Wikipedia:Language order poll

This page is for conducting a Wikipedia:poll about the desired language order of multilingual lists. There seem to be several options, and different ones are used in different cases. See: meta:Interwiki sorting order for more info and Help:Interlanguage_links#Sorting for the current guideline.

General orderEdit


Updated November 8, 2011
By order of... Count %
number of speakers 5 3.2
native language population 0 0.0
article count 3 1.9
English alphabet 2 1.3
alphabet, based on local language 58 36.9
alphabet, based on two letter code 68 43.3
usage pattern in local language 2 1.3
alphabet in the users language 1 0.6
Alphabetical in English 4 2.5
Two-letter code in source, Alphabetical in view 14 8.9
Total 157 100.0

Best of both worlds and "Two-letter code" are actually quite similar since there is no extra sorting script available by now.

By order of number of speakersEdit

Worldwide, regardless of native language. Currently: English, Chinese, Spanish, Russian, Arabic, estimated order:Portuguese, Japanese, German, Korean, Italian

  1. 戴眩sv 21:02, 6 Feb 2004 (UTC)
  2. James F. (talk) 05:44, 13 Feb 2004 (UTC)
  3. Juppiter 04:02, 21 Mar 2005 (UTC)
  4. Round Earther 10:36, 31 July 2005 (UTC)
  5. Megapixie 05:28, 11 August 2005 (UTC)

By order of native language populationEdit

By order of article countEdit

Language Wikipedias by size

  1. Sara 02:11, 9 Feb 2004 (UTC)
  2. --Milaiklainim 08:55, 21 Mar 2004 (UTC)
  3. Fangfufu (talk) 22:57, 22 December 2008 (UTC)

By order of English alphabetEdit

  1. Dante Alighieri | Talk 20:32, Apr 22, 2004 (UTC) I support, nigh on equally, the two below options
  2. Exploding Boy 00:57, August 12, 2005 (UTC)

By order of alphabet, based on local name of languageEdit

e.g. Español (es:) before Esperanto (eo:) and Suomi (fi:) right before Svenska (sv:)

  1. Angela. 22:50, Feb 7, 2004 (UTC)
  2. Michael Snow (ordering by size feels like POV)
  3. Jmabel 08:23, 13 Feb 2004 (UTC) (ordering by size feels like too much thought required when adding, error-prone)
  4. Docu (for Wikipedia:Interlanguage_links, as it is currently defined in the pywikipediabot
  5. Jeandré 2004-02-25t09:47z
  6. Secretlondon 19:00, Feb 28, 2004 (UTC)
  7. Ryan_Cable 14:47, 7 Mar 2004 (UTC)
  8. webkid 22:12, 12 Mar 2004 (UTC)
  9. Kpalion 13:34, 13 Mar 2004 (UTC) No problem to find a particular language in a long list
  10. Melkom 08:47, 20 Mar 2004 (UTC)
  11. +sj+ 03:16, 2004 Mar 21 (UTC)
  12. Lakefall 21:31, 9 Apr 2004 (UTC) Most obvious for the reader. (Easy to find a language in a long list.) (Only for lists of languages by local name such as Wikipedia:Interlanguage links.)
  13. Timtzeptel 08:07, 10 Apr 2004 (UTC)
  14. Stan 05:44, 11 Apr 2004 (UTC) (for lists of languages by name)
  15. Kaare 10:34, 13 Apr 2004 (UTC) The obvious choice (to me)
  16. Beatnick 08:06, 5 May 2004 (UTC) Easier to understand than ordering by two letter code. Harder to sort but bots are our friends (see Docu's remark)
  17. AquaRichy 08:17, 4 Jun 2004 (UTC)
  18. Jonik 16:10, 26 Jul 2004 (UTC) This is the clearest way for the reader. Also remember that nl: (Nederlands) goes before ja: ("Nihongo"), etc. See e.g. this for all the language names in latin alphabet (I haven't found such a list here in en-wikipedia.) Yes, it's more difficult to sort the links this way, but all this interwiki stuff should, of course, be (better) automated anyway...
  19. Gabbe 14:20, Aug 9, 2004 (UTC)
  20. Hapsiainen 21:34, 2 Sep 2004 (UTC) The link lists like "Esperanto Español Eesti Suomi French" are awkward for the reader.
  21. Jpk 11:18, 21 Sep 2004 (UTC)
  22. David.Monniaux 06:57, 21 Apr 2004 (UTC)
  23. ChristianGlaeser 9:47 pm, 27.04.04 EST
  24. Lussmu 18:35, 26 Jul 2004 (UTC) This one, but only if it's done in software.
  25. VerdLanco 19:47, 15 Nov 2004 (UTC)
  26. Rje 16:07, Feb 14, 2005 (UTC). Although we would still need to have a policy for non-Latin alphabet languages.
  27. Sounds like a good idea. JuntungWu 14:15, 15 Feb 2005 (UTC)
  28. Trilobite
  29. Mononoke 18:32, 20 Feb 2005 (UTC)
  30. BlankVerse 10:34, 16 Mar 2005 (UTC)
  31. ··gracefool | 02:49, 28 Mar 2005 (UTC)
  32. Marsian // talk 09:04, 2005 Apr 4 (UTC) : I agree with Jonik. This is probably a little bit more complicated, but eventually we'll get this way somehow.
  33. Jannex 13:54, Apr 25, 2005 (UTC)
  34. Marianocecowski 09:59, 9 May 2005 (UTC)
  35. Most logical idea, IMHO. Linuxbeak | Desk 20:13, May 15, 2005 (UTC)
  36. Craigy   (talk) 04:14, Jun 23, 2005 (UTC)
  37. Kaldari 6 July 2005 18:44 (UTC) - (Moved from the ===Comment=== section. If wrong, please revert. Marsian // talk 7 July 2005 02:59 (UTC))
  38. Certainly the simplest for the reader, and who else do we care for? KissL 10:39, 28 July 2005 (UTC)
  39. TarmoK 19:50, 2 August 2005 (UTC) Is wikipedia for writers or readers? :o) ....we have different tools for settings things right if writer doesn't know exact place
  40. Instantnood 17:17, August 30, 2005 (UTC) My second choice after by alphabetical order in English.
  41. Noisy | Talk 15:28, 13 November 2005 (UTC). Docu's comment above re bot action seems to make this a pointless poll.
  42. Mardus 13:11, 7 February 2006 (UTC) /For usability's sake. Is it possible to add a vote for more than one resolution (at a time)? You can reply to this one in my Talk page. Because if we were to sort them by their country codes, those codes would have to be visible to everyone else in text, but that seems superfluous.
  43. Eddi (Talk) 16:31, 5 March 2006 (UTC) For reader friendliness, not editor friendliness, at least while automatic sorting isn't implemented yet. And the poll is important as both editors and robots would have to comply if it were to become policy.
  44. Alensha 19:43, 9 March 2006 (UTC) This is the best for readers. If we don't care about the readers, then 1) we're not a good encyclopedia, 2) we might as well leave out the wikilinks since people can type it in the searchbox too. For languages with non-Latin alphabets we could have a list like the Finnish have.
  45. Robots can do the ordering, it needs to be determined only once. -- nyenyec  19:57, 9 March 2006 (UTC)
  46. I absolutely agree. Why should we confuse poor readers with the inner operation of the software? Adam78 13:39, 11 March 2006 (UTC)
  47. As this is the best-looking on the page ('Suomi' looks odd sandwiched between 'Farsi' and 'French') and closest to current commonest usage (so involving the least change) - MPF 13:26, 21 March 2006 (UTC)
  48. Two years, and still no software support? Make *pedia for readers. --William Allen Simpson 07:55, 22 March 2006 (UTC)
  49. This way seems to most logical and easy to navigate in for the readers. Who the hell knows the language ISO codes well enough to easily find his language in a long list? Jeltz talk 11:39, 30 March 2006 (UTC)
  50. Most logical, though the ISO proposal is also valid; even though I usually prefer ISO codes, it is true that we can't expect everyone who's looking for a certain language to know its ISO code. —Nightstallion (?) 20:12, 21 March 2007 (UTC)
  51. Robin Stocker 17:08, 11 October 2007 (UTC) List looks better and is easier to search.
  52. I'm thinking for the readers' best, and this one's certainly the simplest for the reader. There is no sense in listing the languages like this: "Esperanto Español Eesti Suomi French". --Dogah (talk) 17:44, 26 February 2008 (UTC)
  53. Yes. Siliamaav (talk) 22:29, 26 February 2008 (UTC)
  54. Most logical. Just transliterate other alphabets. Bart133 (t) (c) 00:33, 2 July 2008 (UTC)
  55. The Jade Knight (talk) 09:39, 9 September 2008 (UTC)
  56. It looks best. BartekChom (talk) 19:41, 30 October 2008 (UTC)
  57. Most user friendly for people just reading articles. Most interwiki links are added by bot so no problem if it can be a little confusing for a person contributing to find the right position to place a interwiki link. Kinamand (talk) 20:16, 6 February 2009 (UTC)
  58. Oda Mari (talk) 14:31, 3 February 2010 (UTC)


What's the collating order for this? The notion of "alphabetical" order is different from country to country. Do you sort éèê as e, or separately?
--David.Monniaux 06:57, 21 Apr 2004 (UTC)
é,è, ê are sorted as e, but ñ comes after n. Question is, what with Japanese (Chinese, etc) symbols? --Marianocecowski 09:59, 9 May 2005 (UTC)
The currently proposed order of ja: is as "N" from "Nihongo", which is the pronunciation of 日本語(Japanese) in Japanese. For other languages, please see fi:Wikipedia:Luettelo Wikipedian kielistä(as Jonik has pointed). From the point of the appearance on the side bar, putting kanji-written langs bottom might be an option? But in that case, I don't know about arabic alphabets or kind of that... (I guess this might imply that using alphabet for ordering itself has a PoV aspect, but still, I feel slightly better than 2-letters) - Marsian // talk 7 July 2005 02:59 (UTC)

I think there is confusion about the phrase "local language". Does it mean local to the wiki, or to the language name? I would think that the English wikipedia should have the names in English - and other languages should use their own names for English etc.. Otherwise the en wikipedia (for example) becomes unitelligable to some English speakers. Rich Farmbrough 22:10 25 March 2006 (UTC).

It's explained so there shouldn't be any confusion "e.g. Español before Esperanto and Suomi right before Svenska". Jeltz talk 11:51, 30 March 2006 (UTC)
No confusion for editors perhaps. But for readers it's not obvious that "Anglo-Saxon" will be after Greek and before English. Rich Farmbrough 16:52 30 April 2006 (UTC).
Ah, misunderstod your message. I think the opposit actually. This system would be more obvious for the reader and less obvious for the editor. It's not obvious that Irish is ga while Finish is fi. ga is native spelling while fi is not. One option would be if the names where in the language of the Wikipedia you currently are at (i.e. names in English at English Wikipedia) but that could create new problems. I'm not against that option but I really dislike ordering by ISO 639 codes since that is just maiking it easy for the editors without thinking of the readers. You can't expect the readers to know the ISO-codes. Jeltz talk 21:41, 30 April 2006 (UTC)

The fallacy here is for the many languages that don't use a Latin alphabet. Unless the suggestion is Latin in alphabetical order, followed by other alphabets... The two letter language code is clearer than all of that. An objective standard is better. Dovi 11:46, 28 June 2007 (UTC)

By order of alphabet, based on two letter codeEdit

e.g. Esperanto (eo:) before Español (es:) and Suomi (fi:) somewhere before Français (fr:)

  1. Jmabel 08:23, 13 Feb 2004 (UTC) I'd be just as happy with this
  2. Dori - easiest way by far (regarding interwiki links, poll not to clear)
  3. Danny 22:10, 14 Feb 2004 (UTC)
  4. Tuf-Kat
  5. andy - as the easiest way
  6. Andres 08:44, 20 Mar 2004 (UTC)
  7. Arwel 19:52, 9 Apr 2004 (UTC) - easiest way
  8. Mustafaa 09:19, 10 Apr 2004 (UTC) - by far the easiest, and uniform across wikis.
  9. Stan 05:44, 11 Apr 2004 (UTC) (for Wikipedia:Interlanguage links)
  10. Henrygb 16:23, 14 Apr 2004 (UTC) - where would Chinese/Japanese/Korean/Arabic/Hebrew/Thai etc. fit in on local language?
  11. Jonel 22:44, 14 Apr 2004 (UTC) Makes sense to use the abbreviation used for all the pages
  12. Zeitgeist 20:02, 27 Apr 2004 (UTC) It's the easiest to handle.
  13. Jusjih 08:22, 28 Apr 2004 (UTC) It is the easiest for editors to alphabetize. It should be uniform in all versions.
  14. Woggly 06:54, 23 May 2004 (UTC) simplest way.
  15. pne 12:32, 11 Jun 2004 (UTC)
  16. Andre Engels 23:15, 21 Jul 2004 (UTC) - simplest and most common in the various Wikipedia languages
  17. Dittaeva 11:34, 24 Jul 2004 (UTC) - agree with Andre, but this should be in source, the software should order it by article count on view.
  18. Yekrats 14:36, 25 Nov 2004 (UTC) Otherwise, how would we order languages which do not use a Latin alphabet?
  19. eo:Montanesko
  20. Harriv 22:14, 7 Jan 2005 (UTC)
  21. Ejrrjs | What? 21:38, 20 Feb 2005 (UTC)
  22. Tony Sidaway|Talk 17:50, 23 Feb 2005 (UTC)
  23. CF 09:36, 16 Mar 2005 (UTC)
  24. Jon Harald Søby 19:46, 24 Mar 2005 (UTC)
  25. Lysy 07:43, 2 Apr 2005 (UTC)
  26. r3m0t talk 12:46, Apr 3, 2005 (UTC)
  27. missmarple 18:22, 20 Apr (CET) anyone that has ever added iw-s knows that this is by far the most simple way
  28. Deco 05:50, 4 May 2005 (UTC) Because it's the least prone to human error. However, it's much better to have the software reorder them appropriately, since this can be done globally, can be more easily changed globally, and doesn't depend on humans not screwing up. Deco 05:50, 4 May 2005 (UTC)
  29. Palnatoke 11:13, 1 Jun 2005 (UTC) agree with Deco and Yekrats
  30. Manscher 07:20, 2 Jun 2005 (UTC) I agree with Woggly, Yekrats, Missmarple and Deco
  31. JFW | T@lk 01:21, 6 Jun 2005 (UTC) Least confusing
  32. --TheGrappler 14:41, 17 Jun 2005 (UTC) Not only what would we do with Hebrew, Chinese etc. but if this was consistent across all wikipedias, it would be very easy for anybody to update the interlanguage links anywhere (rather than having to learn the order that the languages would appear in in the other wikis)
  33. grubber 11:17, 2005 Jun 22 (UTC) Easiest to maintain
  34. billlund 04:19, Jun 23, 2005 (UTC) Not the easiest to search, but less ambiguity
  35. Denelson83 04:21, 23 Jun 2005 (UTC)
  36. --Urod 3 July 2005 00:26 (UTC)
  37. Li-sung 5 July 2005 20:13 (UTC)
  38. Node 8 July 2005 02:48 (UTC)
  39. Radiant_>|< 14:30, July 18, 2005 (UTC) it's easiest; as an alternative I wouldn't mind 'order of alphabet based on local language' or on English name, but I prefer this.
  40. Pekinensis 15:32, 27 July 2005 (UTC) simple and easy. I don't believe that any sort order will be obvious to the user, so there's no point in complicating our task.
  41. --Agari 19:02, July 28, 2005 (UTC)
  42. By far the easiest, and I think also the most common. Rd232 12:27, 3 August 2005 (UTC)
  43. -shuri 17:52, 9 August 2005 (UTC)
  44. Sarge Baldy 08:43, 10 August 2005 (UTC) I do a lot of interlanguage work, and this is by far most convenient.
  45. Phil | Talk 09:18, August 10, 2005 (UTC) Less room for argument, as I have commented elsewhere (This is to avoid collating problems with non-Latin scripts: should they all be sorted to the bottom? do we assume that people just know that "فارسی" means "Farsi"? "中文" is sorted to the end because its code is "zh": I don't know enough to say whether that is Cantonese, Mandarin or something else, but it looks kind of odd using different collation styles for some links.) HTH
  46. Gene Nygaard 11:50, 10 August 2005 (UTC) The only sensible way. It is totally unrealistic to expect users to learn and know some "alphabetic" nonsense when some of the names are displayed in other alphabets and syllabaries and whatever.
  47. Huaiwei 16:38, 13 August 2005 (UTC) Agree that this is far simpler, and helps to raise less questions over non-alphabet languages like Chinese.
  48. ᓛᖁ  03:28, 14 August 2005 (UTC)
  49. Trevor MacInnis 05:11, 14 August 2005 (UTC) easiest, and impartial.
  50. Head 02:47, August 15, 2005 (UTC) KISS
  51. But amenable to other ideas in the longer term. Rich Farmbrough 14:25 20 March 2006 (UTC).
  52. *drew 08:18, 24 March 2006 (UTC)
  53. Y Ynhockey (Talk) Y 10:48, 30 March 2006 (UTC)
  54. no problems with non-alphabet languages. no article cout or speaker estimation needed. system could be used in all wikis. Tobias Conradi (Talk) 15:09, 24 June 2006 (UTC)
  55. Wim van Dorst (Talk) 20:20, 20 July 2006 (UTC)
  56. AZ t 19:06, 22 July 2006 (UTC)
  57. Yath 13:25, 6 August 2006 (UTC) - though there is no need to say "two letter code" as it is perfectly simple to sort strings of different lengths.
  58. Kazubon 09:53, 18 March 2007 (UTC) - Simple and easy.
  59. Dovi 17:07, 11 April 2007 (UTC) The only option that makes sense in a multilingual project, especially given the fact that "local names" of numerous languages are not written in Latin letters at all. Functional, objective & unambiguous. Also especially intuitive given the fact the the project domains are already based on these codes precisely.
  60. JonnyJD 22:13, 5 May 2007 (UTC) - The only way without an integrated ordering script. The ordering should be the same in every wiki.
  61. Otto 17:03, 26 May 2007 (UTC) See comment from missmarple.
  62. Skullers (talk) 04:23, 20 December 2007 (UTC)
  63. Hanfresco Hanfresco (talk) 10:13, 18 January 2008 (UTC) - Makes the most sense to me.
  64. Bender235 (talk) 14:15, 21 July 2008 (UTC)
  65. Ency response? 22:41, 29 October 2008 (UTC) it's simple,usefull, obvious, however non political correct :-))
  66. Deror (talk) 08:53, 13 March 2009 (UTC)
  67. gidonb (talk) 15:46, 28 May 2009 (UTC) most simple to apply *leads to* fewest mistakes *leads to* most consistent usage across articles and Wikipedias (although for non-English-language Wikipedias I also advise to put English on top. No, I am not a native speaker of English).
  68. ترجمان05 (talk) 15:08, 30 December 2010 (UTC)

By order of usage pattern in local languageEdit

We should study which links are actually clicked on by the users, to determine which ones they most want to use, and order at least the first few that way, for ease of use.

  1. Jimbo Wales 19:58, 20 Jul 2004 (UTC)
  2. ··gracefool | 07:08, 24 Jun 2005 (UTC)

By order of the alphabet in the users languageEdit

We ask the users browser what's her preferred language, provide all the language names in that language and order them in whatever way the writers of that language generally order language names. (Yes, this probably requires changes in the software.)

  1. If possible. The Jade Knight (talk) 09:30, 9 September 2008 (UTC)

Alphabetical order, according to language name in EnglishEdit

e.g. Chinese, German, Spanish (in that order) rather than Deutsch, Español, Zhongwen (in that order).

  1. David Cannon 22:43, 25 Jul 2004 (UTC) As this is the English Wikipedia, I believe it would be best to identify other languages by their English names, and to alphabetize them accordingly.
  2. This is easily the most efficient. It's simple and obvious to users and it implies no value judgment. Maurreen 18:43, 13 Feb 2005 (UTC)
  3. Instantnood 16:03, August 13, 2005 (UTC)
  4. Rschmertz 06:38, 28 February 2006 (UTC) I support this if it means using the English words in the actual list, ideally followed by the local word for the language in parens.


I don't think anyone's suggesting forcing (or even encouraging) Japanese Wikipedians to follow our rules. Trouble is, both section titles, even with examples, are a little ambiguous. The #By order of alphabet, based on local language description seems to say something a little different what you seem to think: sort based on the English alphabetic order of the local names of the language, when we're talking about the English Wikipedia. Said section says nothing about how to handle languages written in non-Latin alphabets.
I read the description of 'this' section as implying that that Wpedias in other languages (say, Ossetian) will, of course, follow the sorting of the relevant local names for the languages (in this case, the Ossetian words for "Japanese", "Lithuanian" and "Ojibwe", all in Cyrillic, since this is the usual alphabet for Ossetian). --Rschmertz 05:49, 28 February 2006 (UTC)
I think this discussion should only apply to the English Wikipedia. Other languages don't have to follow the same sorting rule. If they'd rather have the language names appear in those respective languages, they should be allowed to have it that way. And they should discuss this topic on their domain, perhaps in their language. -- Unregistered User 21:50, 13 September 2007 (UTC)

Best of both worlds: Two-letter code in source, Alphabetical in viewEdit

When editing pages, enter interlanguage links in (English) alphabetical order by two-letter language code (eo, es, fi, fr...). For now, let them appear in that order. We also request a future software change so they may soon appear in order of alphabet, based on local language (Suomi, Svenska...). (Option inspired by comments below.)

  1. Krubo 00:41, 4 May 2005 (UTC)
  2. Palnatoke 10:54, 28 July 2005 (UTC)
  3. ChongDae 09:43, 29 July 2005 (UTC) -- Two-letter code in source, user's choice (alphabetical order/two-letter order/any other order) in view
  4. Echo Chongdae - user choice always a good thing. NB I voted for two-letter codes above, which I see as complementary. Rd232 12:30, 3 August 2005 (UTC)
  5. Sarge Baldy 02:03, 17 August 2005 (UTC) This would be fine by me as well, as long as I could optionally have them displayed in order of the two-letter code for myself.
  6. I like this on as well. Rich Farmbrough 14:27 20 March 2006 (UTC).
  7. This is definately the obvious way of doing it. I was very surprised the software didn't already do this. Carl Kenner 03:12, 1 May 2006 (UTC)
  8. It would be optimal if Mediawiki did the sorting so that it always is the same on all pages. Jeltz talk 21:38, 1 May 2006 (UTC)
  9. This sounds good, as long as it's possible to implement without any problems. Ardric47 07:44, 8 May 2006 (UTC)
  10. Rigadoun 19:42, 17 June 2006 (UTC)
  11. LX (talk, contribs) 13:08, 18 July 2006 (UTC)
  12. The Jade Knight (talk) 09:29, 9 September 2008 (UTC)
  13. NaBUru38 (talk) 16:25, 11 February 2009 (UTC)
  14. Hluup (talk) 10:18, 20 March 2010 (UTC)
  15. Hyacinth (talk) 13:49, 18 April 2013 (UTC)

Simple English in EnglishEdit

Updated 21:14, 20 July 2006 (UTC)
By order of... Count
Alphabetize 13
List at the end 6
Don't standardise order 2


e.g. between Shqip and Slovenian

  1. Angela
  2. Jusjih 08:23, 28 Apr 2004 (UTC)
  3. pne 12:31, 11 Jun 2004 (UTC)
  4. Davidcannon 22:45, 25 Jul 2004 (UTC) - but only until such time as the "simple" wikis are established in other languages. When that time comes, we should list all simple wikis together, at the end.
  5. eo:Montanesko David Cannon is right.
  6. Jon Harald Søby 19:45, 24 Mar 2005 (UTC)
  7. --TheGrappler 14:44, 17 Jun 2005 (UTC) agree with David Cannon
  8. 500LL 08:16, July 18, 2005 (UTC)
  9. Easiest. Or, list it under "English, Simple". Radiant_>|< 14:32, July 18, 2005 (UTC)
  10. Sarge Baldy 08:44, 10 August 2005 (UTC)
  11. ᓛᖁ  03:30, 14 August 2005 (UTC)
  12. Trevor MacInnis 05:12, 14 August 2005 (UTC)
  13. Wim van Dorst (Talk) 21:14, 20 July 2006 (UTC)
  14. English, Simple. Failing that, alphabetise normally. —Nightstallion (?) 17:13, 22 March 2007 (UTC)

List at the endEdit

e.g. after Zhongwen.

  1. Docu (for Wikipedia:Interlanguage_links)
  2. Tuf-Kat
  3. +sj+
  4. Roozbeh
  5. Muhammad
  6. grendel|khan 17:08, 2005 Mar 30 (UTC) Well, I'd prefer at the beginning, but so long as it's not between Shqip and Slovenian...

Don't standardise orderEdit

  1. Docu (except for Wikipedia:Interlanguage_links)
  2. 80.255 18:35, 9 Mar 2005 (UTC)


  • Ideally I think language link should be sorted alphabetically by lang. code in the source, because this will be the easiest for editors to follow. How these links are displayed on view is an entirely different matter that should be solved by the software (could also be solved by software in source). In my opinion it should then be sorted either by langs. with most articles, or article size/quality, or simply a list made on meta (that could be based on "weighted" article count. ---Dittaeva 11:43, 24 Jul 2004 (UTC)
  • I agree arranging the displayed links should be done in the software, so that the order does not differ from page to page (by accident). I think the displayed order should be one that makes it as easy and fast as possible for the reader to find whatever language she is looking for from a long list. We cannot assume she knows the article count of every Wikipedia, the lengths of the articles, the language codes (which are not shown) or other things like that. --Lakefall 11:47, 25 Jul 2004 (UTC)
  • Even if all Wikipedias were to sort the interlanguage links by the language code, the order of the links would not be the same in all Wikipedias. This is because the alphabetical order itself is different in different languages. Consider for example the placement of Z in the Estonian alphabet or the fact that the letters V and W are considered to be equivalent for all purposes in the Finnish alphabet. --Lakefall 15:17, 27 Jul 2004 (UTC)
    • Exactly. So, please make the algorithm sort after translittering and sort by the country's alphabet. --Lussmu 15:33, 27 Jul 2004 (UTC)
    • I agree with Lakefall and Lussmu. I do not buy user Dittaeva's comment that sorting by language code "will be the easiest for editors". When adding an interwiki link, you usually know the name of the language you are adding as well as the language code. Alphabetizing by one is not "easier" than the other, per se. It is just a matter of choice. Even languages written without the latin alphabet do have names in your language, and those names can be alphabetized like any other.     --VerdLanco 19:47, 15 Nov 2004 (UTC)
      • Except some languages have more than one name and/or more than one way of transliterating them (especially when you consider transliterating every language into every other!). Clearly (to me), the solution is to let editors do what they wish in the wikitext, perhaps simply recommending that they use alphabetical order based on language code, but then order them at page-rendering time by whatever method the community and/or user (i.e., in Preferences) decides on. - dcljr 03:29, 23 Apr 2005 (UTC)
    • Of course you'll know the name of the language you are adding, but you probably don't know the names of the other languages, if there are many of them. Therefore, sorting according to the language code makes the most sense to me. Jon Harald Søby 19:43, 24 Mar 2005 (UTC)
      • You should know the names, because you can read them from the Wikipedia sidebar. The only problem are transliterations, but they are often alredy established. -Hapsiainen 22:21, Mar 24, 2005 (UTC)
  • There should be option available for users: languages sorted by name or by real use or by combination of these. Pavel Vozenilek 18:10, 30 Mar 2005 (UTC)
  • By the way, the category below doesn't seem to exist... - dcljr 03:29, 23 Apr 2005 (UTC)
  • I don't have a strong opinion either way, but I'm interested in this question because I'd like to create a tool for correctly updating all interwiki links for a topic across all Wikipedias, and the Wikipedias seem to be divided on this matter. In one test article, ar, en, fr, nl, id, sv, and pl ordered by language name, while ca, da, de, et, es, eo, hu, it, ja, pt, zh ordered by two-letter codes. This is the main point of contention; the only other main difference is that they use HTML % codes to varying degrees (probably due to auto-conversion on save) and the French Wikipedia puts theirs at the beginning on one line. I'd like to see one consistent order across Wikipedias to make this easier and facilitate browsing one topic in various languages. Deco 23:49, 27 Apr 2005 (UTC)
  • I've returned to this page, and noticed that people had two, contradicting ideas of what they were voting about. Was this poll about the order of the links when you see the page or the order you can use when editing it? They are two different things, if someome develops software that orders the links as desired. And some people speculated that it is possible and voted assuming that automatic link ordering will be added to Mediawiki software. Due to the poll's ambiguity its results can't be binding. Also, option "Best of both worlds: Two-letter code in source, Alphabetical in view" was added after the poll had started. Not all voters saw it. People's comments are the most useful part of this poll, I think. -Hapsiainen 09:29, August 26, 2005 (UTC)
  • The sorting based on local names for languages (e.g., "Suomi" for Finnish) makes no sense to me when you take languages like Chinese into account (I had no idea that was spelled Zhongwen, BTW). If someone in Tbilisi has written up a page on bats in Georgian ("Kartuli"), he needs to know the Korean word for Korean ("Hangook-eo", or "Han-kuk-oh", or...) to know that his listing should (probably) go after Korean rather than before. The preponderance of Latin-based language names mitigates this somewhat, but how long can we count on that? Even if we come up with an authoritative easy-reference list to solve the Georgian W'pedian's problem of where to list his language (or go with the unambiguous two-letter code sorting), it's not very convenient for an ordinary user to have to basically grope through the list to find his language. --Rschmertz 07:03, 28 February 2006 (UTC)
    THis only applies to en.wikipedia. Othe 'pedias have their own rules, e.g.:
From m:Interwiki_sorting_order. Rich Farmbrough 14:31 20 March 2006 (UTC).
  • I think that ordering by two-letter codes is bad. That is good for the editor but bad for the reader (you have to scan through the whole list to be sure to find your language since you don't know the order as a reader). Sure the codes have some relation to the names of the languages but not that much in some cases. I think that we should concentrate on something that is easy to use. What ordering that is I'm not sure but it isn't the ISO codes since that is a system that we can't assume the average reader knows. Jeltz talk 11:59, 30 March 2006 (UTC)
  • If "By order of alphabet, based on local language" is implemented, how on earth will editors be able to learn the transliterations for the many languages that don't use the Latin alphabet? Ardric47 07:43, 8 May 2006 (UTC)
  • Did anyone ever suggest ordering them based on the article content? for example:
  • articles about Spain or other countries that use Spanish have Spanish listed first.
  • articles about Catalunya have Catalan listed first, then Spanish.
  • articles about South America have Spanish listed first, then Portuguese.
etc... - MTC 19:14, 29 November 2006 (UTC)
  • About the Best of both worlds: Two-letter code in source, Alphabetical in view option: People who have voted for this should consider voting for this entry in the MediaWiki bugtracker, and taking place in the discussion there. It deals exactly with the desired behavior, but has been rejected by some developers for somewhat strange reasons. --Head 15:16, 27 June 2007 (UTC)