contents of each page before deleting it. Benwing2 (talk) 05:41, 1 October 2020 (UTC)
 * 1)  Dentonius (my politics | talk) 18:34, 14 October 2020 (UTC)

Abstain option 1 (Delete and automatically redirect)

 * 1)  Automatic redirects aren't reliable enough for me, but I do want Module:languages/data3/a and Module:languages/data3/e to be edited to convert ⟨ƿ⟩ in   and   links to ⟨w⟩. —Mahāgaja · talk 07:06, 24 September 2020 (UTC)
 * , is it possible to remove the js redirect timeout in special cases? -- 21:27, 28 September 2020 (UTC)
 * Yeah, I could mark the wynn links differently in MediaWiki:Noarticletext and have the JS use a shorter timeout for them specifically. It won't be instant, but it can redirect as soon as the page content and JavaScript loads. — Eru·tuon 21:04, 29 September 2020 (UTC)
 * Oh, nice. I'd be really interested to see how that looks. -- 22:03, 29 September 2020 (UTC)
 * Fast redirects are in force now. I can also implement a redirect from the edit page of wynn titles. Normally the JavaScript redirect doesn't work on the edit page, but if we decide not to have wynn entries, we will want the edit page to redirect because redlinks go to the edit page. — Eru·tuon 00:29, 16 October 2020 (UTC)
 * 1) Comment. I'm not knowledgeable about this, but surely our entries should reflect how these words were actually written in Old English, where this is known, not modern transliterations thereof. Modern transliterations or reconstructions can be included and noted as such, but the main entry(ies) should be at the spelling(s) found in actual OE manuscripts. Mihia (talk) 21:32, 14 October 2020 (UTC)
 * FWIW, I don't know that that's "surely" the case: we currently do the opposite with e.g. Old Norse (where normalized spellings are the lemmas, and manuscript spellings have only relatively recently and incompletely been included as "form of"s, since normalized spellings are what many modern editions of the texts and reference works use), and indeed modern English to some extent, where we don't include forms using long s at all even if that's the only spelling a word is attested in. (The logic here is similarly that almost all modern editions and reference works use w, and the replacement, as with long s, is 1-to-1 and applies in all cases.) - -sche (discuss) 18:19, 15 October 2020 (UTC)
 * As I understand it, long s is a variant written form of the same letter, while wynn is a different character altogether. By the same token, "normal s" is a character in modern English, so having modern English entries use it in place of the older variant is less problematic than having OE entries use w, which, and someone correct me if I am wrong, did not actually exist at that time. If OE is to be treated as a distinct language, then characters that did not exist within it should not be used in the canonical dictionary forms, IMO. Mihia (talk) 23:29, 15 October 2020 (UTC)


 * 1) Comment. No strong opinion. One of the best-known texts is Beowulf and most scholarly editions of that text use W/w. --EncycloPetey (talk) 00:22, 15 October 2020 (UTC)

Support option 2 (Hard redirect)

 * 1)  —Μετάknowledge discuss/deeds 20:11, 23 September 2020 (UTC)
 * 2)  To my knowledge, there's no meaningful distinction between this and option 1. Hazarasp (parlement · werkis) 23:34, 23 September 2020 (UTC)
 * Actually, there is (see below), but I don't think this is sufficiently better to justify rescinding my support of option 1, at least for now. Hazarasp (parlement · werkis) 01:25, 25 September 2020 (UTC)
 * 1)  Most reliable option. —Mahāgaja · talk 07:06, 24 September 2020 (UTC)
 * How so? -- 07:32, 24 September 2020 (UTC)
 * Automatic redirects don't work if you click on a red link, do they? If I click on ſnow I get taken to a creatable page which suggests both snow and Snow, but I don't get taken automatically to either of them. And if I put  directly into the URL bar, after a few moments I get taken to Snow, not snow. So I just really prefer hard redirects using the #REDIRECT command. —Mahāgaja · talk 11:45, 24 September 2020 (UTC)
 * , why would anyone ever click on such a redlink, though? Bare links like that shouldn't exist for Old English terms, and those in linking templates will be automatically converted as part of this solution. —Μετάknowledge discuss/deeds 18:57, 24 September 2020 (UTC)
 * I like the reliability of a redundant system: both hard redirects and character conversion in links. As the Germans say, doppelt hält besser. —Mahāgaja · talk 19:29, 24 September 2020 (UTC)
 * Option #1 is also redundant -- ƿorld automatically redirects to world without the need of an entry, albeit, slower. -- 05:01, 28 September 2020 (UTC)
 * 1)  Having full entry pages for every Old English word with wynn replacing W doesn't seem to add much value. However, since these forms were used and did exist, I think that pages of hard redirects is a good amount of acknowledgement of their use. So when a user searches for "ƿesan" and lands on the page for "wesan", but they don't understand that ƿ and W are functionally equivalent, the little "**redirected from ƿesan**" will reassure them of the validity of the "ƿesan" alternative spelling.  – Guitarmankev1 (talk) 13:21, 26 September 2020 (UTC)
 * When someone searches for wynn form, it already returns a w-form. Try searching for ƿorld. -- 05:01, 28 September 2020 (UTC)
 * Right, when using a wynn but no page exists with the wynn form, the software will automatically read the wynn as a W and direct to that page instead if there is one, even if that word would have never used the wynn form. Try searching for "beƿilderment" and you'll see this is the case. The difference with hard redirects is that a (Redirected from "xxx") notification will pop up at the top of the page, which may show that the searched form might have some validity. This way, you'll get the redirected notice when you search for a valid term like ƿesten, but not when you search for an invalid form like beƿilderment. – Guitarmankev1 (talk) 12:46, 28 September 2020 (UTC)
 * 1)  فين أخاي (talk) 02:47, 27 September 2020 (UTC)
 * 2)  (though I, too, would prefer option 1, and the preferences of users who've said they prefer one option or another should probably be taken into account so people don't have to try to strategically switch votes at the last minute) - -sche (discuss) 21:16, 28 September 2020 (UTC)
 * Sounds like we need to enact a system. --  04:35, 17 October 2020 (UTC)
 * support vote struck Benwing2 (talk) 05:24, 29 September 2020 (UTC)
 * If so, you should oppose option #2. -- 05:33, 1 October 2020 (UTC)
 * Last chance to downvote option #2 for option #1 to win out. -- 16:10, 14 October 2020 (UTC)
 * 1) . I don't know what the plan is for how this will work, but I'm worried about there being cases where pages with wynns have more information on the word than pages without wynns. If this is done manually, it'll decrease the chances of losing information. —Globins (yo) 04:51, 1 October 2020 (UTC)
 * 2)  I agree with some of the previous comments. I think it would be quite helpful for users copy-and-pasting terms with wynns to be able to see that they had been redirected from their search input. I'm not knowledgeable about this terminology, but that seems to me to be the case here Hk5183 (talk) 20:28, 14 October 2020 (UTC)
 * 3)  opposed this initially out of a slight preference for option #1, but would rather see this pass than both fail. — Mnemosientje (t · c) 10:46, 15 October 2020 (UTC)
 * 4) . In addition to aforesaid arguments, some languages have romanized entries in spite of being attested with other writing system(s) (like Tocharian A and B). Thus, W-written forms are not that "shocking". We need one entry for one word, and wynn-using forms being put as alternative forms (only "redirected from" isn't enough) provided that it is attested (with a reference to make it reliable). With W-forms only, we ensure not to display actually unattested wynn-forms. All others little but useful features are already propounded above, like for instance the information-loss avoidance, W-forms proposed when typing in search bar, no unnecessary pages (whence unification of the data on a single page and a gain of place, since one generally doesn't type the wynn character), and so on. Malku H₂n̥rés (talk) 14:11, 18 October 2020 (UTC)

Oppose option 2 (Hard redirect)

 * 1) . The aforesaid. —  inqilābī  [ inqilāb   zindabād  ] 17:46, 26 September 2020 (UTC)
 * 2)  Hard-redirecting is already automatically done through javascript on the 404 page. --  05:01, 28 September 2020 (UTC)
 * 3) . I have to join this in order to avoid innecessary pages. HeliosX (talk) 21:09, 28 September 2020 (UTC)
 * 4)  Dentonius (my politics | talk) 18:34, 14 October 2020 (UTC)
 * oppose vote struck — Mnemosientje (t · c) 13:49, 4 October 2020 (UTC)
 * as things stand, both options are set to fail (with each one having less than 2/3s support), leaving us with the status quo of nearly-complete entries with language and part-of-speech headers and sometimes differentiated etymology or pronunciation headers for wynn entries, as at ƿesan. If any one of you switched to also supporting this option (and not just option 1), we could at least reduce that level of duplication down to hard redirects, even if that's not quite your preferred option of no entries at all. - -sche (discuss) 03:07, 15 October 2020 (UTC)
 * I'd rather it fail in no-consensus and start a second vote than vote for option #2. -- 03:25, 15 October 2020 (UTC)
 * What option would you propose in a second vote? - -sche (discuss) 04:27, 15 October 2020 (UTC)
 * Option #1 or bust. We'll see how this vote pans out. -- 06:46, 15 October 2020 (UTC)
 * 1) . I am not a fan of redirects at all, as presently implemented (albeit I may have been guilty of creating one or two myself in the past), as it may not be clear to the user why they are being given something different from what they typed, or what the connection between the two is. Contrary perhaps to some opinions above, I personally don't think that the "redirected from ~" notice is enough (even if people notice it). I know sometimes at Wikipedia I type in a term that I don't know the meaning of, and I am redirected to an article with a different title, and sometimes the term I typed in is not even mentioned anywhere in that article, so then I am just confused. Are the two things exact synonyms? Just related in some way? Related how? Mihia (talk) 19:40, 16 October 2020 (UTC)
 * 2)  If the software redirects automatically then creating these redirects is unneeded. J3133 (talk) 03:28, 21 October 2020 (UTC)
 * 3)  I prefer option #1. Benwing2 (talk) 03:54, 21 October 2020 (UTC)

Decision

 * Option 1: passes 9–4 (69%).
 * Option 2: no consensus 10–7 (59%). —Mahāgaja · talk 09:18, 29 October 2020 (UTC)