Language committee/Archives/2008-01

For a summary of discussions, see the archives index.

Spanned discussions

The following discussions span multiple months and are archived in the first applicable archive:

"Wikipedia Korean Hanja" (June 2006)
"Wikipedia Ottoman Turkish" (October 2007)
"Wikipedia Classical Japanese" (November 2007)
"Wikiversity Greek" (November 2007)
"Code conflict between Alemannic Wikipedia and Tosk" (December 2007)
"Wikisource Low Saxon" (December 2007)

Wikinews Hungarian

The request for a Hungarian Wikinews was approved.

Gerard Meijssen (GerardM)
04 January 2008 16:48

<this user has not agreed to public archival.>
Tangotango
06 January 2008 23:46

I agree, they are quite active and I expect the number of articles per day to increase if this becomes an actual project.
Bèrto 'd Sèra
07 January 2008 13:54

Ok 4 me, too
Jesse Plamondon-Willard (Pathoschild)
09 January 2008 03:01

< http://tools.wikimedia.de/~pathoschild/ls-testanalysis/?prefix=Wn/hu >
The activity with regards to content creation since November looks good, although I'm a little concerned that none of the editors seem to stay longer than a month. However, there are existing Hungarian wikis that are very active, and there are enough interested editors that I don't think it will lose momentum.
I think they're ready for approval.
Gerard Meijssen (GerardM)
CC Florence Devouard (Board chair)
14 January 2008 05:37

<this user has not agreed to public archival.>

Wikipedia Sakha

No decision was taken on the request for a Sakha Wikipedia.

Gerard Meijssen (GerardM)
05 January 2008 05:34

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
06 January 2008 17:30

Hello,
I'm working on the new analysis script which provides comprehensive statistics for test projects (not just for the last 30 days). Please hold any decision for a few days until it's complete and I can generate the better statistics.
Gerard Meijssen (GerardM)
06 January 2008 17:55

<this user has not agreed to public archival.>
Bèrto 'd Sèra
06 January 2008 18:16

We also still have to check that it really IS written in sakha.
Jesse Plamondon-Willard (Pathoschild)
06 January 2008 18:38

~~Bèrto~~GerardM,
The new analysis script I'm working on uses the toolserver database to get a huge amount of data about the entire lifetime of the project. It generates statistics about edits (in particular new pages and minor edits) and distribution by user overall, and for every month of its lifetime. It also generates comparative statistics showing trends (for example, the overall number of edits or users per month).
The statistics are not all-important, of course, but they're certainly a powerful tool for informing decisions. Sometimes they *are* binding, too; I don't think we'll ever accept a project with zero editors. ;)
Bèrto 'd Sèra
06 January 2008 18:45

LOL, no, that much we are not going to accept :)
Sabine Cretella
07 January 2008 01:51

Hmmmm ... are we about real people and community or software and numbers?
There are people who work mainly in the background of a project, that is: they do proof reading offline and do not even go to the wiki ... so on one hand you have those who make a small edit and save each minute and then those who work offline and copy and paste the whole text after it is finished or even pass the text on to somebody else to upload it and don't even care about attribution. This happens often with small wikipedias. Statistics are not good to show such situations and could influence decisions in a negative way, something I'd not like to see happen.
Statistics IMHO can never be binding because each project is a separate community and therefore separate case.
Jesse Plamondon-Willard (Pathoschild)
07 January 2008 02:36

Sabine,
Some of those problems can be addressed by statistics, like measuring how much content is added or removed; others will show up in discussion if we're transparent about our decision (ie, the editor will tell us). Further, the statistics are best viewed over the long term to show tendencies (which the new script does), not simply as counts; is the project growing or stagnating, or is one editor putting on a short burst of activity to impress us?
I agree that we shouldn't *only* look at statistics in judging a request, but I think they should definitely be one tool we use in making informed decisions. There's generally no single tool or criterion that we can consider alone.
Jesse Plamondon-Willard (Pathoschild)
08 January 2008 22:29

< http://tools.wikimedia.de/~pathoschild/ls-testanalysis/?prefix=Wp/sah >
They seem ready for approval.
Bèrto 'd Sèra
09 January 2008 00:15

Hi!
Nice tool Jesse :) FMI, why are some editors' names striked? I will proceed to look for a source that can duly conform this as Sakha asap.
Jesse Plamondon-Willard (Pathoschild)
09 January 2008 00:33

Bèrto,
Thanks. :) Users that have less than 10 edits are struck out; it's just a visual aid to distinguish active editors from bots and passing editors.
Bèrto 'd Sèra
09 January 2008 00:37

LOL, I kinda should have guessed it :) I clearly need more sleep :)
Jesse Plamondon-Willard (Pathoschild)
08 February 2008 09:54

Hello Berto,
Any luck finding a Sakha source? I'd like to forward this request to the board for approval with this weekend's batch.
Bèrto 'd Sèra
09 February 2008 05:21

Hi!
sorry, I had totally forgot about it, my fault. Apart from the baby I had a very unpleasant disk crash and I spent the last days finding out how many bad words fit in a minute. I'll send the mail by today and copy LangCom.
Jesse Plamondon-Willard (Pathoschild)
19 February 2008 16:05

Hello Berto,
Do you have any news about this? The next batch is almost ready for approval, and I'd really prefer to include the Sakha Wikipedia in this one so they don't have to wait another month.

Wikipedia Shakhordoy

The request for a Shakhordoy Wikipedia was rejected.

Jesse Plamondon-Willard (Pathoschild)
10 January 2008 03:22

Hello,
I've rejected the Shakhordoy Wikipedia. It has no ISO 639-3 code, its test project on the Incubator was rejected by the community, and Google returns no search results for such a language.
< http://meta.wikimedia.org/wiki/Requests_for_new_languages/Wikipedia_Shakhordoy >

Wiktionary Pitjantjatjara

The request for a Pitjantjatjara Wiktionary was conditionally approved.

Gerard Meijssen (GerardM)
10 January 2008 21:27

<this user has not agreed to public archival.>

Wikinews Simple English

The request for an Simple English Wikinews was rejected. This was accidentally proposed twice due to the archival backlog at the time.

Jesse Plamondon-Willard (Pathoschild)
11 January 2008 22:14

Hello,
I propose the rejection of the Simple English Wikinews. It does not have an ISO code (there is one for English, which has a wiki, but not for "Simple English"), it's completely redundant with the existing English Wikinews, and "Simple English" is also vague and poorly defined, with no standard vocabulary or orthography.
If there is no objection in 48 hours, I'll mark it as rejected.
Bèrto 'd Sèra
12 January 2008 00:42

Ok 4 the rejection.
Gerard Meijssen (GerardM)
12 January 2008 03:49

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
12 January 2008 14:51

Hello,
The request is already unnecessary: every editor is more than welcome to write simple English on the English Wikinews, or even simplify other articles; there's no policy that says they must use difficult words. Further, we do not consider requests predating the subcommittee when deciding current requests. Furthermore, it fails nearly every requirement in the policy.
Jesse Plamondon-Willard (Pathoschild)
30 January 2008 12:40

Hello,
I propose the rejection of the Simple English Wikinews. This language is an undefined subset of English arbitrarily invented by the contributors, it has no ISO 639 code (although there is one for English itself), and it is not sufficiently unique that it could not exist on the English Wikinews.
There are four existing simple English wikis (Wikibooks, Wikipedia, Wikiquote, Wiktionary), but two are largely inactive and would not pass our activity requirements for approval. We have also set a precedent that wikis created before the policy was implemented are not taken into account. Furthermore, approving simple English sets a precedent for any other abitrary language (simple Megleno-Romanian or scientific Gilaki?).
If there is no objection, I'll mark it as rejected within 48 hours.
Shanel Kalicharan (Shanel)
30 January 2008 16:52

I concur. Based on our current policy this is clearly not an acceptable project, and accepting it would create said bad precedents. Since they seem to have an active test project on Wikia (yay Incubator Plus!) they can continue to write there in Simple English with no great loss.
That said, I challenge you to write in Simple English, good sir. ;)
Bèrto 'd Sèra
31 January 2008 00:26

Yes J agreed

Wikipedia Karakalpak

No decision was taken on the third request for a Karakalpak Wikipedia.

Gerard Meijssen (GerardM)
12 January 2008 13:17

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
12 January 2008 19:41

Hello,
The status page is at < http://meta.wikimedia.org/wiki/Language_subcommittee/Status/wp-kaa >. You can see the test project statistics by clicking on "automated statistics" on the line about needing a test project.
The test project needs more editors. They still have some missing messages to translate, though they're almost done.

Extremaduran Wikipedia

The request for an Extremaduran Wikipedia was approved.

Jon Harald Søby
01 December 2007 18:07

Hi all!
I propose the final approval of Wikipedia in Extremaduran. They are done localizing the interface, and have almost 400 articles ready in the Incubator, as well as an active community. As far as I can tell, all our requirements are met, so they should be approved.
Jesse Plamondon-Willard (Pathoschild)
02 December 2007 02:38

Hello,
Some very important interface messages have not been translated, including the session loss error message, email confirmation messages, and the watchlist lag warning. Once these (and a few others) are translated, I'm all in favour.
Bèrto 'd Sèra
02 December 2007 05:31

Yes, that stuff needs being addressed
Gerard Meijssen (GerardM)
12 January 2008 15:11

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
12 January 2008 17:08

Agreed.
Jesse Plamondon-Willard (Pathoschild)
20 January 2008 07:06

Hello,
The requirements have been met for the Extremaduran Wikipedia request.
They're missing 8 'most used' messages, but these are particularly difficult to translate because they include standards names and computer jargon like "RSS feed" and "bytes". For example, the French translation of "RSS feed" is "Flux RSS", which is almost unintelligible even to me as a native French speaker— the fact that "RSS" is an untranslatable English acronym doesn't help. :)
Given that these are trivial messages for tasks unrelated to editing or admins, and that translating them would not do much good (see above), I suggest that the localization is quite sufficient for final approval. If nobody objects within 48 hours, I'll notify the board of approval.
< http://meta.wikimedia.org/wiki/Language_subcommittee/Status/wp/ext >
Bèrto 'd Sèra
20 January 2008 08:47

LOL funnily enough we translated exactly the same way in PMS. Anyway, as matter of fact the only word that counts for users in that msg is RSS.
"Feed" is totally unusable for non English speakers, but unless they know the whole idea about feeds it's the whole concept that makes no sense at all, in this case.
I don't think many will translate "byte" anyway. Basically it's like Joule, a unit of measure. All the localization we made, for such things, consisted in taking away the plurals. But that's also because 90% of our names have no plural form at all.

Increasing localization requirements

The localization requirements were expanded after additional discussion on the Foundation-l mailing list.

Gerard Meijssen (GerardM)
12 January 2008 04:19

<this user has not agreed to public archival.>
Jon Harald Søby
12 January 2008 07:20

I agree with you. However, I think we should add nuances to this. A model I've been thinking about is this: for the first time a language is to get a wiki, it only needs to translate the most used messages. For the second wiki in that language, they have to have the entire interface (minus exif messages, just because I intensely hate exif messages) translated, and for the third wiki they need to have all extensions used by Wikimedia translated. Does that seem reasonable?
Gerard Meijssen (GerardM)
12 January 2008 07:31

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
14 January 2008 18:00

Hello,
The current policy requirements call for the localization of MediaWiki, which includes 1730 messages. This is already a daunting task which many have complained about, but it makes MediaWiki usable in that language.
Doing as you suggest increases that to 2300 messages, many of which are obscure and difficult to translate (like "Assert failed" or "Begins with combining mark"). These extensions are largely unneeded for editing in MediaWiki, such as boardvote, bookinfo, categorytree, or revreview. This has a major impact on the difficulty in achieving a new project, but has little gain in usability.
I propose educating users on localization files, and encouraging them to put any local translations into the files to share them with other wikis. However, I object to changing the current localization requirements.
Bèrto 'd Sèra
15 January 2008 01:59

Hi,
I will vote against japanese requests if they don't have localisation done. It's a HUGE community, there is no excuse for being "partially localised".
It is reasonable to keep localisation requirements in steps, so communities can grow in time. But steps are steps. After the first you have a second, etc...
We are publishing an open source platform and need the people to help us, not only to use us.
Gerard Meijssen (GerardM)
15 January 2008 02:03

<this user has not agreed to public archival.>
Gerard Meijssen (GerardM)
15 January 2008 01:46

<this user has not agreed to public archival.>

Constructed languages

No decision was reached on the policy for constructed languages. This discussion is closely related to several discussions about historical and constructed languages:

Internal discussion:
- "Wikipedia Ottoman Turkish" (October 2007);
- "Wikipedia Volapük" (October 2007);
- "Wikisource Volapük" (November 2007);
Public discussion:
- [Foundation-l] Fwd: Tokipona-Wikipedia, part one (January 2008)
- [Foundation-l] Fwd: Tokipona-Wikipedia, part two (January 2008)
- [Foundation-l] "Historical" languages and constructed languages (January 2008)

Discussion 1

Jesse Plamondon-Willard (Pathoschild)
13 January 2008 01:27

Hello,
Discussion at <http://meta.wikimedia.org/wiki/Talk:Langcom#Distinguishing_living_languages_from_dead_languages> has raised an important point.
Currently, we do not approve requests in historical languages because there are no native speakers in that community; creating content in languages nobody speaks natively does not help spread the content to more people. A project in a historical language would, however, draw editors away from projects in their native languages. Instead, wikis should be opened in the languages those editors speak and write natively, to make the content available to their language communities.
In my opinion, the same principle applies to constructed languages like Volapük which have no native speakers; the editors should open the projects in their native languages instead. (The way the policy is currently worded, constructed languages would also be rejected.)
What do you think?
Gerard Meijssen (GerardM)
13 January 2008 02:49

<this user has not agreed to public archival.>
Bèrto 'd Sèra
13 January 2008 03:11

Hi!
Well... from a strictly logical pov, the argument about native speaker cannot apply in principle to a constructed language. These are language created to be a "lingua franca", something anyone can use but nobody can claim to own. Call it OS languages, if you wish.
When it gets to real life, I never cared to learn any of them, and most people seem to go my way, yet this are individual choices, and cannot be used as a policy.
It's true that many such languages have a very small "user base" (I'd rather use this definition, than "speakers"), yet this also applies to many an OS software project.
I guess the only criterion one can use is to ask for evidence of a minimal "user base" for the project. Then it really depends on how good they are in marketing their "lingua franca", which does not dipend on us.
Gerard Meijssen (GerardM)
13 January 2008 03:32

<this user has not agreed to public archival.>
Gerard Meijssen (GerardM)
16 January 2008 16:31

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
16 January 2008 17:05

Hello,
Kotava was founded by one person in 1975, with major reforms in 1988 and 1993. It is estimated to have 40 (non-native) speakers worldwide. This is according to its French Wikipedia article; the English Wikipedia article was deleted as non-notable, and it has no Ethnologue article.
Lingua Franca Nova was founded by one person in 1965, and first presented on the Internet in 1998. It is estimated to have between 30 and 180 users worldwide; it is largely or entirely used on the Internet, so these numbers are based on forum posts, wiki edits, and group membership count. The language development is driven by a Yahoo! fan group and a wiki. It has no Ethnologue article.
I think neither of these languages has a "reasonable degree of recognition" as required by the exception for constructed languages in the policy. Furthermore, the second in particular has a very high degree of the fan-driven neologism creation you use as an argument against allowing historical languages.
Gerard Meijssen (GerardM)
16 January 2008 17:12

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
16 January 2008 17:51

They don't need permission to localize or start a test project, but you verified Lingua Franca Nova as eligible for a wiki. As I've said above, it does *not* seem to meet this criteria:
"The proposal has a sufficient number of living native speakers to form a viable community and audience. If the proposal is for an artificial language such as Esperanto, it must have a reasonable degree of recognition as determined by discussion."
Gerard Meijssen (GerardM)
16 January 2008 17:54

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
16 January 2008 18:12

Hello,
It does not seem to have been particularly difficult for Kotava. Their request was filed by the creator of the language, cites one website (the creator's site) and one wiki as the entirety of the language's literature, and uses the same sites as its references. It states the number of speakers as 40, distributed "where the speakers live".
ISO 639-3 was recently updated to add 59 new languages and 71 split from previously grouped languages. Their requirements don't seem very draconian; the only difficulty seems to be waiting times. An ISO 639 code does not seem to be very good evidence of recognition.
Jesse Plamondon-Willard (Pathoschild)
16 January 2008 18:23

The link to the request in question is <http://www.sil.org/iso639-3/chg_detail.asp?id=2007-145 >, by the way.
Gerard Meijssen (GerardM)
16 January 2008 18:26

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
16 January 2008 18:46

You did officially mark Lingua Franca Nova as meeting the requirements for a wiki. I don't see any evidence of significant recognition, as required by the exception in the policy I quoted. As I argued above, having an ISO code is not very good evidence of such recognition. Neither language seems to meet the requirements.
If there's no such evidence, we should either reject this request, or change the policy to remove the requirement or disallow constructed languages (which would fix inconsistent treatment as related to historical languages).
Gerard Meijssen (GerardM)
16 January 2008 18:51

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
16 January 2008 19:15

Hello,
There is a specific exception for constructed languages, so that they do not need living native speakers if they have international recognition based on discussion (like Esperanto). These two languages do not appear to meet this exception. Activity on their wiki is not relevant if this requirement is not met, just like the activity on the Ottoman Turkish test project was not relevant because it failed the requirement for native speakers.
You suggested that having an ISO code is evidence of recognition, which is not the case as I've argued above in relation to Kotava.
They can certainly localize on Betawiki, but they should not be marked as eligible until there is subcommittee consensus to do so or until we change the policy.
Gerard Meijssen (GerardM)
16 January 2008 20:19

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
16 January 2008 23:05

Hello,
No, I will not forget about native speakers. The single overriding goal of a Wikimedia project is to provide information to people <http://wikimediafoundation.org/wiki/Vision >. Opening projects in languages with no native speakers is an academic exercise that does not further this goal. If I speak English natively and learn Esperanto, I will not refer first to the Esperanto Wikipedia if there's an English Wikipedia.
Constructed languages are a special case, not because they're constructed but because they're intended to cross language communities (linguas franca). If many contributors from different languages want to put their efforts together to create a resource in a common language to more effectively reach more people, that is wonderful and fits perfectly within our goals. However, this is not the case for a language popularized less than 10 years ago and having 30 speakers worldwide, only a subset of which can be expected to edit. A wiki in such a language is primarily intended to promote the language rather than provide information to more people.
For this reason, constructed languages do not need native speakers, but they *do* need to meet the requirement that applies instead. They must have sufficient recognition and usage to further our mission. I have presented several different arguments above about why these two languages do not.
You raise a concern with new terms, but new terms are continuously created in all languages that are in usage, be they historical or constructed or living. For example, "vira ordinatralia" ("computer virus" in Latin) takes on a different meaning than the two words put together to represent a modern concept. Similarly, "cracker" is a new term in English recently invented by a single person to distinguish good-faith and malicious hackers. You'll notice that the term "cracker" is used with this meaning on the English Wikipedia.
You also raise the question of "continued usage" without defining it. Does academic or online enthusiast usage count, as with Kotava or Ottoman Turkish? The Ottoman Turkish requesters, as I recall, provided links to several current websites written in Ottoman Turkish; far more numerous and comprehensive than the sites written in Kotava, for example. Ottoman Turkish is also taught in schools, while Kotava is not. I do not see anything about Kotava or Lingua Franca Nova that makes them more suitable for a project than Ottoman Turkish.
Neither Kotava nor Lingua Franca Nova appear to conform to the current requirements, not even with the exception for constructed languages. Either we find evidence to determine otherwise, change the requirements, or reject the request.
Shanel Kalicharan (Shanel)
16 January 2008 23:05

Yo,I think you are both agreeing that constructed languages do not have a requirement for native speakers (although I am a bit confused as to what Mr. Jesse means by "There is a specific exception for constructed languages, so that they do not need living native speakers if they have international recognition based on discussion (like Esperanto). These two languages do not appear to meet this exception." Other than that I'm pretty sure you're both talking about two different things (it seems that way to me anyway). I think Jesse is talking about his objection to these two constructed languages, and about using ISO codes as a metric for how well-known/used a constructed language is, and you about linguistic continuity and historical languages.
IMO, Jesse tends to look at things from a project perspective, while Gerard tends to look at things more from a foundation goal/usability perspective. Hence why you butt heads quite often (although that's not necessarily a bad thing). Jesse and Gerard, is that correct? I now return to the sidelines. :)
Shanel Kalicharan (Shanel)
16 January 2008 23:16

I'm pretty sure he means continuous spoken usage. It could be taught in school and used academically without new words being invented, but if it's being spoken, then new words are being made all the time. Now I go away again.

On Jan 16, 2008 11:05 PM, Jesse Martin (Pathoschild) wrote:
You also raise the question of "continued usage" without defining it.
Does academic or online enthusiast usage count, as with Kotava or
Ottoman Turkish? The Ottoman Turkish requesters, as I recall, provided
links to several current websites written in Ottoman Turkish; far more
numerous and comprehensive than the sites written in Kotava, for
example. Ottoman Turkish is also taught in schools, while Kotava is
not. I do not see anything about Kotava or Lingua Franca Nova that
makes them more suitable for a project than Ottoman Turkish.
Bèrto 'd Sèra
16 January 2008 23:19

This is a part of the policy that I would really deem as "utterly subjective". Who and how can present data we can use, apart from real activity as observed in the incubator?
Define "reasonable degree". Being dexcribed in a wiki article is NOT a recognition. Lots of anime charachters are and this doesn't make them real. I'm okay with setting parameters for minimal recognition, but if they are used they must be clearly identifiable.
Ethnologue silence is interesting. From one hand they give them a code, and from the other they decide they don't care... Then again, maybe ethnologue cares for ethnos (i.e., natural languages)... while SIL cares for languages in general, no matter where they are from. Just throwing out a possible logical explanation, in reality I have no idea of what can stand on the back of such a schizofrenic behavior.
Bèrto 'd Sèra
16 January 2008 23:38

Hi!
Yes, the goal is quite clear here. There is no need to develop a post office if there's no people living on the island, to whom someone from the mainland may want to write. This is pure logic.
New terminology is a boundary fixed by ISO. A somewhat "closed" language is not supposed to experience semantic growth, while living languages are. Constructed languages (when alive) are naturally expected to experience such growth. Whether they are acceptable for a WMF project cannot be based on this issue. Constructed or not, a living thing is changing, a museum specimen is not.
I welcome extended feasibility criteria for constructed languages, yet "reasonable" is not a criterion. Once translated to human speech it only means "when LangCom is happy with it". I'd rather have a set of criteria saying that a constructed language needs:
1) a valid ISO 639-3 code
2) <fill in>
We have made NPOV a religion; we are now supposed to believe in it, aren't we?
Jesse Plamondon-Willard (Pathoschild)
16 January 2008 23:41

Bèrto,
Indeed. The problem is that there's really no objective way I can see to distinguish historical languages (which have no living native speakers, but are frequently spoken by a small number of enthusiasts online or in academia) from most constructed languages (which can be described exactly the same way). GerardM raises the issue of creating new terms for the wiki, but constructed languages consist *entirely* of the creation of new terms by a small number of enthusiasts (who edit the Wikipedia-to-be, in the case of Lingua Franca Nova, making the Wikipedia a central place for language development).
This is partly shown by the way usage estimates are mapped directly to the number of editors on the official wiki or members in the Yahoo! group. There doesn't seem to be language communities for these two languages, so there's nobody for the speakers to talk to except other enthusiasts online.
The only reason we have this vague exception at all is because we could not reach a consensus on constructed languages when the requirement for living native speakers was introduced.
Bèrto 'd Sèra
16 January 2008 23:55

All true. But can there be such a community at all? I mean, a constructed language is a tool, pretty much like software. You cannot expect any new OS tool to have anything more than a small group of enthusiasts, in the beginning.
They obviously see a wiki as a marketing tool. This is not bad in principle, as 99% of our small languages see their wikies exactly along the same line. Yet I do agree that the WMF may not desire to invest resources on a one-man band. Swahili maybe has but one editor, but whatever he/she writes tomorrow can be potentially read by millions of people, whereas a constructed language is more unclear as per future usability.
Yet, esperanto is a success story, and where you had one good product you can have many more. I do mean "product", because this is exactly what a constructed language is.
My suggestion is that for cosntructed languages we may want to have FULL localization since the very start, plus a proven number of editors and activity we can track with your stats. And we should fix clear numbers for them. Also, I'd want to see this stuff happening for (say) 6 months in a row. If they go under the mark during one month, the count starts from zero again.
It's a product, so basically it can be assessed in industrial units.
Jesse Plamondon-Willard (Pathoschild)
17 January 2008 16:19

Hello,
I've moved the Lingua Franca Nova request back to the discussion phase in the meantime, given that we're still discussing the requirements that apply to constructed languages and there's no consensus yet on this language. This is precisely the kind of situation where we want to invite comments and arguments on the request page. :)
Gerard Meijssen (GerardM)
17 January 2008 16:22

<this user has not agreed to public archival.>
Gerard Meijssen (GerardM)
17 January 2008 16:24

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
17 January 2008 16:31

Hello,
Please read the emails I have sent in this thread; I have provided pages of arguments which you seem to have ignored so far. These languages do not meet the current exception for constructed languages. The requirement asks for more that a language simply being constructed.
Adding new words to a language does not create a new language. The ISO 639 standard classifies languages, it does not freeze them in time.
Gerard Meijssen (GerardM)
17 January 2008 16:52

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
17 January 2008 17:29

Hello,
A "dead" language means that it is no longer in use for everyday communication, not that it has ceased to exist.
An example of this is Latin, which has no living native usage, is not even used by scholars as an everyday language, and is classified as extinct by SIL < http://www.ethnologue.com/show_language.asp?code=lat >. However, it is used for liturgical purposes (because it is considered the most accurate), which includes the arbitrary creation of new words to cover modern concepts. The official Vatican website hosts a list of such invented words, called the Lexicon Recentis Vaticana, at < http://www.vatican.va/roman_curia/institutions_connected/latinitas/documents/rc_latinitas_20040601_lexicon_it.html >.
For more information about Latin as used by the Vatican, see <http://en.wikipedia.org/wiki/Ecclesiastical_Latin >.
A "dead" language can indeed have new terminology without becoming living as a result. This is not relevant to having a wiki for that language.
Bèrto 'd Sèra
17 January 2008 20:44

Yes. They are to be «eligible». Whether they are «elected» or not is matter for discussion, elegibility isn’t.
Bèrto 'd Sèra
17 January 2008 20:46

Jesse, if you have problems wuth SIL, you should discuss them with them :) Dead means Dead. If it's "alive", then it's alive.
I would myself argue that Latin can be classified as "dead" (same reasons you quote), but when it is...
Jesse Plamondon-Willard (Pathoschild)
17 January 2008 21:16

Bèrto,
"Eligible" means that the language is perfectly acceptable as a project, and can be approved once have localization and a test project done. It does not simply mean that it has an ISO 639 code, which is only one requirement. We should not mark a project as eligible until we are reasonably sure we will approve it if the other requirements are met.
I have no problem with SIL; their classification simply does not provide the dead/living classification for constructed languages as it does with natural languages. With regards to constructed languages, we therefore need to look at other references besides the ISO 639 classification (which is always "constructed", whether the language is living or not).
Just to clear up any confusion, I was not arguing that Latin is a living language.
Gerard Meijssen (GerardM)
17 January 2008 22:40

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
17 January 2008 23:36

Hello,
The Ottoman Turkish test project had 1643 pages, 102 editors, and 7340 edits. Even after they were rejected, the test project is more active than many wikis we approve. Nonetheless, the language is historical and the editors were enthusiasts writing an academic language. Their work was redundant, because there were no native speakers; their efforts would be more meaningful on the modern Turkish Wikipedia.
When you show me a language that has 30 users *in the world*, all of which are enthusiasts using a language learned from one website's tutorials, it does not seem "glaringly obvious" that the language is living. The best argument I've seen in favour of allowing constructed languages is that they allow many people across many languages to collaborate on a common resource. This is not applicable when you have such a small number of users in the world.
The goal of a Wikipedia is to provide knowledge to more people, not to promote small invented languages. This is particularly true when, according to the distribution mentioned code request, the majority of users have large Wikipedias in their native languages.
Even leaving that aside, the language fails the policy. We cannot justify marking it as eligible unless we change the requirement, and we cannot justify reducing the requirement further unless we revisit historical languages.
Gerard Meijssen (GerardM)
18 January 2008 04:10

<this user has not agreed to public archival.>
Bèrto 'd Sèra
18 January 2008 04:29

I repeat, all right, then DO define "acceptable". This is NOT a defnition.
Bèrto 'd Sèra
18 January 2008 04:40

Jesse....
The Othman project is NOT to be compared to this. In ANY way. You are saying that 2 liters are more than 3 Kgs... they belong in 2 different classes... damn Jesse, you're supposed to be a programmer, after all...
A constructed language CANNOT in principle have native speakers. So okay, we need numbers. Which ones?
1) state sources. I believe the only source we can use are your stats. If you have any other, you are welcome to present them, but please do it and quit playing with words. We are all busy in life.
2) state numbers. How many of what? For how long?
THIS is a policy. We are not here to discuss whether angels are males or females. We are here to fix criteria. You are sorry for othamns? Fantastic, open a bug at SIL.
Jesse Plamondon-Willard (Pathoschild)
18 January 2008 15:44

Hello,
Historical and constructed languages are not completely different issues. They are both languages. I think it is silly that something created in someone's basement or office a decade or two ago and written by 30 people on a Yahoo! group and a wiki should get better treatment than something with thousands of years of culture and history and literature with hundreds or thousands of writers today. Both have no native speakers, but both have demonstrated non-native usage.
Yet, the natural language is rejected while we argue about what special treatment we should give to the invented language so *that* can be approved instead. I would personally favour rejecting constructed languages entirely, but agreed to a compromise whereby a constructed language is allowed if it is in widespread usage (poorly phrased in the policy as "recognition"), like Esperanto. A language written by 30 people online is not in widespread usage.
I do *not* think constructed languages should get special treatment. I think we should treat them exactly like we do any other language. You obviously think they should get special treatment, so I think it is up to you to define in the policy *how* they do. Based on this discussion, it looks like you just want to exempt constructed languages entirely from the fourth requirement for approval, so that any invented language in the universe with an ISO 639 code is acceptable.
You want me to define what I want before I approve. I want either widespread usage as required by the policy, or a policy change. 30 people online does not qualify.
Jesse Plamondon-Willard (Pathoschild)
18 January 2008 17:39

Hello Gerard and Berto,
I've sent the following email to the other subcommittee members to invite comment. I'm hoping a personal message in their inbox and an introduction to the debate will pull them into the discussion, so we can decide it either way.

-------------------
Hello,
There is a debate on the subcommittee mailing list about constructed languages. We are currently in a deadlock with only three members discussing, so please consider skimming through the discussion and commenting.
Following is an introduction to the issue if you've missed it, but it is of course biased towards my own position. Please read the full discussion for a more balanced view. :)
==Summary==
The subcommittee recently decided that historical languages would not be approved, but constructed languages are currently exempt from the requirement for native speakers that disallows them. Instead, they must have "a reasonable degree of recognition as determined by discussion", a vague compromise requirement.
The issue centers around Lingua Franca Nova, a constructed language popularized in 1998 and having 30 users worldwide, all centered on the official wiki and a Yahoo! group. The language's development is driven by the Yahoo! fan group and wiki. It has no Ethnologue article.
GerardM and Berto favour accepting it, and I favour rejecting it.
===ISO 639 code===
GerardM argues that it has an ISO 639 code, considering that to be sufficient recognition. He argues that historical and constructed languages are completely different classes, and that constructed languages should get preferential treatment.
I argue that an ISO 639 code is not difficult to obtain and therefore does not qualify as sufficient recognition. As an example, Kotava was given an ISO 639 code after a request filed by its creator, citing one website (the creator's site) and one wiki as the entirety of the language's literature, using the same sites as its references, and stating the number of speakers as 40 distributed "where the speakers live". < http://www.sil.org/iso639-3/chg_detail.asp?id=2007-145 >
===Wiki activity===
GerardM argues that they have a wiki with regular activity, so that they are already ready for a Wikipedia. <http://lfn.esef.net/index.php/Special:Recentchanges>
I argue that although there is a Lingua Franca Nova wiki, this does not make them exempt from the other requirements. For example, the Ottoman Turkish test project was the most active test project on the Incubator, but was rejected as failing other requirements.
===Historical and constructed languages different classes because of neologisms===
GerardM considers historical and constructed languages to be entirely different classes, because a wiki in a historical language would need to invent words for modern concepts.
I argue that historical and constructed languages should be treated the same way; for example, historical languages are rejected because they fail the requirements that apply to any natural language. Furthermore, constructed languages are entirely made up of invented words, and a Wikipedia in Lingua Franca Nova would involve prolific invention of new words to expand the language's small vocabulary. I also argue that new words are invented continuously in all languages in use. For example, "vira ordinatralia" ("computer virus" in Latin) takes on a different meaning to represent a modern concept than the two words put together. The Vatican keeps a comprehensive lexicon of new words at < http://www.vatican.va/roman_curia/institutions_connected/latinitas/documents/rc_latinitas_20040601_lexicon_it.html>.
===Native speakers===
GerardM considers native speakers to be irrelevant to the creation of a new wiki (he provides no arguments for the point).
I argue that the single overriding goal of a Wikimedia project is to provide information to more people <http://wikimediafoundation.org/wiki/Vision>, and that opening projects in languages with no native speakers is an academic exercise that does not further this goal. If I speak English natively and learn Esperanto, I will not refer first to the Esperanto Wikipedia if there's an English Wikipedia.
===Continued usage===
GerardM argues that "continued usage" is more important than native speakers.
I argue that many historical languages are also in "continued usage"; for example, there are many Ottoman Turkish websites maintained by enthusiasts, far more than the two websites cited for Lingua Franca Nova. The Ottoman Turkish test project was also far more active than the Lingua Franca Nova wiki. I also ask what he means by "continued usage", and whether that includes online-only usage like Lingua Franca Nova. (He does not respond to the question.)
Gerard Meijssen (GerardM)
18 January 2008 17:58

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
18 January 2008 18:23

The only reason constructed languages do not need native speakers is because the policy says so. That exemption is vague and undefined, as Berto has mentioned, and under discussion. Saying that the policy is that way because the policy says so is not a good argument.
The fact that someone recently invented them does not mean they deserve special treatment. Historical languages are rejected because they have no native speakers. You say that constructed languages obviously don't have native speakers, but that does not mean they obviously should be exempt from needing them. It simply means that they obviously fail the requirement that applies to all (other) languages.
I have repeatedly said that I think we should not allow constructed languages that don't meet the normal requirements. Your problem is not that I have not said so, but simply that you do not read what I write before you respond. I made it pretty clear in my last message: "I would personally favour rejecting constructed languages entirely, but agreed to a compromise whereby a constructed language is allowed if it is in widespread usage (poorly phrased in the policy as "recognition"), like Esperanto. A language written by 30 people online is not in widespread usage."
The same applies to your claim that I have not provided any counterarguments. I have provided pages of arguments; the summary you quoted in your response is over two pages long.
As I have said before, our goal is not simply to approve wikis that are edited; otherwise, we would allow projects in historical languages like Ottoman Turkish, which was an extremely active test project. Our goal is to approve *useful* projects that provide information to native readers.
Sabine Cretella
19 January 2008 01:09

As for Ottoman Turkish: even if they are the most lively project around on incubator: the rules are the same for all.
Now if it is a living language whoever is interested can file a change request with SIL. Once the change is accepted by SIL they can of course get their Wikipedia, but until this happens they can of course also go ahead to work on incubator. I believe it is only relevant that they know they need the amendment of the ISO 639-3 code. Again: we do not determine which language is living and which one not.
Have a great day :-)
Jesse Plamondon-Willard (Pathoschild)
19 January 2008 01:35

Hello,
I agree completely; I was only using Ottoman Turkish as an example to prove the same point you raised, that test project activity does not exempt them from the policy. :)
Gerard Meijssen (GerardM)
18 January 2008 16:23

<this user has not agreed to public archival.>
—
18 January 2008 19:26

<this user has not agreed to public archival.>
Bèrto 'd Sèra
18 January 2008 20:34

Apples and oranges are both fruits. Yet they get placed in different baskets.
It's YOU requiring a restriction for constructed languages (and when you do so you imply you feel quite comfortable in recognizing them as "different"). You'd never care to measure the number of online people for a natural language, just read your message history: YOU NEVER DID. Now you do it. So the difference IS there, no matter the number of U turns you make to hide it.
Make a proposal, and we'll discuss it. But for heaven's sake, make it a fully qualified proposal. I don't care WHY you want a limit, that's your problem. You want something -> you say what you want, and THEN we discuss the practical impact of it. Full stop.
"widespread" is NOT a number. And you are still not saying how and who will measure the number. Jesse... be honest with yourself, call things their names and pls be practical.
Bèrto 'd Sèra
18 January 2008 20:42

Hi!
Now THIS is a criterion. I may agree or not, but it's clear (ant that's why I am able to agree or not).
Just a question: would you say the same for Esperanto? It's not for the sake of discussion, it's a real question.

<this text is quoted from a user who has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
18 January 2008 21:46

The solution I consider ideal is simple: remove the special exception for constructed languages.
I'm willing to accept a compromise objective exception if necessary, but not one that allows a constructed language apparently used only online by about 30 people. I haven't thought of one, or I would have proposed it. (I also don't think it's up to me to invent the better exception for constructed languages, since I think there should be none.)
Bèrto, Could you quote the part of Karen's message that you think is a good criterion? I read it as generally not favouring constructed languages.
Karen, thanks for joining the conversation. :)
Gerard Meijssen (GerardM)
18 January 2008 21:51

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
18 January 2008 22:14

GerardM,
That implies that the only "constructive position" is to approve everything. I'll point out that there are many cases where you disapproved of something.
I've supported my position with several pages worth of arguments. I have indeed considered your position, and I've carefully responded to every single point you've raised. In return, you've consistently failed to even acknowledge the contents of my emails, claiming that I have no position, that they're "absurd", or that you "fail to observe any counter arguments". It is difficult to have a productive discussion with that attitude.
Sabine Cretella
19 January 2008 01:09

Some days ago I talked with Gerard about this project and I thought about it over and over again.
As for constructed languages:
1) the actual policy does not fit, since there cannot be any native speakers.
2) having a Wikipedia serves these languages to become better known and maybe is the only collaborative space they can have.
3) we need special rules for constructed languages and these must be somewhat stricter than for living languages, simply because there cannot be any native speakers.
Therefore for now I would put such a wikipedia on hold until we figured out a particular policy for them.
Bèrto 'd Sèra
19 January 2008 01:35

She is saying NO constructed languages allowed. Full stop.
This is a criterion. The tag say constructed -> we say no.
I will not express an opinion until we have clear alternative models. This is one. Any other?
Jesse Plamondon-Willard (Pathoschild)
19 January 2008 01:38

That is perfectly acceptable to me, as I've said previously. I have no other criteria to propose.
Sabine Cretella
19 January 2008 01:49

Well the actual policy requires native speakers - so anyway: constructed languages are excluded by this point.
That means that if we want to allow for constructed languages we need a different policy. Now Esperanto might come up since some families use it, but Esperanto does not transmit real feelings - you'll soon read a blog about this - there is something I understood lately which is/has been completely wrong in German and other societies. Esperanto has not been talked long enough to be considered a "living language" by SIL, right? It still is considered a constructed language.
So: our statement is no exceptions for anyone - this means: they need native speakers and even if I will repeat it 1000 times, sorry, but or we create a special policy for constructed languages or they will not be able to get a Wikipedia.
Constructed languages with a certain diffusion can help in communication, but they must have "people".
Now when is the point when a living language is to be considered as "not living, but almost extinct"? I suppose 30 writers/speakers is not enough to keep a language alive.
Hmmmm .... <which means "thinking" ...>
Bèrto 'd Sèra
19 January 2008 01:52

Hi again,
As a marginal note. There are historical cases of constructed languages that managed to become "natural". Italian is one such example. When Dante writes his De vulgaris Eloquentia, he clearly states that his language has... 2 speakers.
The proceeding to define "a language for Italy" took centuries, and when the nation was united this construct was used by les than 1% of the population. This also clearly shows if you read stuff written in it just 80 to 60 years ago. In time it became a natural language and it is classified as such, presently.
Yet, the languages we deal with in this case are still at their Divine Comedy age. So they are a product of craft, not a product of nature. As such, we cannot only classify them as a "communication technology" and state a number of "users", rather than speakers.
The question is whether we want to spend the 5$/year it takes to help a potential Mr. Alighieri or not. If this can be considered a WMF mission, than be it.
In that case, one will want to say what minimal number of users are requested (30 is not enough is not a criterion "at least X users" is). And it's also a matter of saying whether we can identify an independent source that can produce an acceptable estimate, or we are going to rely on in-house technology (our stats).
Jesse Plamondon-Willard (Pathoschild)
19 January 2008 02:07

Promoting languages is a noble goal, but not part of the Foundation's mission. I don't think it should be a factor in our decisions.
I would favour simply removing the exception, so that any constructed language can be approved once it starts having living native communities like Italian. That's fair and consistent with how we treat other languages. We also don't have to make arbitrary decisions; we can just ask for an update to the language's ISO 639 code, and SIL's linguists will do the research for us.
Gerard Meijssen (GerardM)
19 January 2008 03:38

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
19 January 2008 04:15

Sure, I'm in favor of a consistent policy with quantifiable, objective criteria. I've said this repeatedly.
You repeatedly state that constructed languages and historical languages are different. This is true in the same way as "historical" languages are different from "extinct" and "ancient" languages. They're all languages with no native speakers treated in the same way by the policy, regardless of how many non-native speakers use them in academia or in enthusiast groups. The only reason constructed languages are treated differently is because we did not make a decision about them at the time we changed the policy.
You again mention activity. The activity on the Lingua Franca Nova wiki does not put any approved request to shame; it seems to be mostly edited by a single user, who happens to be the creator. <http://lfn.esef.net/index.php?title=Special:Recentchanges&limit=500&days=30>. Again, the Ottoman Turkish test wiki was vastly more active, but it got rejected anyway. Activity is not relevant if the language is not eligible.
The current policy requires "a reasonable degree of recognition as determined by discussion", which it does not have— certainly not if you go by either subcommittee or community discussion. Furthermore, this very requirement is under discussion, as you may have noticed. You seem to be the only member who has commented that does not disagree with it.
If you need to ask what's left, you need to read my messages as well as those by the other subcommittee members. The introduction I sent to every subcommittee member (including yourself) might also be helpful.
Bèrto 'd Sèra
19 January 2008 04:15

This IS a criterion. Okay. Any alternatives?
Gerard Meijssen (GerardM)
19 January 2008 05:16

<this user has not agreed to public archival.>
Sabine Cretella
19 January 2008 06:02

One note: the difference among extinct/historical languages and constructed languages is that constructed languages have "new terminology" and this is wished. While it does not make sense at all to force an extinct or historical language to "incorporate" modern terminology.
That is what makes the status of these languages VERY different.
Maria Fanucchi (Arria Belli)
19 January 2008 06:44

I have followed this discussion since it began, though quietly as usual. I'm very discrete on all mailing lists, more so when heated discussions are going on involving people more experienced than myself.I'm not sure you're really interested in my opinion on this matter, but just for the record, I do not think that Lingua Franca Nova (or most constructed languages, for that matter) should be allowed a Wikipedia. Wikipedia is here to transmit the world's knowledge to the most people possible; when these people have a native language but prefer to work on a constructed language with a few dozen people, this does not justify, at least in my mind, a Wikipedia in that constructed language. I think it will serve mostly as a place to develop their language further, and perhaps promote it, rather than build an encyclopedia that others can use and learn from; if they want people to learn their constructed languages, they have their own preexisting tutorials for that.
I do not think Wikimedia is the appropriate place for them. Perhaps they could get a Wikia project.This is not in any way, shape or form an attack on constructed languages. I think these are fine exercises in linguistics that can produce some lovely works (read: I'm a Tolkien geek); I just don't feel that Wikimedia is the right place to develop them.
Bèrto 'd Sèra
19 January 2008 07:44

So...
At least positions are getting clearer. There is one side saying «no to constructed languages as such», and another side willing to allow them. I haven’t seen any specification of the conditions at which they would be allowed, though. «Discussion» is clearly not a condition. It looks like we can keep this thing going until our next lives, and «discussion» won’t lead us anywhere. The current result is «no result».
I’ve been thinking for a few hours now about «why would they want a wiki». The only answer I can find is that having a wiki has lately become a sort of status symbol. You seemingly become «certified as language» in the eyes of the masses because you are accepted into this club. Otherwise... why would they want a wiki? They already have one, as it seems, and there are lots of places that will host them for free (I can point them to one myself, if they have no better luck anywhere else). Hmmm.... I mean, I’m not sure I want the WMF to become something so snobbish that people would want to «be in», otherwise they are «out». Yet maybe the Board would find it nice for the marketing. I don’t really know, but this is yet another possibility.
I am positive that a wikipedia is made to deliver encyclopedic content to the people. Yet I’m also positive that nowadays it’s only us fools who believe in our own gospel. For everybody else out there, this is «the scene» and LangCom is but «face control», let’s face it. From a strictly functional POV, we are but a bunch of gorillas placed on the club’s entry to decide whether candidates are «trendy enough» for them to shake their legs on our dance floor.
Only, when it comes to «constructed languages» this pretty much results in a perturbation of the market. While natural languages have a full world for them to evolve even in the absence of a WMF wiki, most of this «wannabe languages» need a lot of marketing to make it. And if we close the door now, we will give an unfair advantage to those «constructed languages» that are already in. More hits, more exposition, more potential users. Exposition is surely not enough to win their campaign, but it DOES help a lot.
So you will excuse me if I require more time to think about it. Principles are a nice thing, but reality is made of cruel numbers. I won’t cast a vote until I can find a proper way to express a balanced solution.

Discussion 2

This discussion occurred alongside and after the related Foundation-l discussions.

Gerard Meijssen (GerardM)
25 January 2008 10:11

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
25 January 2008 10:39

Hello,
I replaced the redundant requirement for constructed languages with a note that constructed languages are under discussion, since it seems clear that no constructed language will be approved using that criteria ("sufficient recognition").
You are interpreting the policy in a way that does not match its intended meaning, and you are ignoring general agreement on this list to replace or remove the redundant requirement. You claim on Foundation-l that "As long as there is no alternative, the current policy stands"; this is arrogant, technically incorrect (see below), and practically incorrect (constructed language will not be approved with this requirement). The subcommittee discussion on constructed languages seems to have stalled for the moment, but that certainly doesn't mean the subcommittee generally supports the existing requirement.
Concerning the use of that phrase as an exception, quoted below is the message I sent to Foundation-l (twice!), which you've previously ignored:
------------
The exception for constructed languages that GerardM mentions is not an exception at all.
The line he's referring to is the second sentence in this requirement: "The proposal has a sufficient number of living native speakers to form a viable community and audience. If the proposal is for an artificial language such as Esperanto, it must have a reasonable degree of recognition as determined by discussion."
That phrase has been in the policy since the very beginning, before there was a requirement for native speakers. You can see this in the very first draft written on 11 November 2006, at <http://meta.wikimedia.org/wiki/Meta:Language_proposal_policy?oldid=466496>. (This draft predates my joining the subcommittee, so no subcommittee discussion shaped it.)
That original draft reads as such: "The proposal has a sufficient number of speakers to form a viable community and audience. If the proposal is for an artificial language such as Esperanto, it must have a reasonable degree of recognition as determined by discussion."
It was then intended not as an exception, but as an _additional requirement_. The requirement for native speakers was introduced nearly a year later on 17 October 2007 (see <http://meta.wikimedia.org/wiki/Meta:Language_proposal_policy?diff=711692>). The _extra requirement_ for constructed languages did not then exempt them from the new requirement; it was simply left behind by accident, and only noticed recently and misinterpreted.
As such, the current policy prohibits constructed languages *and* has a special requirement for them (which is contradictory, but that's because it's just an omission), it does *not* exempt them from needing native speakers.
This is the current matter of discussion: should we have an exception for constructed languages after all? If we exempt them from needing native languages, do we apply a special requirement for them or not?
------------
Gerard Meijssen (GerardM)
25 January 2008 11:10

<this user has not agreed to public archival.>
Jon Harald Søby
25 January 2008 12:01

It seems to me the only one not agreeing is you. The hint then is "compromise". You can't have all you want. Also you have a bad habit of just disregarding your opponents' arguments, and then repeat yourself over and over without addressing the issues that are being raised. It is impossible to reach an agreement then.

2008/1/25, Gerard Meijssen:
<this text is quoted from a user who has not agreed to public archival.>
Gerard Meijssen (GerardM)
25 January 2008 12:29

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
28 January 2008 17:23

Hello,
Again, the current policy does not make an exception for constructed languages. Repeatedly ignoring this does not make you right. Until there is consensus for your interpretation, the original interpretation remains in force. As you yourself said, "you do not change a policy until we agree on one".
I'm not concerned about pacifying naysayers; my concern is maintaining a balanced policy that is geared towards achieving the Foundation's mission. Constructed languages are different from historical languages in some ways, but not in any way that makes them any more suitable for achieving that mission. The practical facts as relates to a wiki are:
1. both have zero native readers;
2. both have non-native enthusiasts using them (with varying degrees of proficiency);
3. both must invent new words for the wiki, although constructed languages far more so than historical languages due to their small starting vocabularies.
Therefore, regardless of any irrelevant differences, their practical situation with regards to a wiki is generally identical. Any exception that applies for constructed languages should equally apply to historical languages. If we allow any constructed language that has a very active test project, we should revisit requests with very active test projects in historical languages.
(Further, languages should not be exempt from a requirement simply because they fail it; that's an illogical argument.)
Gerard Meijssen (GerardM)
28 January 2008 17:26

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
28 January 2008 17:37

I strongly disagree with yours; there is no such exception, as I explained in my first email to this thread. I supported this with a lengthy explanation and links to evidence; you have not supported yours at all. Votes without arguments are meaningless unless you want to put this issue to a subcommittee vote.
Gerard Meijssen (GerardM)
28 January 2008 18:48

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
28 January 2008 19:09

The requirement for native users was implemented in October 2007. What the policy said six months before that is not relevant.
If you want policy quotes, here is what the policy says now: "The proposal has a sufficient number of living native speakers to form a viable community and audience."
If you want outdated quotes taken out of context, here is something you said on November 24th 2007 in a thread called 'Orang Seletar': "Given that according to Ethnologue there are only some 1700 people that speak this language, I would not consider this a language that is likely to do well as a project." The constructed languages in question have approximately 30 isolated users.
Gerard Meijssen (GerardM)
29 January 2008 01:09

<this user has not agreed to public archival.>
Gerard Meijssen (GerardM)
29 January 2008 02:56

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
29 January 2008 07:51

I did not change it against your protests; indeed, you just blanket revert anything you disagree with. Here is what the policy said in October 2007, long before we even began to look at constructed languages: "The proposal has a sufficient number of living native speakers to form a viable community and audience" (identical to what it says now).
Your proposed policy, <Language subcommittee/Archives/Policy#GerardM>, was rejected. You can compare my original draft and the compromise we both agreed to at <http://meta.wikimedia.org/wiki/User:Pathoschild/Sandbox?diff=855514 >; both include an identical requirement for constructed languages.
Semantics is not relevant here, no matter how often you repeat it. There are 5 language types defined by ISO 639-3 <http://www.sil.org/iso639-3/types.asp >. "Living" is acceptable, others are not. It's not relevant that extinct, ancient, historical, and constructed languages are different types, because their situation with regards to readers and editors is identical.
Jesse Plamondon-Willard (Pathoschild)
29 January 2008 07:52

Fixed link to your policy draft: http://meta.wikimedia.org/wiki/Language_subcommittee/Archives/Policy#GerardM
Gerard Meijssen (GerardM)
29 January 2008 09:26

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
29 January 2008 09:59

There is no history for your policy draft. That is the exact text you emailed to us on 02 January 2007 at 17:20 EST in a thread titled "Proposal for a new language policy". I can forward your message to you if you don't have it archived, since it predates the langcom-l mailing list and you don't allow public archival.
I was the only significant editor of the compromise draft that was approved, as you can see at <http://meta.wikimedia.org/wiki/Special_projects_subcommittees/Languages/Policy/GerardM-Pathoschild?action=history>: "copied from langcom wiki, tweaked (other authors: minor changes by Cormaggio and Hillgentleman)".

External comment about historical languages

Gerard Meijssen (GerardM)
29 January 2008 02:18

<this user has not agreed to public archival.>
--------
Dear Language subcommittee,
I would like to register my disagreement with the recent rumblings in the Language subcommittee against any and all Wikipedias in "historical" languages, as well as the possibility of that being extended to any and all Wikipedias in constructed languages.
Now, all "historical" languages are not created equal. Some languages have no contemporary literature, like Anglo-Saxon. There can be no purpose to writing a Wikipedia in such a language, nor even a modern vocabulary for writing one.
Others languages have an active contemporary literature, like Latin, in which significant works are still written by the Vatican and others. Languages like Latin I would classify as living "classical languages" that have a contemporary literature, but few or no modern speakers.
Some people would say that languages without native speakers are useless. I disagree profoundly. When Newton wrote Principia, was he writing in a 'useless' language?
Principia is a book that changed the whole history of science. Yet, when this volume was published, Latin was already a primarily-written language whose last native speakers had died out a thousand years before. Would we have denied a Latin Wikipedia in Newton's time? Should we deny one today, when Latin, though it has declined, still has an active literature?
If a language has an active literature, it is not useless. Yes, primarily written languages are not ideally suited for teaching young children basic facts about the world. But they do have an important place in the intellectual sphere. Imagine Catholic seminary students, from different parts of the world, writing articles on church history, using the original Latin sources. Would not such articles be ripe for translation into many different languages?
And the argument that people are being siphoned off from their native language Wikipedia to work on Latin just doesn't make any sense; it is far more likely that the unique prospect of a Latin Wikipedia is drawing people in who would not otherwise be associated with Wikimedia projects at all.
Of course, the big question is, where do you draw the line? And how do you draw it effectively, so that we don't exhaust the resources of the Language subcommittee in fruitless research? As you might have guessed, I'm a strong proponent of requiring active contemporary literatures. ISO doesn't evaluate this, so we need alternate criteria.
One really simple and quantifiable method to measure that is this: Is a language's contemporary literature notable enough to be the subject of a Featured Article on English Wikipedia?
Or indeed, if that is too English-centric, I would have no problem (and would in fact prefer) expanding the criteria to cover FAs and GAs in any of the major Wikipedias, per David Gerrard's suggestion on the public foundation mailing list. What we're really looking for here is demonstrability of the notableness of contemporary literature.
Yup, simple as that. So, can Modern Latin literature make it?- probably, with some work. Modern Anglo-Saxon literature?- no way. Modern Ottoman Turkish literature?- maybe (I'm honestly not familiar enough to know). This way, -you- don't have to do the research. The Featured Articles process (which involves quite extensive peer review mechanisms) does it for you.
And these criteria for living "classical languages" (demonstrability of the notableness of contemporary literature) could be applied similarly to constructed languages, which are also (just like "classical languages"), really primarily-written languages .
Gerard has kindly forwarded this letter on my behalf, as I am not a member of this list. If you have any questions about this proposal at all, please send them to my personal e-mail and I will respond promptly.
Thanks,
User:Pharos
Jesse Plamondon-Willard (Pathoschild)
29 January 2008 07:55

Hello,
Could you provide his email address so we can include him in the discussion?
Gerard Meijssen (GerardM)
29 January 2008 09:23

<this user has not agreed to public archival.>
Bèrto 'd Sèra
29 January 2008 10:02

Latin had a very important role thru the centuries, and up to Newton’s age was still very much in use as a technical Lingua Franca. In THAT period, ISO would hardly have classified it as “historical”. It was rather an acrolect of the general Romance and Germanic language set. Nowadays, the only surviving role for Latin is basically ecclesiastic literature. Same thing can be asserted for Slavonic, used by us orthodox. In those days Latin had the value that today has been taken over by International English, and Newton used it as such. He wrote using the medium that ensured him the widest European audience. Today, I cannot think of any scientist who would write in Latin, for the simple reason that nobody would read him. The argument that Latin might attract to the WMF people who otherwise wouldn’t work in the WMF is valid for just any linguistic code, Klingon included.
As I said before for constructed languages, hosting one more or one less language makes little difference if any, in terms of WMF expenditure. So this is not about money. It’s about WMF mission. If the Board thinks that the mission must be changed in order to allow “training places for linguistic entities” instead of only “delivering encyclopedic content” it will say so; this is not up to LangCom to decide. Once the mission is enlarged to such an extent that training places are included there is no problem in approving just anything having a valid ISO code (Klingon included). At the moment the situation is different.
Bèrto 'd Sèra
29 January 2008 10:24

Oops… just saw this. I’ll repeat the message to him.
<this text is quoted from a user who has not agreed to public archival.>
Bèrto 'd Sèra
29 January 2008 10:40

Hi Pharos,
Sorry, I didn’t notice you weren’t included in the list of addresses, so I repeat the message for you to be included. Pls be aware that when I put Latin, Slavonic and Klingon in the same message I do not mean to be derogatory to anyone. I’m just pointing out that we have no “ad hoc” policy, and that whenever we make a step we must consider the implications in terms of actual implementation.
Also… when Catholic or Orthodox students want their material to be read by a general audience they use a living language (no matter which one). I understand and appreciate the concerns of those who believe that Slavonic and Latin carry a “mystic” value for religious matters. I am myself one of them, as I’d never be satisfied on hearing a Mass read in contemporary Russian or Ukrainian (or in Latin, for that matter). Yet this, once again, has nothing to do with encyclopedic content. When I wear my Christian hat I wear it, here I take it off and wear the encyclopedic hat. They are different in nature and cannot be mixed at any level. If one isn’t ready to do this he cannot be into a project requiring NPOV, because “believing” is an Article of Faith, and as such it is 100% POV by definition.
Bèrto 'd Sèra
30 January 2008 23:54

I'm sending Pharos' answer, since he cannot access the list himself
Bèrto ‘d Sèra
-----Original Message-----
From: Pharos <email addressed censored>
Sent: Thursday, January 31, 2008 5:09 AM
To: Bèrto 'd Sèra
Subject: Re: Historical languages
Hi Bèrto,
I would like to answer your letter, but I would also like you to please share my answer with the rest of the Language subcommittee, by forwarding it to the mailing list.
The text of my letter follows below.
Thank you very much for listening, and for your help in forwarding this.
--------
Dear Bèrto and Language subcommittee,
I think it is important to remember that ISO simply has no category under which a primarily-written language like Latin would fit. It just does not recognize the situation that a language could still be actively written in, but have no native speakers.
And I understand why this is so- ISO was drafted in the modern age, and Latin (as well as other "classical languages", like Sanakrit and Classical Chinese) have indeed declined greatly in relative importance. This reflects a long-term historical shift in many cultures toward "national" languages and literatures.
But there is a difference between declining greatly and going extinct. Ladino, the traditional language of Sephardi Jews, had dramatically declined in recent decades, going from millions to mere thousands of speakers; but we judge a language on its current use, not how far it has fallen.
And Latin, all will agree, has certainly fallen from a very great height. There are no scientific journals published in Latin anymore, it is true (as with the vast majority of languages, which unfortunately have never had scientific journals in the first place). But there is -still- some new material being published in Latin, as there has been continuously for many centuries.
So I would say that this whole category of "classical languages", not just Latin, has declined greatly, but that the category is not extinct.
Supporting a Latin Wikipedia does not have to do with any mystical respect for the language, or for ancient texts written in that language.
I support Wikipedias in Latin and other living "classical languages" simply because of the new literature (recently published articles and books), that have come out in these languages, certainly comparable to the literary output for many of our smaller languages that have Wikipedia editions.
New works are published in Latin, and though a proportion are of a religious subject matter, these are generally philosophical treatises (which is an idiom not dissimilar to encyclopedia-writing), not devotional poetry or something.
And we should judge every "classical language" on an individual basis, because some are truly extinct nowadays, and some are not. And I do think requiring an FA or GA on the contemporary literature of a language is a good criterion that can be used in -actual implementation-. Certainly "Klingon" would be quite excluded by such a criterion (because 'Klingon literature' is certainly not a notable subject).
And I am not arguing that we should adopt Wikipedias in living "classical languages" simply -because- they will draw new contributors to Wikimedia projects; I am simply pointing out the flaw in the thinking that Latin is so "enticing" it would somehow draw people away from Wikipedias in their native language, such as English or Italian.
I think Wikipedias in Latin-like languages are very much a part of our goal of "delivering encyclopedic content". Latin is a language that is read by many thousands, and in which there are still active writers, and it is firmly within our mission I think to enrich this living intellectual tradition. It also a language in which the original sources for many important historical and religious encyclopedic topics are written, and the Latin Wikipedia is a potential source of collaboration for content that can be profitably translated into many other languages.
And all of these arguments apply equally to all living "classical languages" (that can be demonstrated to still have active contemporary literatures), and are not specific to Latin.
Bèrto has kindly forwarded this letter on my behalf, as I am not a member of this list. If you have any questions about this proposal at all, please send them to my personal e-mail and I will respond promptly. Also, I would appreciate a CC of any replies.
Thanks,
User:Pharos
Bèrto 'd Sèra
31 January 2008 00:26

Hi!
Well... there are a LOT of possible improvements to just any standard, and ISO cannot be an exception. Yet, if you have proposed amendments to ISO 639-3, LangCom is not the place to discuss them.
One of the reasons why we are very strict on ISO application is that rarely people realize to which extent industrial standards rule their lives. Making ISO 639-3 an evident conditioning factor helps people in understanding how the industrial system will act towards their linguistic entities in the future. This, in turn, pushes real people to consider that they can contribute to the making of better standards by providing feedback to the Boards that publish them.
This is not only about languages, obviously. It's thousands of standards out there that rule one's life, and mostly people ignore even just their existence. So it's in the interest of democracy if people become aware of what standards rule what, and how they can ask for improvements.
Your point is that Latin is not "historical", but rather "living" (although living a particular life). This is something for SIL to evaluate, not for us. We do not assign labels, we simply use those assigned by ISO on an "as is" basis.
The implications of an "historical" label are that semantic production for that entity should be closed. The language has been frozen at a certain point in time and it is not supposed to produce new semantic concepts anymore.
Now, there are two technical possibilities:
1) a new label is released for "Contemporary Latin" as a "living" language, and as a distinct entity from "historical Latin"
2) Latin is "living" and not "historical"
None of the two depend from a LangCom decision, but obviously as soon as SIL decides that Latin is "living" and publishes such a decision the consequences in terms of our policy are automated.
Personally I don't think that having a Latin wikipedia can damage anyone. Knowledge of Latin can only help both English and Italian speakers in getting familiar with case declinations, which are of great importance if they want to study German or Slavic languages.
My personal opinion is just an opinion, though. A language doesn't need my opinion to be granted a wiki, it needs a "living" ISO 639-3 code. BTW, why so much attention for Latin? Nobody is asking for Latin wiki to be closed, AFAIK.

Limburgish Wikinews

The request for an Limburgish Wikinews was conditionally approved.

Gerard Meijssen (GerardM)
13 January 2008 14:47

<this user has not agreed to public archival.>

Wikipedia & Wikisource Coptic

The second request for an Coptic Wikipedia was rejected, and no decision was taken on the request for an Coptic Wikisource.

Maria Fanucchi (Arria Belli)
13 January 2008 14:52

http://meta.wikimedia.org/wiki/Requests_for_new_languages/Wikisource_Coptic
http://incubator.wikimedia.org/wiki/Wp/cop
This proposal has been stagnating for some time now. There's one user interested in the project, which has been inactive for quite some time on the incubator. What say you? Close or suspend our decision until they get more stuff done? (The link provided on the incubator page leads to a site in French about the Coptic Orthodox Church and not texts in Coptic.)
Gerard Meijssen (GerardM)
13 January 2008 16:59

<this user has not agreed to public archival.>
Bèrto 'd Sèra
13 January 2008 17:13

A wikisource would be fit for such a language, provided that they can find material to publish. I know pretty well what a coptic christianity is, but I have no immediate awareness as per the availability of original coptic text. Maybe we can ask these people to be more specific? Unless the material is really a lot and it draws a sufficent number of active users, I’d suggest they stay on the multilingual wikisource, anyway.
Gerard Meijssen (GerardM)
13 January 2008 17:21

<this user has not agreed to public archival.>
Gerard Meijssen (GerardM)
13 January 2008 18:24

<this user has not agreed to public archival.>
Bèrto 'd Sèra
13 January 2008 18:35

Definitely yes.
Maria Fanucchi (Arria Belli)
13 January 2008 18:43

Shall I close the request, then? (Just to be clear, I mean the Coptic Wikipedia proposal.)
--Maria
Bèrto 'd Sèra
13 January 2008 18:45

Given they have but one user, I’d say yes. Point him to the general wikisource, where he can publish all the text he wants
Maria Fanucchi (Arria Belli)
13 January 2008 18:48

I'm talking about closing the Wikipedia (WP) request, not the Wikisource (WS) one. I don't think he'd be allowed to put encyclopedia articles in a Wikisource. Or am I wrong?
Bèrto 'd Sèra
13 January 2008 18:54

You are right. Sorry, I was misguided by the title J
Anyway yes, it’s tagged as extinct, so all we can do with the pedia is close the request.
Maria Fanucchi (Arria Belli)
13 January 2008 19:14

It's been done. Both requests rejected and archived and a message sent to the initiating user explaining that as far as the Wikisource goes he can work on oldwikisource.
Hopefully, these being my first requests handled, I didn't mess up.
Jesse Plamondon-Willard (Pathoschild)
14 January 2008 13:40

Hello,
I think we need to clarify the Wikisource issue. According to the policy, we allow Wikisource subdomains in extinct languages (based on a compromise we reached when we disallowed extinct languages). According to past practice, we do not close requests for inactivity that otherwise meet the requirements for eligibility/conditional approval.
Which of these has changed? Are we closing a valid request due to inactivity, or changing the policy with regards to Wikisource subdomains in ancient languages?
Bèrto 'd Sèra
15 January 2008 05:25

You are right and nothing has changed. The request should remain open.
Maria Fanucchi (Arria Belli)
15 January 2008 05:58

Request reopened.

Wikipedia Jeju

The request for a Jeju Wikipedia was rejected.

Maria Fanucchi (Arria Belli)
13 January 2008 15:06

http://meta.wikimedia.org/wiki/Requests_for_new_languages/Wikipedia_Jeju
Artificial language, no ISO code, one interested user, two supports, proposal stagnating since April 2007. Close?
Gerard Meijssen (GerardM)
13 January 2008 17:02

<this user has not agreed to public archival.>
Bèrto 'd Sèra
13 January 2008 17:15

ok
Maria Fanucchi (Arria Belli)
13 January 2008 19:26

I may close this request tomorrow if no one opposes.
--Maria
(Yes, I decided to actually work a bit for a change. However, given how disorganized I am this may not last.)
Jesse Plamondon-Willard (Pathoschild)
18 February 2008 22:40

Hello,
I've marked this request as rejected.
—
22 February 2008 22:49

<this user has not agreed to public archival.>

Wiktionary Aramaic

No decision was taken on the request for an Aramaic Wiktionary.

Maria Fanucchi (Arria Belli)
13 January 2008 15:14

http://meta.wikimedia.org/wiki/Requests_for_new_languages/Wiktionary_Aramaic
They initially looked very motivated, but there has been no activity at all on the proposal page since April 2007 (excluding Jesse's template tweaks a month later). I think I'd approve them (ISO code, enough interested users perhaps inactive because they gave up on the request being treated, large amount of texts with which to work, historically important language). What do you think?
Gerard Meijssen (GerardM)
13 January 2008 17:03

<this user has not agreed to public archival.>
Bèrto 'd Sèra
13 January 2008 17:14

If this is a second project they also need full localization. It can be a good chance to see if they are willing to work.
Shanel Kalicharan (Shanel)
13 January 2008 18:03

The problem is that Aramaic (living Aramaic anyway) is not a single language. According to the article on en-wiki "Aramaic is really a group of related languages, rather than a single monolithic language," and "Modern Eastern Aramaic exists in a wide variety of dialects and languages. There is significant difference between the Aramaic spoken by Jews, Christians, and Mandaeans." From the proposal, it looks like they want to include all dialects and languages. *wonders what the Aramaic Wikipedia does*.
Bèrto 'd Sèra
13 January 2008 18:26

Then the answer is no. They should wait for MLMW, too. This is what MLMW is made for. I understand the need to share admin resources, and they have a code they can use to «legally» group, but we still need a tool for them to do it.
Jesse Plamondon-Willard (Pathoschild)
14 January 2008 13:35

Hello,
An additional problem is that Aramaic is classified as "ancient", which means that it went extinct centuries ago. This fails the requirements for native speakers, so I propose it be rejected.
Gerard Meijssen (GerardM)
14 January 2008 13:49

<this user has not agreed to public archival.>
Shanel Kalicharan (Shanel)
14 January 2008 14:48

Yes sillypoo, it depends on what Aramaic you are referring to. :)
Jesse Plamondon-Willard (Pathoschild)
14 January 2008 17:42

Sorry, I was confusing what they were proposing. They're proposing a wiki for all Aramaic languages, not just the language named 'Aramaic' (which is extinct).
In that case, as GerardM said, the problem is having no ISO code. Maybe we can propose a 'macrolanguage' or 'collection' code for Aramaic languages?
Bèrto 'd Sèra
14 January 2008 18:11

As said, MLMW is necessary here. Aramaic languages (whatever ISO will say about codes) are not necessarily mutually understandable to the point in which such a project can work "as is".
We need something that will minimize friction, otherwise the battle to define "what is the RIGHT aramaic" will start asap.

Wiktionary multilingual

No decision was taken on the request for a multilingual Wiktionary.

Maria Fanucchi (Arria Belli)
13 January 2008 15:20

http://meta.wikimedia.org/wiki/Requests_for_new_languages/Wiktionary_multilingual
A bit controversial, I'm afraid, though stagnating for many months. What shall we do?

(The incubator page has not been created, I think: http://incubator.wikimedia.org/wiki/Wikt/mul In any case, I see an edit page.)
Gerard Meijssen (GerardM)
13 January 2008 16:08

<this user has not agreed to public archival.>
Bèrto 'd Sèra
13 January 2008 17:17

No. Not until MLMW is out. This is going to produce a lot of work that later will either have to be remade or will become useless and/or out of standard.
Jesse Plamondon-Willard (Pathoschild)
14 January 2008 13:17

Hello,
I requested more input from the community a few days ago, but I only got one response < http://lists.wikimedia.org/pipermail/foundation-l/2008-January/037340.html >.
I suggest requiring a demonstrable consensus on the Wikibooks community, like a vote similar to the one that created the Wikisource subdomains < http://wikisource.org/wiki/New_vote_on_language_subdomains >. Since this is a question of organization with no nationalist or linguistic stakes, I don't think such a vote would be problematic. Further, since it is more a question of organization than of creating a new language, I think it is outside our scope to decide < http://meta.wikimedia.org/wiki/Langcom#Scope >.

Statistics and policies

Sabine Cretella
13 January 2008 17:37

Hi - first of all thanks Jesse for your work which in itself is great and will be helpful. Well you know that I don't like statistics since I am more about people, direct contacts, mutual understanding etc. and I took a long time thinking.
When we started with LangCom the policies were created to take out as much discussion as possible since if we start to discuss about each single new proposal and make any exception to our rules we will get into trouble. Actually there is little room for discussions when we base our decisions on clear points so only the "real problems" remain to discuss and most of the process can become automatic. This might be sometimes hard for new projects and also us, because often it could happen that we know a project would work well even without meeting this or that goal, but in the end it is a clearer way to get to discussions.
Considering that the languages we will have to deal with will become more and more particular and we will understand only very few of them we will not have much of a chance to have proof that things are proceding fine, but we then can only rely on the statistics. My problem now is that all should have the same rights and therefore, even if I don't like it really myself I propose to base our decisions on the statistics only when it comes to contents creation. The reason for this is that otherwise we would get into endless discussions why we do accept one project to be opened (maybe because we can consider it differently knowing the language) and another one is rejected (just because we don't know that language and have to rely on statistics). To remain objective we only have one chance, if we like it or not: follow the figures.
You can believe me that this mail was not easy to write since it is somewhat against my beliefs ... but I don't see a different chance if we want to do proper work and remain objective.
Thanks and have a great week!
Jesse Plamondon-Willard (Pathoschild)
14 January 2008 13:28

Hello,
Statistics are not the only objective tools we can use. For example, Bèrto sometimes asks contacts to verify that the language is actually what is advertised. They might tell us that it is another language, or poorly written by non-native editors; both are objectively bad.
But I do agree that we should only use objective criteria when deciding projects.
Gerard Meijssen (GerardM)
14 January 2008 13:40

<this user has not agreed to public archival.>

Gilaki Wikibooks, Wikinews, Wikiquote, Wikiversity

The requests for a Gilaki Wikibooks, Wikinews, Wikiquote, and Wikiversity were conditionally approved.

Gerard Meijssen (GerardM)
14 January 2008 09:52

<this user has not agreed to public archival.>

Rename conditional approval

The policy and templates were modified to use updated wording, in particular renaming "conditional approval" to "verification as eligible".

Jesse Plamondon-Willard (Pathoschild)
15 January 2008 18:15

Hello,
I suggest renaming 'conditional approval' to 'verified as eligible'. Conditional approval has been a source of confusion for a long time, particularly given that our criteria can change. Last week I changed the templates as an experiment to use the new wording, with a clearer body text. This seems to have significantly decreased confusion, despite meaning the same thing. It also clarifies the distinction between criteria for eligibility and for approval in the policy.
You can see a comparison of the previous and new versions at <http://meta.wikimedia.org/wiki/User:Pathoschild/Sandbox?oldid=836287>.
If there are no objections, I'd like to make this change permanent and update the wording in the policy in 48 hours.
Bèrto 'd Sèra
15 January 2008 18:21

Yes, it makes sense to me.
Gerard Meijssen (GerardM)
15 January 2008 19:42

<this user has not agreed to public archival.>

Lingua Franca Nova

GerardM noted that Lingua Franca Nova's test project is external.

Gerard Meijssen (GerardM)
17 January 2008 16:13

<this user has not agreed to public archival.>

Wikipedia Sassarese

No decision was taken on the request for a Sassarese Wikipedia.

Sabine Cretella
19 January 2008 09:00

Hi, could we please get the stats for the Sassarese incubator project? I was contacted by one of the group of editors and he wanted to know what was missing. I already had a look and the UI is ok. I am quite sure that this project will succeed well since they have a mixed group of peoople who work online and offline and I got the names of the anonymous editors.
I already talked with some of you about this, to understand if generally it makes sense to send this request right now. Of course we now need to have a look at it together.
Shanel Kalicharan (Shanel)
19 January 2008 14:16

Hello,
The statistics are at http://tools.wikimedia.de/%7Epathoschild/ls-testanalysis/?prefix=wp/sdc :)
Jesse Plamondon-Willard (Pathoschild)
19 January 2008 16:13

Hello,
They meet all the requirements except for an active test project. If you look at the edit distribution per month, you'll see that participation dropped after September 2007, so that only one user has more than 3 edits per month. Once they can spur more activity for a while, they'll be ready for approval. :)
Sabine Cretella
19 January 2008 21:22

Hi,
first of all thanks to Shanel for the link, but it does not load here for some reason (sometimes this happens).
I passed this info on to Federico who asked me. Their problem is: like for Neapolitan many people are not literate enough for wikis and he gets things in a file. He told me that people were somewhat shocked when they heard that to write bold they had to use and .... well I know that - I have exactly that problem with people who could write on nap.wiki and don't. It is one of the reasons why I am thinking about throwing actually all templates out and why I am starting to avoid wiki links right now since it will show people that writing plain text without any formatting is ok - all the rest can be added later by anyone who is able to read. So their are exactly in five people - he is the one mainly doing things and keeping them together.
I know what it means to keep a small wiki going also - and I was informed by one of the other main editors that he will not be able to work on nap.wiki for months. I myself are still under time pressure for several reasons and I wanted nap.wiki to be able to go alone, but again it comes back to me so that I cannot really leave it and just care about the technical stuff - I will have to do some bits and pieces there, hoping that other people will come up to help as well. This is one of the problems all small wikipedias will have. I know that Federico is serious since he is working on the wiki even if he is under time pressure having to give exams and with other trouble around.
So I hope that Sassarese will make it, really.

Technical reforms

Several reforms aimed at increasing the subcommittee's effectiveness were described.

Jesse Plamondon-Willard (Pathoschild)
20 January 2008 06:01

Hello,
I've been busy over the last few weeks on some sweeping technical reforms. If you're interested, following is a description of the most important changes.

==Analysis==
I've rewritten the analysis script from the ground up. The new script accesses the replicated database directly to obtain every bit of information stored by MediaWiki. Analysis is now performed in a single step (previously we had to generate a list of pages, paste it to the relevant wiki, then process the relatedchanges), is vastly more efficient, and is now open to public usage (previously only I could run it, for performance reasons).
With the replicated database, the analysis script provides information on the test project's entire history since the very first edit (previously only the last 30 days were available for technical reasons). In addition, the script newly provides:
- lists of editors and redirects;
- edit distribution by user per month;
- number of edits and minor edits per month;
- number of new pages per month;
- amount of content added or removed in bytes per month;
- number of editors per month;
- and overall statistics, including total number of non-redirect pages, redirects, editors, and edits.
If you have any suggestions for statistics, feel free to respond to this email.

==Status pages==
Status pages have been almost entirely automated and moved to Meta. They've been completely redesigned to provide users with every documentation page and tool needed to track their own progress, and integrated with the automated analaysis tool above. These changes were also intended to minimize the work required for investigation, so that a thorough investigation of a project can now be done in a matter of seconds instead of taking ten minutes or more.
You can see the difference yourself by comparing these two links:
- old format (manual): <http://incubator.wikimedia.org/wiki/User:Pathoschild/Status/wp-bcl?oldid=131393>
- new format (semi-automatic): <http://meta.wikimedia.org/wiki/Language_subcommittee/Status/wp/bcl >
==Documentation for subcommittee members==
The documentation at <http://meta.wikimedia.org/wiki/Language_subcommittee/Handbook_%28subcommittee%29> has been completely rewritten, expanded, and updated. It now covers every primary subcommittee task in detail, from creating status pages to approving a wiki.

==Verification for eligibility==
As I previously proposed, "conditional approval" has been renamed to "verified as eligible", to reflect the current practice of giving it automatically to requests that meet the requirements for eligibility. Conditional approval has been a source of confusion for a long time, particularly given that our criteria can change over time.
There is no practical difference, since they mean exactly the same thing, but this is much less confusing for requesters. It also clarifies the distinction between criteria for eligibility and for approval in the policy.

==Pages renamed==
The language subcommittee pages have been renamed from "Special projects committee/Languages" to "Language subcommittee", in order to make subpage naming less awkward. This was already a problem with the number of subpages we had, but made status pages very messy.
< http://meta.wikimedia.org/wiki/Special:Prefixindex/Language_subcommittee >

==Archival script==
The archival script at < http://pathos.ca/tools/ls-archival > has been greatly improved, which makes archival easier. I'll still be the only one archiving, but I thought I'd mention it anyway. :)
Jesse Plamondon-Willard (Pathoschild)
20 January 2008 07:48

Hello,
I also sent a modified version of the summary to Foundation-l, including the recent change in the localization requirements. Some users on that list are periodically critical of the subcommittee, so a little extra transparency and showing them some of the changes we're implementing to improve processing times might help. :)
Gerard Meijssen (GerardM)
20 January 2008 07:57

<this user has not agreed to public archival.>

Wikipedia Sranan Tongo

No decision was taken on the second request for a Sranan Tongo Wikipedia.

Gerard Meijssen (GerardM)
20 January 2008 11:34

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
20 January 2008 14:41

Hello,
The activity for January so far looks excellent, but I'd prefer to see two consecutive months of activity at the very least. Activity often spikes for a few weeks and then returns to its normal level, so one month is not a good indication of long-term viability. (Generally, projects we approve already have a few months of regular activity.)
Gerard Meijssen (GerardM)
20 January 2008 14:45

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
20 January 2008 14:54

Hmm.. Now that you mention it, they've added 26297 bytes in 145 articles, which comes out to an average of 185 bytes. That's exactly the size of my email so far, including this comment. The quality is definitely something we'll need to look at.
I suggest waiting a month or two to see what the normal activity level turns into, and what the quality of the articles will be. The test project has essentially started off this month, so we shouldn't be too quick to judge it.
Jesse Plamondon-Willard (Pathoschild)
20 February 2008 13:47

Hello,
I propose the final approval of the Sranan Tongo Wikipedia. If there's no objections within 48 hours, I'll include it in the batch recommendation to the board this week.
< http://meta.wikimedia.org/wiki/Language_subcommittee/Status/wp/srn >
Jon Harald Søby
20 February 2008 14:03

I concur.
--
Jon Harald Søby
Gerard Meijssen (GerardM)
21 February 2008 13:49

<this user has not agreed to public archival.>

Marking localisation requirements as completed on status pages

GerardM suggests not marking localization requirements as completed on status pages until the requests are approved. This suggestion is not enacted.

Gerard Meijssen (GerardM)
25 January 2008 13:08

<this user has not agreed to public archival.>
Jesse Plamondon-Willard (Pathoschild)
25 January 2008 13:22

Hello,
If you're referring to the status pages, not marking requirements as met when they are met would defeat the purpose of status pages. They serve to let the community know what they have left to do before they are approved. If we need updates on a few messages before final approval, we can simply mark the requirement as not met or ask them to do that before granting final approval.

Wikipedia High Norwegian, Värmlandic

The requests for Wikipedias in High Norwegian and Värmlandic were rejected.

Jesse Plamondon-Willard (Pathoschild)
30 January 2008 12:28

Hello,
GerardM has rejected the requests for Wikipedia in High Norwegian and Värmlandic 2 due to lack of ISO 639 code. (This is just notification.)
< http://meta.wikimedia.org/wiki/Requests_for_new_languages/Wikipedia_High_Norwegian>
< http://meta.wikimedia.org/wiki/Requests_for_new_languages/Wikipedia_V%C3%A4rmlandic_2>