Template talk:Infobox language/Archive 4

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

Archive 6

→

Archive 10

Français or français?

Please take a look at the infobox for Occitan. Should the native name of the language be capitalised? If so, to what extent?

I have already suggested (Wikipedia talk:Manual of Style) that, this being a datum in a box, we should show what the Occitans do & leave their name uncapitalised. Rothorpe 13:42, 10 October 2007 (UTC)

The case for capitalization is twofold:

• it is functioning as a title, and titles are, I believe, still capitalized in Occitan, as in English.
• Occitan, as the English word for the language, is capitalized anyway.

Whether these are enough together is another question. Septentrionalis PMAnderson 01:41, 11 October 2007 (UTC)

I think it should be capitalized just as being the first word in a table cell. The words "spoken" and "total" aren't normally capitalized in English either, but they are in the infobox because they start a cell. —Angr 06:27, 11 October 2007 (UTC)

OK, then, I've capitalised as in the text. Rothorpe 15:35, 11 October 2007 (UTC)

colspan="3"?

I wanted to edit the layout of the table sligtly because of a problem in an article, but at the start of the template I'm greeted with a 'colspan="3"'... WTF? Is there a special reason this table has got three columns instead of two? I'm not really prepared to hack in this template if I have no clue why this is. Also perhaps this template should have a preview sample to aid editing. Shinobu (talk) 13:39, 21 November 2007 (UTC)

Created languages like Esperanto have an optional third column for the date that the language was created. --CBD 02:20, 15 December 2007 (UTC)

Flags

I noticed in the infobox on German language we have a whole collection of national and municipal flags. As they do not seem to add anything to the article I would propose to remove them. Does anyone have any objections? --John (talk) 21:53, 11 March 2008 (UTC)

Setting doesn't work

I tried to add "setting=Liturgical language" to the infobox at Church Slavonic language, but if there is any content in that field, the template won't display correctly. — —Michael Z. 2008-04-06 06:07 Z 06:07, 6 April 2008 (UTC)

ISO code for old languages

I propose to add a new field to specify the ISO language code for the historic versions of the language, e.g. 'osp' is ISO 639-3 code for "Old Spanish" (the Spanish spoken some centuries ago), or 'ang' for Old English. See for example all code starting by "Old " in [1]. Opinions? —surueña 13:49, 8 May 2008 (UTC)

Those should just be in the articles for the "old" language in question. So Old English language has an infobox showing its code as "ang", and Old Spanish language has an infobox showing its code as "osp". There's no reason for the infobox at Spanish language to show "osp". —Angr 20:47, 8 May 2008 (UTC)

Sure there's a reason. Old Spanish is a kind of Spanish. This is just the kind of reference information I'd like to see at a glance, rather than searching the article text and hoping that I found all of the relevant articles to check. Likewise, French fr/fre/fra < Middle French frm < Old French fro, en < enm < ang, etc. —Michael Z. 2008-05-08 22:07 z

I agree with both of you. However, although ideally all languages should have a specific article describing the historic variant studied by scholars from old texts (and then that specific code should be specified in that article), in practice only a few languages have it. So probably instead of a new field, maybe a new section named "historic variants" linking to those variants, and giving the language code. I think this would also encourage editors to create a new article about old variants. PS: Angr, does your nick have anything to do with the Old English lang code? :-) Cheers —surueña 07:44, 9 May 2008 (UTC)

How about something like this? I would only list direct ancestors (older versions), and not all of the influencing languages. —Michael Z. 2008-05-09 08:22 z

Language family: Indo-European (ine)
                  Germanic (gem)
                   West Germanic
                    Anglo-Frisian
                     Anglic
                      English
Ancestors:       Middle English (enm)
                 Old English (ang)
Writing system:  Latin (English variant, Latn)

Looks like a good approach to me (but I'm not a language expert at all). I suppose new parameters are needed to specify the language code for the family and ancestors, e.g. anc1iso, fam1iso, anc2iso, fam2iso... In my opinion it's more difficult to add the script code (you can't insert the language code in existing parenthesis), and anyway articles for each writing system are far more common. But I support the rest of your proposal. Cheers —surueña 13:37, 9 May 2008 (UTC)

Requested move

The following discussion is an archived discussion of the proposal. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.

The result of the proposal was no move per WP:IBX. JPG-GR (talk) 04:29, 31 July 2008 (UTC)

I request move this template to Template:Infobox language. The word “language” is not a proper noun, so it should use lowercase letter. --✉ Hello World! 10:38, 27 July 2008 (UTC)

I believe Infoboxes standardly have capital letters after the word "Infobox" in their titles. The idea is that this is an infobox whose name is "Language". See WP:IBX, which says "Name the template [[Template:Infobox Some subject]] (Somesubject should be in the singular and capitalized)." —Angr 11:21, 27 July 2008 (UTC)

I Knew that the former name of this infobox was just Template:Language. This is the main reason why now they're using uppercase L. But, anything else? --✉ Hello World! 05:04, 28 July 2008 (UTC)

Well, the page I linked to above is from the Manual of Style for infoboxes, which says that the name of the topic (i.e. the word after Template:Infobox) should be capitalized. If you want to change that, bring it up on that talk page, since it will affect all infoboxes, not just this one. —Angr 05:22, 28 July 2008 (UTC)

The above discussion is preserved as an archive of the proposal. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

add Coordinates via Template: coord

As per discussion with User:Ngio and WP:BOLD, I've added an latd parameter that generates the standard coordinates thingie. See User:Ling.Nut/rere for example. I'll add documentation. Revert both if you are extremely unhappy, but I think it's a useful option. Ling.Nut ^{(talk—WP:3IAR)} 16:09, 5 September 2008 (UTC)

Hmm... useful for very small languages perhaps, but it would be silly to try to implement it for languages spoken across an entire country, let alone in several countries. —Angr 20:54, 5 September 2008 (UTC)

Representing a language as a point on a map is always an abstraction (well, maybe something like Anuta, spoken on a tiny dot of land), but I think it's a useful thing to have. For example, I was wanting this to help us build up a useful set of coordinates for things like the WALS database. Ngio (talk) 21:03, 5 September 2008 (UTC)

Well, where would you put the point for English? How can one coordinate dot represent the UK, Ireland, Canada, the US, Australia, New Zealnd, etc., etc.? And even for a geographically compact language like Irish, I can't think of a place to put the dot that doesn't seem to favor one dialect over the others. —Angr 21:43, 5 September 2008 (UTC)

The template produces two map options. The one that is the most obvious is also the one that is by far the least useful!! If yo have javascript enabled, try clicking the globe image, instead of the blue text. It produces a far superior image... If even that granulation isn't fine enough for you, we could try playing with variable Coordinate Parameters. But then the temptation to make things a little too complicated would arise very quickly... Ling.Nut ^{(talk—WP:3IAR)} 01:50, 6 September 2008 (UTC)

I'm not talking about viewing the map, I'm talking about deciding which coordinates to give. —Angr 10:19, 6 September 2008 (UTC)

(undent) I see. I wouldn't use coords for English language, either. But for all the Formosan languages in Taiwan... it would clearly be more useful. It's optional, at any rate. If it isn't (is) useful, then... it shouldn't (could) be used.Ling.Nut ^{(talk—WP:3IAR)} 12:40, 6 September 2008 (UTC)

I've tested this out on the Reefs-Santa Cruz languages: Äiwoo language‎, Santa Cruz language, Nanggu language. It seems to work fine, and I hope that Angr agrees that these are cases where giving coordinates is appropriate. But I'll wait to see if there's any more feedback before adding them more widely. -- Ngio (talk) 09:47, 9 September 2008 (UTC)

Indic notice

A look at Tamil language shows that the indic notice box occupies only the right column and thus it unnecessarily expands the template. Could someone center it? Eklipse (talk) 12:09, 16 September 2008 (UTC)

IANA tags

The template should accommodate IANA tags and subtags, e.g., Belarusian language#Computer representation. —Michael Z. 2008-11-14 07:03 z

Aren't they the same as the ISO 639 codes? —Angr 07:43, 14 November 2008 (UTC)

Oops, I actually meant the IETF language tag, which includes the language code, and optionally adds subtags for region, language script, and variant. —Michael Z. 2008-11-14 15:34 z

Slight tweaks

I will make slight tweaks to the infobox design. It won't be drastic. A copy can be found at {{User:The Obento Musubi/Language}}. – Obento Musubi (C • G • S) 04:57, 16 December 2008 (UTC)

Wiki policy for lists of countries

Wondering about the spaces for areas "spoken in" (defined as "countries in which it is mainly spoken") and "official language in" (defined as "list of countries in which it is an official language"). It is of course useful to be precise, so that self-declared states such as S. Ossetia are relevant, as are dependent territories such as American Samoa. However, this runs the risk of sparking edit wars, even though we're not making any direct claims about legitimacy. Is there any policy governing this? kwami (talk) 20:30, 27 December 2008 (UTC)

List of countries is a redirect, so it was perhaps not the best talk page to ask at...

Short answer is that there is no policy - at least, as far as I know. You may wish to read Wikipedia talk:WikiProject Countries/Lists of countries for a recent discussion on a related topic.

My feeling would that you should use all the countries on the list of sovereign states, plus any other territories specifically relevant to the topic at hand (so, Hong Kong on the list of English-speaking countries for example). If you want dependent territories as well, you should use the list at dependent territory. This isn't a perfect solution - both lists have grey areas on the edges of their definitions - but I think it's a good general guide.

Where status over an entire country/dependent territory is disputed, you should include a footnote that links to an article which gives detail. On the English-speaking one, you might list the Falkland Islands, and then put "disputed by Argentina, see Sovereignty of the Falkland Islands" in a footnote, for example. For South Ossetia you might put "recognised as part of Georgia by the UN, see International recognition of Abkhazia and South Ossetia".

The other thing is that in an article or section marked out as a "list of countries", you should make it clear in the introduction to the list how it is decided what is included and what is not included as a country. The field concerned on this template doesn't use the word "country", so it's less significant, but it's something to think about. Pfainuk talk 00:18, 28 December 2008 (UTC)

There is an emergent consensus here to use ISO 3166-1 as a base for lists of countries (changing some names to reflect a more NPOV stance). To this base list can be added additional de facto independent entities in italics with a note about their international status. Thus a list might look like this:

United Kingdom, Taiwan, France
Abkhazia (recognized only by Russia and Nicaragua)

The discussion actually has more far-reaching impacts, but for the Language Infobox, that is the most relevant one. (Taivo (talk) 19:14, 2 January 2009 (UTC))

Personally, I see no reason to single out polities which claim independence. Chuvash is spoken in Chuvashia (Russia) and surrounding areas—or maybe we could word it Russia (Chuvashia and surrounding areas). Same with US states. Restricting ourselves to ISO 3166-1 is unhelpful. kwami (talk) 20:07, 2 January 2009 (UTC)

I don't know if you intended this to be as ironic, but it is when you consider how many countries habitually refer to Taiwan as a country and how few officially recognize it as such. — CharlotteWebb 08:58, 3 May 2009 (UTC)

Referencing for Country Lists

The referencing for the lists of countries in which a language is spoken is universally bad. There is an assumption being made that Ethnologue is most likely the source of the list, but editors are constantly adding and subtracting countries in some articles based on no more valid criteria than "I live there" or "It's called Abkhazia because that's where Abkhaz is spoken" or any of a dozen other violations of WP:OR or WP:Syn. (Taivo (talk) 09:03, 31 December 2008 (UTC))

Category:Articles containing language text

This template appears to be populating Category:Articles containing language text for some reason. --Pascal⁶⁶⁶ 21:09, 13 March 2009 (UTC)

It's the ISO 639 codes. They put the article in the appropriate category, e.g. the ISO codes de and deu in the infobox put the article German language into Category:Articles containing German language text. And if the code parameter is empty, it populates the category where the language name is also empty, I guess. —Angr 23:00, 13 March 2009 (UTC)

Could a template expert please change the code

{{#ifexist:Template:ISO 639 name {{{iso1}}}|[[Category:Articles containing {{ISO 639 name {{{iso1}}}}} language text]]}}

so that it does not categorize if iso1 is either empty or “en” As noted above, if it is empty, it populates the non-existent Category:Articles containing language text. If it is “en” it is it inserts English language in Category: Articles containing English language text, which was deleted by this CfD. —teb728 t c 05:23, 15 March 2009 (UTC)

Putting in an automatic category is a problem since a whole lot of language articles contain no text in that language (probably the majority of them). This category should be inserted manually where appropriate and not done automatically. There shouldn't be an entry for "Language articles containing Kofyar text" for example. I removed the offending line completely. Insert this manually only in those language articles that actually contain text in that language, otherwise leave it out of most language articles. (Taivo (talk) 12:27, 15 March 2009 (UTC))

I believe the category is added automatically also when you use the {{lang}} and {{lang-xx}} templates around the foreign-language text, so if you're careful always to use those tags, the article will go into the category without anyone having to worry about the infobox or the category. —Angr 21:31, 15 March 2009 (UTC)

--------->

Personally I think the ASCII-art "arrows" are uglier than a mud fence. Any objections to using real arrows and a more gradual increase in indentation as with Template:Infobox Writing system (see Tibetan script for example). — CharlotteWebb 09:06, 3 May 2009 (UTC)

I'd prefer to see this system of genealogy-by-typographical-trickery abolished entirely, to be honest. How about an inline list, 1 → 2 → 3 → 4 an so on? Chris Cunningham (not at work) - talk 11:43, 3 May 2009 (UTC)

Seems like the width constraints would make that format more difficult to read. I'd say an unordered list would work well if had an easy way to make the indentation less than the default at each level. — CharlotteWebb 14:08, 3 May 2009 (UTC)

Another thing: I can't figure out how to get rid of the blank line at the top of the template, see example here. There is a "p" tag of unknown origin containing only a single nbsp. But only on articles where there is no kind of warning banner above the infobox. — CharlotteWebb 16:34, 3 May 2009 (UTC)

It'll be caused by the parser's interpretation of line breaks in wikitable syntax. Eliminating difficult-to-debug issues like this is one of my main motivations for detangling voodoo code in templates. It's probably in either the family tree or codelist sections of the code, which are unfortunately so hair that I haven't begun to untangle them yet. Chris Cunningham (not at work) - talk 19:21, 3 May 2009 (UTC)

The arrows, ASCII-arty or not, are an abomination in my opinion and I reverted them as soon as I got back from my wikibreak. I'm not code-savvy, so I can't really follow all the arguments here, but I have trouble seeing what's so bad about the old layout. How, where and why is it a problem to readers?

Peter ^Isotalo 22:14, 3 May 2009 (UTC)

For viewers, it's guaranteed to be an accessibility nightmare. For code editors, it's not only a total hack but also a huge amount of additional code (nearly 10% of the entire template logic) for very little gain. Chris Cunningham (not at work) - talk 08:20, 4 May 2009 (UTC)

How is it an accessability nightmare? It's a set of links arranged in a vertical hierarchy with a slight increment for each to improve layout. Do arrows improve legibility dramatically? If this is annoying to code editors, I'm sorry to hear that, but it doesn't merit switching to an interim layout that looks like ASCII-art.

Peter ^Isotalo 08:29, 4 May 2009 (UTC)

Oh, don't get me wrong - all of that applies to the arrows version as well. Accessibility-wise, we should be avoiding the use of non-breaking spaces to impart meaning because obviously screen readers ignore them as whitespace, and likewise with line breaks. The family tree section is very guilty of both of these. Chris Cunningham (not at work) - talk 09:32, 4 May 2009 (UTC)

Figure 1

Language family

Indo-European Germanic West Germanic Anglo-Frisian Frisian West Frisian

Figure 2

Language family

Indo-European Germanic West Germanic Anglo-Frisian Frisian West Frisian

What would be most "readable" to screen readers? My first thought is to use html elements which already imply a hierarchy. However the default indentation of the "bullet list" elements (see figure 1) is too much to for it to fit properly within the infobox. To manually over-ride this we could abandon the asterisk short-hand and use real html to get a more compact appearance (see figure 2). This would be vaguely similar to what I did with Template:Infobox Writing system (except I used divs). Alternatively I believe there are ways to do this automatically from monobook.css, etc. based on the class-name of the div or whatever element contains the bullet list, allowing us to vicariously adjust the margins and line-height while still using the wikitext asterisks. Either way I know I can come up with something that works better than the bloody nbsps. — CharlotteWebb 14:51, 4 May 2009 (UTC)

Nested lists could work - your second example looks good. Might want to run it by WT:ACCESS for input. Chris Cunningham (not at work) - talk 15:45, 4 May 2009 (UTC)

CSS bits were simpler than I remembered. You'd just need something like:

.squeezedlist ul {
    line-height:100%; margin-left:5px [or something in em's or en's etc.];
    }

{| class="infobox"
! Label 
| class="squeezedlist" |
*Foo
**Bar
***Etc.
|}

Or in this particular case it might be something like:

{{#if:{{{fam1|}}}  |*{{{fam1}}}
}}{{#if:{{{fam2|}}}|**{{{fam2}}}
}}{{#if:{{{fam3|}}}|***{{{fam3}}}
}}etc. etc.

Of course there might not be any need to keep these as separate parameters rather than accepting the list as a whole (shrug). — CharlotteWebb 16:20, 4 May 2009 (UTC)

Empty line before this Infobox

I noticed that sometimes this Infobox is preceeded by an empty line (visible as <p>&nbsp;</p> in HTML). I think that  s from this template somehow "leak". I located two sources of them: one is Official status part that contains them directly in the code and the second are Template:Infobox Language/signnotice and Template:Infobox Language/nonotice that also contain them and are somtimes included in this Infobox. I konw that they wouldn't be there without a reason so I didn't just remove them, but instead are asking here: Why are those  s there? And what would happen if I removed them? Svick (talk) 09:55, 3 July 2009 (UTC)

It'd probably break the table rows, which need non-breaking spaces to ensure that the wikitable code reads the line as having a line break. The code in question is exceedingly hairy and of very little actual value; I'm tempted to rip it out completely and replace it with a more straightforward system. Chris Cunningham (not at work) - talk 10:41, 3 July 2009 (UTC)

Let's just re-write it using regular html. You tell me what's more readable:

{{#if:{{{foo|}}}
| <tr class="xyz">
    <td> Foo </td> <td class="pdq"> {{{foo}}} </td>
  </tr>
}}

or

{{#if:{{{foo|}}}
| <nowiki/> <!-- only clean way to force a line break before "|-" -->
{{!}}- class="xyz"
{{!}} Foo {{!}}{{!}} class="pdq" {{!}} {{{foo}}}
}}

Surely the former in most cases? — CharlotteWebb 12:37, 4 July 2009 (UTC)

Couldn't the latter also be written:

{{#if:{{{foo|}}}|
{{!}}- class="xyz"
{{!}} Foo
{{!}} class="pdq" {{!}} {{{foo}}}
}}

though? That's considerably cleaner. Chris Cunningham (not at work) - talk 11:19, 6 July 2009 (UTC)

Interwiki ->pl:

pl:Szablon:Język ->

pl:Szablon:Język infobox OK. --SPL908455, Henryk (talk) 17:04, 21 October 2009 (UTC)

TfD of Infobox Language/Coptic

I nominated Template:Infobox Language/Coptic for deletion, see Wikipedia:Templates for discussion/Log/2009 November 17#Template:Infobox Language/Coptic. Svick (talk) 17:05, 17 November 2009 (UTC)

Analysis of infobox fields

Here is a breakdown of how many times each field is used in language infoboxes in section 0 of articles. Note there are quite a few in pages which the template does not support. There may also be a tiny amount of pollution from fields used in other infoboxes on the same page:

field	count
sil	1
Atlas-UNESCO-language	1
Sister-Languages	1
Atlas-UNESCO-country	1
related	1
poptime	1
langs	1
type	1
popplace	1
time	1
iso15924	1
group	1
languages	1
rels	1
map_caption	2
longd	3
latd	3
longEW	3
latm	3
latNS	3
longm	3
minority	4
ib-part	6
signers	7
date	10
creator	13
setting	13
posteriori	14
family	17
caption	24
image	26
iso2t	32
iso2b	32
notice	46
extinct	48
pronunciation	57
map	61
rank	69
agency	73
iso1	102
nation	104
ll	118
script	135
ld	193
lc	193
nativename	298
iso2	320
region	453
states	464
speakers	513
iso3	518
familycolor	584
name	598
fam	2220

—Preceding unsigned comment added by Hippietrail (talk • contribs) 00:31, 22 December 2009 (UTC)

endangerment status

There should be a parameter for endangerment status, e.g. according to the UNESCO list[2] or other sources cited in Endangered languages. UNESCO has the categories not endangered, vulnerable, definitely endangered, severely endangered, critically endangered, extinct and a separate flag revitalized. This could replace the extinct parameter. Think Template:Taxobox#Conservation status for languages.--87.162.37.147 (talk) 02:09, 26 February 2010 (UTC)

I came to this talk page to suggest the same addition. Any other views on this? sephia karta | dimmi 14:48, 15 March 2011 (UTC)

Extensions

What can we do about including IANA language subtag extensions? For instance, Ulster Scots is a variety of Scots ISO 639 "sco", but will soon have a suffix, e.g. "sco-ulster". -- Evertype·✆ 00:38, 19 March 2010 (UTC)

dialects and standard forms

Added parameters for dialects & standards. Sorry for those of you I freaked out: I made an el-for-one typo and it took me forever to see it.

Currently, it won't look too good if a box has both 'dialects' and 'dia1' or both 'standards' and 'stand1'. Also, you need one of the dialects or standards to be #1 for it to format properly. Maybe s.o. a bit more competent can iron it out if there's a need. — kwami (talk) 05:19, 26 April 2010 (UTC)

Linguasphere

Has anyone any feelings, positive or negative, about adding a line for inclusion of the Linguasphere code? --Taivo (talk) 20:22, 21 September 2010 (UTC)

Is there any point? I just noticed it on the Basque language page and followed some links, the Linguasphere website is dead, which usually is a pretty bad sign. And I can't find anything on it that's not in a "family", where did they stick isolates like Basque? Akerbeltz (talk) 09:35, 9 November 2010 (UTC)

The 10 sectors that the Linguasphere world is divided into consist of 5 "phylosectors"--genetic units--and 5 "geosectors"--non-genetic, geographical units. Basque is in the Eurasian geosector. Each of the sectors is divided into 10 zones, some "phylozones" and some "geozones". Basque (called "Euskara" in Linguasphere) has its own phylozone. While the website is inactive at this time, the Linguasphere Observatory, which was always a small operation, is still operating from the last information I have read. It's still being funded by the U.N. --Taivo (talk) 13:05, 9 November 2010 (UTC)

Ah right. If it's THAT small though, is it (going to be) big enough for their codings to become used or useful? Akerbeltz (talk) 13:07, 9 November 2010 (UTC)

Fixing font in language codes section

The coding in the language codes section needs to be tweaked. The rest of the template is displayed in a sans serif (Arial?) font, but the information in the language codes section is displayed in a serif (Times?) font. It looks odd to have a different font displayed in that section and since the serif font is wider, it can take up a lot of room when several pieces of information are displayed (such as at Croatian language). Can someone fix this? My coding skills aren't sufficient for the task. --Taivo (talk) 13:27, 9 November 2010 (UTC)

Macrolanguages and minor fixes

The display of macrolanguages is currently a bit messy: Sometimes their ISO 639-3 code is in the first dialect parameter (which works but isn't strictly true), sometimes it is in the iso3 parameter (which isn't displayed since the individual languages in the dialect params hide the iso3 param). I suggest putting these codes uniformly in the iso3 parameter and tweak the code so it isn't hidden.

I also suggest displaying all omitted parameters in the same fashion, like an n-dash, instead of a mixture of n- and m-dashes, "None" and an empty field. I'll add these changes myself if no one complains over the next few days. --ἀνυπόδητος (talk) 18:06, 15 December 2010 (UTC)

Another issue is that if the language Infobox is being used on a page discussing a particular dialect, then the dialect gets automatically tagged as pertaining to a macrolanguage (since the fields lc1, ld1 and ll1 are required to mention the dialect name), even though the language itself may not be considered a macrolanguage, or may not even be mentioned by the ISO 639 standard. — Io Katai ᵀᵃˡᵏ 17:13, 27 January 2011 (UTC)

I think the "Language codes" section should only mention the ISO code for the dialect, not the language. You can use the iso3 parameter for that purpose. lc1, ld1 and ll1 are meant to list codes of dialects/individual languages in the article about the macrolanguage/parent language (see e.g. Albanian language). The genetic relationships should be in the top section of the box, using fam1 etc. (see Tosk Albanian). If you object to adding the language to one of these family parameters, a possibility would be using dia1 (which would require to change the template so that it displays "Dialect" instead of "Dialects" if only one is given). I'm not sure this is a good idea, though, because dia1 seems to be intended for a dialect (i.e. a subdivision of the article's topic). --ἀνυπόδητος (talk) 17:36, 27 January 2011 (UTC)

As you noted, the tags dia1, dia2 and so on, are used to signal dialects of the particular language page in question. In some cases, a major dialect grouping may also encompass a number of smaller, but significantly different regional (sub-) variants (or dialects). However, the ld tag simply stands for "language dialect", and from my understanding it was introduced as a way to denote dialects (and to deprecate the Dialect Infobox), so limiting its use to macrolanguages doesn't seem right to me (and why not use mc, md and ml instead?). The other problem with merely showing the language codes is that they infer that the article's subject has its own iso code, as in the case of Tosk Albanian, but not all major dialects have ISO codes assigned independent from their parent language (they may, however, have distinct linguasphere code identifiers). And so the ld tag provided a way to mention that the dialect was included in that language code, without confusing it as having its own unique identifier.

A solution to this would be to remove the dependency on the lc tag for variants which do not have a separate language code (i.e. not macrolanguages), and move the name beside the last ISO code. — Io Katai ᵀᵃˡᵏ 20:20, 27 January 2011 (UTC)

I'm still not sure what you want to do. For what purpose do you want to use ml? The macrolanguage code is in iso3 (if the article is about a macrolanguage), and the lds are the individual languages.

If you have got a language without an ISO code, and its dialects have an ISO code, you can use ld and the box will say nothing about "macrolanguage" if you don't specify iso3. If both language and dialects have ISO codes, the language is a macrolanguage by definition if I'm not mistaken. If the dialects don't have ISO codes, they shouldn't be in the "codes" section for that very reason. Can you give a specific example? Thanks, ἀνυπόδητος (talk) 14:48, 2 February 2011 (UTC)

I thought ld was for lects with distinct iso3 codes, not for dialects in general. — kwami (talk) 00:37, 25 February 2011 (UTC)

Remove rank?

Several years ago I gave up maintenance on List_of_languages_by_number_of_native_speakers as hopeless. I just came back to it, and it's still garbage. Considering that there's no reliable distinction between being 40th on the list and 60th, shouldn't we just delete this field? It has no information value, but can be extremely misleading. It's not just uncertainties in the number of speakers of the language in question, but it's also hopelessly complicated by the inaccuracies in all of the languages ranked higher in the list. I mean, we don't even know how many speakers of English, French, Portuguese, or Persian there are: not to the nearest 10 million anyway. If you want to know how many people speak a language, we have a field for that; IMO, that's the only thing we can responsibly provide. (For the top dozen or so, which we can say s.t. intelligent about ranking, we can always mention the rank it the lede. Or maybe we can modify the field so that it only accepts values < 20 or so.) — kwami (talk) 00:35, 25 February 2011 (UTC)

No objection. If you think the first 20 languages can be reliably ranked, I could change the field so that it displays "lower than 20" or something like that for other languages. Otherwise, I'd just remove the field and ask at WP:BOTREQ to remove the parameters from the articles. --ἀνυπόδητος (talk) 10:46, 25 February 2011 (UTC)

To the extent that we can rank the top 20 (we'd need to say things like "four-way tie for 13th place, given statistical uncertainties" or "somewhere between 13th and 17th place"), IMO it would be best to put it in the lede. Also, "lower than 20" would be odd for a language somewhere around #200. So deleting the field altogether would IMO be the best bet. — kwami (talk) 11:14, 25 February 2011 (UTC)

I agree with the removal of this in the infobox. Counting speakers of X language is always an iffy situation anyway and can only give a rough approximation of the number of speakers no matter how "accurate" the counting in Y country supposedly is. Once you start dealing with international languages that have some official status in one or more countries, the sociological pressures to claim to speak X language will throw any census off. --Taivo (talk) 13:50, 25 February 2011 (UTC)

Deleted, asked at bot requests. — kwami (talk) 14:04, 25 February 2011 (UTC)

collapse coding section when not in use

I've collapsed the iso1 and lingua code sections when not in use. I can't figure out how to do this with iso2, at least not this late at night. Even so, the tables look better. IMO we should display iso3 regardless, as the absence of an iso3 code is almost always relevant. — kwami (talk) 07:25, 20 April 2011 (UTC)

If someone knows how, can they collapse iso2 as well? — kwami (talk) 23:15, 21 April 2011 (UTC)

add ethnicity parameter?

I think it would be useful to add an ethnicity parameter. That would help us avoid stupid-sounding opening lines such as "the Balam language is the language of the Balam people" just so we can get a link in. It might also encourage people to link to the ethnicity articles. Currently it's very hit and miss: The ethnicity articles typically link to the language articles but not vice versa. — kwami (talk) 23:15, 21 April 2011 (UTC)

I agree, despite the fact that this inclusion might provoke an issue with Serbo-Croatian and perhaps some other. Let it. Best regards, --Biblbroks (talk) 23:42, 21 April 2011 (UTC)

It would be optional; we wouldn't have to add it if it causes problems. And we could say 'English (people) and international' for languages like English, Spanish, etc., rather than listing everybody the big languages are native to. — kwami (talk) 23:50, 21 April 2011 (UTC)

International ethnicity wouldn't go I think. But I already agreed, even if we don't make it optional - if it causes problems, then we should face them, not duck by hiding it. I think the ethnicity parameter would really help solve some issues. What might be a problem is the exact name of the parameter. No, I changed my mind. Let's name it exactly ethnicity. And not make it optional - if this is possible. If this causes a problem we will also face it. All the best, --Biblbroks (talk) 07:19, 23 April 2011 (UTC)

Yeah, if it shows up regardless, then that would encourage people to (a) find the ethnicity article, which may be under some other name, or (b) start one. Either way we're better off. — kwami (talk) 21:48, 23 April 2011 (UTC)

Okay, done. Will show up on all articles. I tested it out on one, and it seems to work fine. Currently it displays an en dash if nothing is entered, which may be a bit misleading. We could simply leave it blank, or make it optional. I expect to have some more opinions here now that it's gone live. — kwami (talk) 05:07, 24 April 2011 (UTC)

I made it optional. We shouldn't make it seem like this is required. I can just imagine what people would try to enter for some of the more major languages. Plastikspork _―Œ^(talk) 06:03, 24 April 2011 (UTC)

So what? And if they try, then what? Why should even "most major" languages be treated differently than for example "non most major" Serbo-Croatian. It should be required. --Biblbroks (talk) 22:50, 24 April 2011 (UTC)

I won't take sides on whether it should be required or not, as I'm not convinced either way. If it is required, I'm not sure a dash is what we want as a placeholder. Maybe just a blank space. Also, we might want a different header than 'ethnicity', to make it clear that we don't mean Quechua for Spanish or Cherokee for English. Maybe for inter-ethnic langs we could have an alt header that would read 'original ethnicity' or 'originating ethnicity'? — kwami (talk) 00:20, 25 April 2011 (UTC)

Per WP:BRD, we are now in the "discuss phase". I disagree that this field should be required. If either of you want to go around and add it, then go right ahead, but we should not make it required. Plastikspork _―Œ^(talk) 00:46, 25 April 2011 (UTC)

Optional. I have to agree with Plastikspork that this parameter should simply be an optional one to be used where needed. I don't think that sentences like, "Timbisha is spoken by the Timbisha people", are required or that an ethnic tag in the infobox is required. Sometimes it's just too complicated for an infobox tag. The Timbisha people, for example, don't call themselves "Timbisha", they call themselves "Shoshoni" and the tribe is called "Timbisha". But not all Shoshoni are Timbisha. Sometimes things just aren't collapsible to a line in a template, so this should not be a required parameter. It might be occasionally useful where the ethnonym is different from the linguanym and such a thing can be gracefully expressed, but that is not always the case. English, for example, should have no ethnic parameter at all and if it is required and visible in the template box, that simply invites a million different edits from the POV pushers of the day. It should be optional. --Taivo (talk) 00:53, 25 April 2011 (UTC)

Right. In addition, if you really want to encourage editors to add this field, the way to do it would be to make a bot request to add |ethnicity=– to the 3880 transclusions. This would be basically the same, and make it clear how one changes the "–" to something else. I still don't think we need to do this, but if people insist on making it "required", this would be the way to do it. By the way, the infobox should contain a summary of what appears in the article, and a good start would be to make sure this important information appears in the 3880 articles. Thanks! Plastikspork _―Œ^(talk) 01:05, 25 April 2011 (UTC)

Per BRD, I've reverted to a state before even parameter was included - if we want to play safely. Also, we should excluded it until we find a better term for this "ethnicity" parameter, because the intention behind previous revert by Plastiksport was, I believe, to play it completely safely. Regards, --Biblbroks (talk) 01:15, 25 April 2011 (UTC)

Actually per BRD we should keep it required until we discuss the change - I think. But I am not sure what others think about BRD in this case. --Biblbroks (talk) 01:17, 25 April 2011 (UTC)

Which ever of the two states you prefer is fine with me. My only issue was having it appear in 3880 articles while we were in the middle of a discussion. Thanks! Plastikspork _―Œ^(talk) 03:20, 25 April 2011 (UTC)

No one is debating whether or not this should be an option in the template box. The issue is whether it should be optional or required. The original state was optional, therefore WP:BRD applied when Biblbroks changed it to mandatory and Plastikspork objected. Therefore the text should stay at optional (where I changed it to) until a consensus is built for required. If no consensus for requiring this parameter is reached, then it stays at optional. For the reasons I stated above, I think this should not be required, but should remain an optional parameter. --Taivo (talk) 04:46, 25 April 2011 (UTC)

Plastiksport, I understood your concern, and I addressed it. I think you would agree I addressed it more appropriately than you. What I don't understand is which two states you think I prefered: excluded EXOR required?

Taivo, actually I am debating. More original state was no option - the potentially dubious (in this case) syntagm "more original" was used because I don't think BRD should be applied here. I don't think template's entry should stay optional, because this will be just postponing the problem. The problem which I believe exists because the current issue exists. What we will have are local disputes/editwars whether include the entry or exclude the entry. And those "localities" will, I believe, eventually lead to this discussion - "a less local" (!?!) dispute. Because I think there is no middle "road" how to deal with this: either play the safe road or play the liberative road. I believe you can "catch my drift". Or is it "catch my grift"? Anyway, maybe I am wrong, maybe there will be no 'include vs. exclude' localities. If we should test this "hypothesis" I think then we should make it optional. And then I am wrong about the two roads, but perhaps not about the two paths that lead from the third road. Liberative vs. safe. :-) What do you think? All the best, --Biblbroks (talk) 06:02, 25 April 2011 (UTC)

Taivo, actually you are right, but I think you believe am too - about the two paths leading from the third road. In the end of the road. That is when thinking about the end of paths. :-) So we are both right, right? Right - as in opposite of wrong. :-) --Biblbroks (talk) 06:31, 25 April 2011 (UTC)

So, I'm assuming that you now support this as an optional parameter, Biblbroks. If it turns out to be a problem, we can always revisit the issue. --Taivo (talk) 07:46, 25 April 2011 (UTC)

Your assumption is based on facts. ;-) Np, I support this "optionality" for now. All the best, --Biblbroks (talk) 01:50, 26 April 2011 (UTC)

I have just been experimenting with this new parameter "at large", and the problem I find is that it doesn't show if included as empty in some handpicked language article's infobox. I know this is the default behavior but can it be changed in this case and that without the "–"? Just plain empty - so as to experiment with experiments mentioned earlier. You know, encouraging the editors? Even "more neutrally". ;-) Best regards, --Biblbroks (talk) 02:09, 26 April 2011 (UTC)

If the parameter is empty, then it should not show. That's the point of an optional entry. But can you give an example? --Taivo (talk) 06:54, 26 April 2011 (UTC)

Yes, Taivo, I understand that this is the default behaviour. What i am suggesting that in this case we should consider the entry to be required to show - to be visible (even if empty) when an editor lists it in some handpicked language article's infobox. I think an example of what i am trying to point is instead of this:

...

---

Region the Western Balkans
Ethnicity Serb, Croat, Bosniak, Montenegrin
Total speakers 16.3 million[1]

---

to show it like this:

---

Region the Western Balkans
Ethnicity
Total speakers 16.3 million[1]

---

... that is when an editor specifically mentions it in some given infobox. If i understood kwami correctly "...Currently it displays an en dash if nothing is entered, which may be a bit misleading. We could simply leave it blank, ..." existence of this alternative also might be endorsed as one of the alternatives of behaviour by one person: kwami. :-) Regards, --biblbroks (talk) 13:08, 26 April 2011 (UTC)

• That's the total opposite of the idea of "optionality" that everybody else has been arguing for. "Optional" means just this: if it's empty it remains invisible. And in my opinion that's the only acceptable version. A visible "ethnicity" field (which, even if empty, would always come with the implicit appeal to fill it with something) would be trivial for some cases, stupid for some others, and downright dangerous for quite a few. Fut.Perf. ☼ 20:23, 26 April 2011 (UTC)

• I thought of optional as the content of this entry. Not optionality of entry itself. I was intending to address those of the trivial cases so as to not be dangerous for those quite a few. But perhaps I was wrong when presenting my thoughts. Maybe it doesn't matter at all. Anyway, all the best. --biblbroks (talk) 20:49, 26 April 2011 (UTC)

Familycolor

Quick question: If a language isn't an isolate, but rather part of a group of related languages, and if it's classification within one of the major groups is controversial, what colour should it take? For example, should Japanese, Okinawan, Yonaguni and so on take on the Altaic family colour, or the language isolate colour? Or should it not be specified at all? — Io Katai ᵀᵃˡᵏ 19:59, 22 April 2011 (UTC)

Since "Altaic" may not be a legit family but an area, IMO it doesn't hurt to color them green, just as we color Aussie, Papuan, and American languages areally despite them being isolates. If half of our articles are colored grey, the color coding IMO doesn't help that much. — kwami (talk) 21:52, 23 April 2011 (UTC)

I have a feeling that edit warring like this is going to continue to perpetuate itself as long as the classification remains controversial. Is there no way to provide a more generic, geographical name than simply 'Altaic'? The term "American languages" is already much more generic, couldn't we have something like "Asian languages" as well? Just a thought. — Io Katai ᵀᵃˡᵏ 02:02, 14 July 2011 (UTC)

The reader never sees the name, and 'Altaic' is the only name I'm aware of, whether you consider it genealogical or areal. — kwami (talk) 19:37, 30 July 2011 (UTC)

sign language auto-cat

Is there an auto-cat for sign languages? I don't see it, but cat:SL remains in the SL articles even when removed from the text. If this template is the problem, could it be removed? Not every SL needs to show up in the main category. — kwami (talk) 04:07, 5 May 2011 (UTC)

Removed. Was in a subtemplate. All SL's are still in the SL cat. — kwami (talk) 06:00, 31 July 2011 (UTC)

no iso3 code

If you write 'none' for the iso3 code, the article will now be added to Category:Languages without iso3 codes. These can be checked when the next edition of Ethnologue comes out to see if they've been assigned a code. Or they might be assigned one at Multitree. — kwami (talk) 22:25, 22 May 2011 (UTC)

Multitree codes are not ISO 639-3 codes and should not be listed as if they were. --Taivo (talk) 04:26, 23 May 2011 (UTC)

Some are ISO, some are not. SIL is shifting languages extinct before 1950 over to LingList, and they have ISO codes. For those which do not, we have a separate Linglist parameter. — kwami (talk) 19:09, 30 July 2011 (UTC)

iso3 non-inline ref vs wp:el.

Can another iso3 parameter be added (iso3i?) which doesn't create an automatic external link, so we can use a standard inline ref per wp:v and wp:el? -- Jeandré, 2011-07-30t13:45z

Sure we can. But what is it exactly that you want, and how would it look on the page? — kwami (talk) 19:14, 30 July 2011 (UTC)

A parameter that can have an inline reference, unlike the current iso3 parameter which automatically makes an external link and prevents inline references. -- Jeandré, 2011-08-02t12:29z

Colors need to be fixed

The background colors for some of these languages families need to be fixed. Hmong–Mien is on almost pure red, which makes the text difficult to read. The pure yellow and green colors need to be changed as well. These should all be light, soft, subtle colors. Not colors that make your eyes bleed :) Kaldari (talk) 19:06, 30 July 2011 (UTC)

Yes, I don't like that red either – nor the one used for ST. Why don't you propose new colors here, and if people like them, we can adopt them? There's a lot of variation in how browsers and screens display colors, and some of your or my choices may look bad to other people. — kwami (talk) 19:11, 30 July 2011 (UTC)

I tried to fix some of the worst ones. It could still use more work though. Kaldari (talk) 00:36, 31 July 2011 (UTC)

I kinda liked the old colors. Not a really major detail, though. I am missing the suggestions that kwami proposed, though.

Peter ^Isotalo 12:30, 1 August 2011 (UTC)

I didn't make any suggestions, except for saying I didn't like the jarringly bright red. I don't care for some of the new colors either: too washed out, not distinct enough. — kwami (talk) 10:29, 7 August 2011 (UTC)

notices now opitional

Since we don't actually have IPA on all that many articles, and there's been no way to cancel the often redundant SL links if the parameter 'signers' is used, I've made both optional as the Indic-font notice is.

However, there is an IPAChartEng parameter embedded in {{Infobox language/IPA notice}} that adds a link to the IPA-for-English chart. Since I don't know of any articles where that is used, I haven't tested to see if it still works. It's not included on our doc page, so if it is used we should note it there as well.

I was thinking of using a bot to cancel the IPA notice if there is no IPA on the page, but that would be difficult to maintain as new stubs are created. Easier to make the notice optional and to use a bot to add it in wherever there is IPA, and then to add it in manually when expanding an article. (There are 3033 articles which transclude this template but not {{IPA}}, 1127 which transclude both.)

If there are no objections, I will make a bot request. — kwami (talk) 21:12, 30 July 2011 (UTC)

Bot ran today. — kwami (talk) 10:28, 7 August 2011 (UTC)

revamping maps & images - will take some time to update

All of our maps are manually sized, which makes it difficult to keep them in sync w changes to the info box. Per the lang family box, I've made a default size of 150px for the image and 350px for the map; free free to adjust. Manual overrides w 'imagesize' and 'mapsize'. True captions for both. Previous 'caption' now called 'imageheader'; people kept using it as a normal caption. — kwami (talk) 08:58, 6 August 2011 (UTC)

I thought I could get through this quickly, but the server has gotten very slow and there are 400 transclusions. It's not fixing them up so much as waiting forever for them to save. So I'm going to go through and do the big languages and leave the rest for tomorrow. — kwami (talk) 09:35, 6 August 2011 (UTC)

• Do we want the infobox and map sized to the same scale (both ems or both px) so that they match width on all browsers?
• Do we want to add 'border' and 'alt' parameters for the map and image? They had been used in a minority of cases, but are no longer supported. We could make the caption auto-double as the alt text, since we are supposed to have s.t. in case the image d n load.

— kwami (talk) 10:26, 6 August 2011 (UTC)

Alt text is not the same as the caption, almost ever. This should be added as a separate set of variables; "alt" and "mapalt", probably. --Izno (talk) 19:59, 6 August 2011 (UTC)

Okay, will add that eventually. Hardly any of our images or maps used it anyway.

Only ten articles skipped family levels (you can skip child2/dia2 and child3/dia3 will still display, but you can't do that with fam2 and fam3), and only half of those had anything worthwhile that wasn't displaying. I thought that was going to be a mess. Next catching unsupported or misspelled parameters, but there's a bug in AWB I need to iron out first. — kwami (talk) 18:11, 7 August 2011 (UTC)

I've added alt text fields. Did you keep a record of which articles you removed alt text from? Kanguole 11:21, 17 September 2011 (UTC)

unsupported parameters

purging transclusions of parameters not documented here.

caption, official, state, altname, country, spoken in, family color, writing system, Script Writing, IPAChartEng, Atlas UNESCO country / language, posteriori (on non-conlangs), status ('endangered'), codic, locality, Language Family, GRN, SIL, sil, othername, native_name, nativeename (quite a few), iso6, linguasphere, Classification, title, extinction, language extinct, pronounced, spoken, map_legend, dial1, l1x (for llx), name2, Mewati.com, scale (w coords), fontcolor, regions, w.

exceptions:

• 'protolanguage' at Armenian (left in)
• 'iso3comment' at Zapotec languages (seems to work)

Apart from recent changes like 'caption', 'state' was probably the most numerous error.

'altname' might be a good thing to add. This would make conversion from langfam easier, and most of our entries under 'nativename' are not the native name.

— kwami (talk) 13:29, 10 August 2011 (UTC)

Also 'alt' text for image (as in previous section), currently unsupported but needed for accessibility reasons. — kwami (talk) 08:15, 3 September 2011 (UTC)

Ethnicity parameter

I've changed it to read as "Ethnicity for which this is the ethnically native language". Perhaps this is somewhat cumbersome so I expect feedback. It appears that some editors don't understand its usage and use, as I reckon was the situation lately with the change in the article English language. That's why I thought it may need some rewording. --biblbroks (talk) 11:35, 15 August 2011 (UTC)

Change the image!

File:Anglophone map.png — Preceding unsigned comment added by ThisguyYEAH (talk • contribs) 03:10, 3 September 2011 (UTC)

Edit request from 99.254.27.102, 21 September 2011

The Sicilian language is also widely spoken in Toronto, which is known to have the highest Italian diaspora in the world.

99.254.27.102 (talk) 13:57, 21 September 2011 (UTC)

Not done: This can be fixed on the article you were looking at, which probably isn't protected. — Bility (talk) 20:14, 22 September 2011 (UTC)

Colour choice

The colour for Sino-Tibetan was recently changed from  Tomato  to  IndianRed . The former colour matched the map in the Sino-Tibetan article, and is also the family colour used on the es, ja, ru and zh wikis, while the darker colour is harder to read. I think it should be changed back.

Also, if it is intended that the rows of {{Infobox language/quilt}} should be variations on yellow, green, red and blue, could Tai–Kadai and possibly Hmong-Mien have more reddish colours? At present Tai–Kadai might be mistaken for a family from the Americas. Kanguole 11:56, 2 October 2011 (UTC)

Could you suggest some colors? Yes, IndianRed is a bit dark, but Tomato is horrible to read too. Maybe something not so glaring? — kwami (talk) 08:44, 5 October 2011 (UTC)

I thought Tomato was fine: light enough for easy reading of black text and it had the virtue of matching the widely used map. But maybe  Salmon  or  LightSalmon  are similar but less bright. Kanguole 09:11, 5 October 2011 (UTC)

Salmon looks good on my browser. (Tomato is almost blinding.) And it's a good match for that map. — kwami (talk) 12:41, 6 October 2011 (UTC)

Thanks, much better. Kanguole 13:08, 6 October 2011 (UTC)

Request for a parameter: ISO 639-6

To represent dialects. ––虞海 (Yú Hǎi) ✍ 13:26, 6 October 2011 (UTC)

automated referencing?

I was wondering if we might want to automate the speaker-data references, since that's what people are most likely to question. I was thinking of s.t. like this:

If there is no manually input reference in the 'ref' field, and

if there is an ISO3 code,

then the 'ref' field would be auto-completed with <ref name="ethno">{{Ethnologue|{{{iso3}}} }}</ref>

That way every single language article with an entry at Ethnologue would have an overt ref, rather than just the semi-hidden iso3 link. It would also make it obvious that that's where our data came from (or at least it should; there's a lot of reference falsification in our articles) and would hopefully encourage people to cite their data if they update the population.

I would probably want to change the 'no date' warning to 'date missing' if we did this, since 'no date' (or 'undated') can be used for when Ethnologue is lacking a date. (Should probably automate that, too: a set response if s.o. enters 'none' in the 'date' field.)

Does that sound worthwhile? — kwami (talk) 03:59, 15 November 2011 (UTC)

(there would be some problems with dialects that are listed under the iso3 code of their parent language, but that's just a matter of cleaning them up.)

Also, should we have an 'era' field for ancestral languages, such as Middle English, that never went extinct but are no longer spoken? It's weird to have to use either existing field. Or should we just use 'speakers', and give the era when it is judged to have had native speakers? — kwami (talk) 04:53, 15 November 2011 (UTC)

It seems to me that automatically added references would give an unjustified appearance of assurance. If an editor fills in |speakers= and a later editor fills in |iso3=, the presentation would imply that the speakers figure was supported by a reference, which neither editor had stated directly. At the very least it would need very precise documentation, but I think even that wouldn't be enough. Kanguole 13:06, 15 November 2011 (UTC)

I don't see any difference between that and the situation we have now, where one editor puts in a speaker population of, say, 4 million and refs Ethn, and then a second editor changes the figure to 16 million and leaves the Ethn. ref. We have hundreds of articles with population inflation and falsified references, and the only way to fix it is to go through all 4,000 articles with iso3 codes and check. (Something for our Wikiproject to team up on, perhaps? 40 people checking 4,000 articles is manageable.) If we make the ref accessible, do you think general readers might check some of them for us, and point out unjustified figures? — kwami (talk) 05:26, 16 November 2011 (UTC)

In the case you describe, the second editor has changed article text so that it no longer matches the explicitly attached reference – they have falsified a reference. In my example, it would be the template that was giving the impression that the reference supports the number, when no editor had asserted that it did. As for ease of checking, the infobox already has a link to the Ethnologue page, which can be reached in one click, instead of two going through the generated reference. Kanguole 12:09, 16 November 2011 (UTC)

Actually, it takes two clicks either way. But general readers often don't recognize the existing link. Making the link visible would allow more readers to spot errors. That was the reason we were supposed to add visible references to all our articles, but no-one's gone through them and done it, presumably because it would be too much work. — kwami (talk) 15:55, 16 November 2011 (UTC)

OK, I miscounted the clicks, but I don't believe the solution to poor referencing is to add 4,000 unchecked references. Kanguole 16:17, 16 November 2011 (UTC)

I'd like to get our Wikiproject together to check the # speakers and dates. It would only take a couple weeks if we get a reasonable number of volunteers. But I suppose we could make a semi-manual ref: say, entering "e16" would trigger the template, making it less time consuming to go through them than the failed manual version; that would also let us know which articles are still on ed. 16 when the 17th comes out. — kwami (talk) 20:36, 16 November 2011 (UTC)

Patrick at WikiMedia helped me code a manually triggered auto-ref. If you enter ⟨e16⟩ in the 'ref' field, you'll get a footnote based on the ISO code at iso3. Given the number of 2ary codes at some articles, I didn't want to link to all of them, so if there is nothing at iso3, you just get a generic reference and link to Ethnologue. Currently it's called whatever we have under 'name'. This should make it convenient for people to document that it's supported at Ethn. — kwami (talk) 16:50, 18 November 2011 (UTC)

Oh, another difference: the link in the infobox connects to the ISO page. The link in the footnote connects to the language article. If the ISO page has been changed since the language article, it may no longer connect, so it's not always easy to access the Ethnologue language article from our info box. — kwami (talk) 16:53, 18 November 2011 (UTC)