Wikipedia:WikiProject Council/Proposals/TRANSWIKI

The following discussion is an archived proposal of the WikiProject below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the project's talk page (if created) or the WikiProject Council). No further edits should be made to this page.

The resulting WikiProject was created at Wikipedia:WikiProject Transwiki

---

Wikipedia:WikiProject TRANSWIKI
Wikipedia:TRANSWIKI

Description

In light of the recent events showing concern about the creation of unreferenced inefficient manual stub creation I think the time has come to organise a project Wikipedia:WikiProject TRANSWIKI which concentrates on trasferring content from other wikipedias but in a way which is much more efficient and can be done with no community concerns. This project would tie together the Missing Encyclopedia Articles project with the stale current translation project. Given that there thousands upon thousands of referenced good content articles on other wikipedias, particularly German, Dutch and French wikipedia I think it is very important to transfer content from other wikipedias but done adequately as part of a project coordination. The idea is a bot, offered by User:ThaddeusB, which can run through categories on a different wikipedia and extract any main information from an article and create it on english wikipedia with a reference to an outside reliable source. I know the community expresses an extreme indifference to automation in regards to content but if programmed correctly bots can do things much more consistently and efficiently than us. The idea is not that the bot writes the articles, the idea is that it draws up missing lists of articles from other wikipedia in the project space, members of a group check them for notability and then the bot is assigned to blue link them in the best possible way without community concerns and which adheres to our policies. The ultimate ideal of cause would be bot which can translate whole articles into English but as we know present, google translate is far from perfect. If it is somehow perfected in the future see google toolbar the new translation thing they have going then I think it would be possible to instantly translate articles but would need to be proof read. But at present I think something which can extracts some basic facts and reference them is most needed. The first phase would be to use a bot to draw up lists of missing articles by wikipedia in the project space. . Something like Wikipedia:WikiProject TRANSWIKI/de/Politicians etc etc. but it would seme this project would have the potential and capacity to do more than just list missing encyclopedia articles, it would attempt to fill a void for translation in which the other project failed in and help improve existing content. I've lost count for instanc ehow many articles I've come across on German and Swiss municipalities and the article on German wikipedia is full and well referecned. We'd need a bot to be able to run off categories on other wikipedias and list them on here in the workspace so in effect it would begin the start of making articles on other wikipedias avialable in english and at least highlighting exactly what is missing. So it could generate lists from a diversity of topics and wikipedias such as Wikipedia:WikiProject TRANSWIKI/nl/Writers, and Wikipedia:WikiProject TRANSWIKI/pt/Portuguese novels etc. They would be stored in the workspace into subcategories and topics for each wikipedia.

The first task of the new project would be to create sections of the new project related to content on the different wikipedia. Then the bot would raid the categories on other wikipedia for various topics and list the articles missing from these categories. The ones that we already have maybe can be moved manually or the bot could by pass them. So eventually we'd have a directory of missing articles organised by each wikipedia and neatly by topic/sub topic so we know exactly what is missing. Inevitably the task is a tremendous one to do so which only a bot could achieve but I am certain that a bot is able to be programmed to copy categories from the other wikipedias and insert them into lists in the project space. Once we have that done or are happy with the missinglists for one topic maybe then the bot can be programmed to start the missing articles, a lot of related categories use similar sources etc so that should make it easier at least.

Given the scale of this proposal I think a seperate wikirpoject dedicated to the transfer of content from other wikipedias and finding missing sources to reference them etc is EXACTLY what is needed at present. I've tried my very hardest to start as many articles as I can which are missing but I have received bad press from doing so in that quality is jeopardized as at present it is an effort done badly manually or by AWB which is really not the way it should be done. If we had a project and bot able to draw up lists of missing articles (rememebr the missing articles are enormous from other wikis) which this project would benefit immensely from the trasferring process could not only be done much more efficiently but it would also provide a monitoring process over what is being trasferred and makes sure say for instance that articles adhere to BLP concerns etc and the project would also provide a forum to discuss these issues and provide a means in which the community can collaborate and work together in doing so in the best way possible without one or two editors being shouted at for not complying with standards. What we need is an organization of language translators, bot operators, quality overlookers to esnure that articles can be transferred correctly and without referencing or other issues. Dr. Blofeld ^{White cat} 12:42, 27 July 2009 (UTC)

Support

Please specify whether or not you would join the project.

1. I think this is exactly what this project needs if it is to take content and missing information seriously Dr. Blofeld ^{White cat} 13:11, 27 July 2009 (UTC)
2. I'll support this if we're sticking to the proposed plan. There are lots of things that could go wrong here and the manual review proces is crucial if this is to gather any widespread support. Also, far from all foreign language wikis are up to standard regarding use of reliable sources. The Danish Wikipedia, for example, is notoriously riddled with unreliable references. So, support the idea but let's move forward with care. Vyvyan Basterd (talk) 13:45, 27 July 2009 (UTC)

   Before any articles are created, wikiprojects for the relative country will be notified in advance and invited to discuss any potential issues such as this prior to creation. Some wikipedias are barely usable, Indonesian wiki for instance is abarely usable and several of the small wikis are riddled with POV and other issues. So there would be a monitoring and approval process in place and drawing up notable content. Initially though this project is more geared towards the high quality more developed wikis, particularly German and Dutch and the referenced good content that exists on French wikipedia etc. Ideally I'd like the project to gain enough momentum to have a way to transfer content to existing articles more efficiently (Swiss and German municipalities for instance) and have people building up consistency in exisiting articles as well as a side of the project dedicated to generating missing content. What is required is a large wikiproject and sheer numbers which as time goes on I hope the project will attract. One of the main problems I find is a lack of real collaboration it many parts to the project even when many people share the same goals. I don't expect this project to create miracles but I do think it would help in planning and regulating our growth across the wikipedias. The key to the success of transferring articles effectively in the long run is manual translators who are able to develop the starter articles and who often work outside central projects independently. On any project I expect there to be at least people from the relative country/subject wikiprojects willing to nurture articles generated but as usual finding numbers will be a problem given the huge variety of interests on wikipedia. Dr. Blofeld ^{White cat} 13:51, 27 July 2009 (UTC)

3. Strong support. I have been working behind-the-scenes to do an overhaul of the translation system on Wikipedia, and I have seen just how much content there is to be gleaned from foreign-language Wikipedias. Compare, for example, Category:Chilean painters (15 entries) with es:Categoría:Pintores de Chile (67 entries). The first article I picked at random (es:Jorge Délano Frederick) has eight different sources and every sentence is cited. Obviously not all the articles are like this, but it's defintely not an anomaly. The 4x multiplier is typical for a category involving a non-English-speaking topic. Many or most of the translations I have observed where an {{Expand Spanish}} or other expand tag was on the article were by IPs or redlinked editors, so starting stub building blocks is crucial to encouraging these translations. As the lingua franca among wikipedias, it is especially crucial to get foreign-language information here so it can be diffused to other wikipedias as well. I feel strongly that the risk of introducing some errors is greatly outweighed by the benefit of access to information that is difficult to find in English-language sources and the rapid expansion of coverage that is possible only through translation. Any bot processes will obviously have to be approved separately before running, so any bot issues can be worked out as they arise. Calliopejen1 (talk) 14:40, 27 July 2009 (UTC)
4. Support. Given current outside demands on time (lead coordinator of WikiProjec tChristianity), I may not be able to offer as much input as I would like, but would be more than willing to do what I can as time and circumstances permit. John Carter (talk) 14:45, 27 July 2009 (UTC)
5. Support. I dont have a huge amount of time, but im suppose to be practicing french anyway so id be happy to lend help from time to time on translations into english and lending suggestions if needed. Ottawa4ever (talk) 15:05, 27 July 2009 (UTC)
6. Sure, might work. I'll help out when I can. –Juliancolton | ^Talk 15:06, 27 July 2009 (UTC)
7. Support. I'm going to try to help with German politicians. Pzrmd (talk) 15:09, 27 July 2009 (UTC)
8. Definitely. I would like to see credit extended to the original authors on the other-language wikipedias. This could be something as simple as a line saying "This article was created by translating information from the article in the [Whatever-language]-Wikipedia". I think it would be even better if we could import the article history from the other Wikipedia. In general, I think this is a fantastic idea. There will be kinks to work out though, and I think we should start slowly at first (maybe a couple categories from a couple different WPs), and see what problems we encounter. Once we work out how best to deal with them, we can go at a faster rate. LadyofShalott 15:22, 27 July 2009 (UTC)

   {{iw-ref}} should be used for this. –xeno^talk 15:42, 27 July 2009 (UTC)

   Thanks - I thought we had something for this, but did not remember what the template was. What about importing the history though? Surely that should be possible? LadyofShalott 15:48, 27 July 2009 (UTC)

   I inquired on this a while back, see Wikipedia:Village pump (technical)/Archive 59#Special:Import and Wikipedia:Village pump (technical)/Archive 43#Importing article histories –xeno^talk 15:49, 27 July 2009 (UTC)

9. I don't have time to review this in its entirety, but I support it in spirit. I think the best thing to do would be the bot create articles, noindexed, in project space and then it must be manually reviewed, improved (i.e. beyond sub-stub), and moved into the mainspace by editors. –xeno^talk 15:49, 27 July 2009 (UTC)
10. Support. This is something I'd idly wondered about before. Copyediting and finding better referencing would be my forte. Fences&Windows 17:32, 27 July 2009 (UTC)
11. Strong support. As I've been doing some of the transwiki work myself, I see the need and benefits for a coordinated transwiki project. --Rosiestep (talk) 17:38, 27 July 2009 (UTC)
12. Strong Support - anything that builds the 'pedia in positive ways is a good thing. — Ched :  ?  17:46, 27 July 2009 (UTC)
13. Yes, we certainly ought to be making sure that the information available on other languge Wikipedias is available to us too, and this sounds an excellent way of going about it (ideally, of course, there would be one properly integrated Wikipedia containing different language versions of articles, but reuniting the now diverse communities to achieve that would be no easy task). Anyway, I have used a bot for this sort of thing (importing info on Polish and now Czech villages), so will be glad to help as time permits.--Kotniski (talk) 08:13, 28 July 2009 (UTC)
14. Good idea. A combination of bots and editor collaboration could give us access to a lot of good articles very efficiently. --Apoc2400 (talk) 11:26, 28 July 2009 (UTC)
15. Supporting, of course. :-) Conditionally, tho' - there are a couple of points I'll bring up below. --User:Ser Amantio di Nicolao ^{Che dicono a Signa?}_{Lo dicono a Signa.} 15:24, 28 July 2009 (UTC)
16. Certainly I will help do what I can. I can offer the technical skills as far as bot programming and such, but I will need the language skills of translators to make this happen. I will elaborate below. --ThaddeusB (talk) 00:24, 29 July 2009 (UTC)

Discussion

This seems like a fine idea, of course, anything which gets reliable, sourced, BLPs is good. But are you planning to use Google Translate? If so, you will need to be approved by Goolge, per their Terms of Sevice:

5.3 You agree not to access (or attempt to access) any of the Services by any means other than through the interface that is provided by Google, unless you have been specifically allowed to do so in a separate agreement with Google.

If you are using a different translation service, which allows you to use it, then that is fine. - Kingpin¹³ (talk) 13:18, 27 July 2009 (UTC)

I don't think articles would be generated fully using google translate if that is your concern. It is not an effective enough yet tool yet to translate whole articles anywayI think the first task which may take many weeks is drawing up a missing article directory from other wikipedias in the most full way possible. Then we would have to sort out how we go about starting articles from other wikis later and find how extracting the main points and referencing them would be done. Remember, the articles will not be fully tranlsated immediately upon creation but they will be started in a way which are mostly important adequate and consistent starter articles which provide a decent platform to build on afterwards and still be an adequate starter article in its own right. I would suggest that the bot also draws up lists of top visited missing articles on other wikis and referenced articles from other wikis and lists them as first priority for starting. If we did require google translate it would indeed have to be arranged and google sent a request. Anything starter using google translate though would require a monitoring process and testing prior to creation to ensure that it is done properly as we know the tool is as yet imperfect. Dr. Blofeld ^{White cat} 13:26, 27 July 2009 (UTC)

Yeah, I'm aware that the articles aren't going to be direct translations from Google Translate, but if you use Google Translate at all, it will have to be done using their interface. But yeah, overall this seems like an okay idea. - Kingpin¹³ (talk) 13:36, 27 July 2009 (UTC)

Agreed. What I mean in regards to content would be something like Kobylá nad Vidnavkou which Kot is working on at present. It takes similarly programmed articles and manually it would be found which word means which in the foreign language first. Once that prior programming is sorted it would then be free to generate more articles based upon a set template. The main problem would be that many articles are not consistent or don't contain set information so extracting the main facts may be considerably more difficult and that is when a translation programme would be more required I believe. Dr. Blofeld ^{White cat} 13:39, 27 July 2009 (UTC)

A few possible problem I foresee. Wikipedia is generally not counted as a reliable source for other wikipedia articles. One of the concerns, unfortunately, is that sometimes the references added don't in fact support the content citing them. I have no idea how prevalent a problem that is, although I have myself been invovled in at least a few instances where that has been the case in the English wikipedia. Using just the foreign language wiki articles as sources, and particularly letting the foreign language wikis know that such is being done, might well spur the creation or addition of such content. Another one is whether the other articles to be translated have sufficient content to really merit an article here. If it were possible to notify some of the other, more active, WikiProjects in the English language about the proposed creation of these articles before creating them explicitly, to ensure that they are of sufficient content, don't have any real obvious POV or other problems, etc., that would probably help alleviate a lot of concerns. Maybe it would be possible to create them in some non-article space first, notify some of the extant projects, and then have members of those projects decide on whether to move the article as is or merge the content into another article? If these concerns were addressed, I don't see any particular problems with the idea. I would also ask that the notification not be extended simply to the national projects. In several cases, and I'm thinking specifically of [{WP:MILHIST]] and WP:FILM here, other projects may be more active regarding a given subject and probably be able to offer better input in some cases than the national projects. John Carter (talk) 14:17, 27 July 2009 (UTC)

Wikipedia is generally not counted as a reliable source for other wikipedia articles.. I know dearest John!! That's not what I proposed. I 'm talking specifically about referenced content to OUTSIDE RELIABLE SOURCES to avoid exactly the problem that was created by referencing to other wikipedias. Content can be extracted from other wikis and relative outside reference added to solve this problem. Referencing to another wiki is not valid as was potined out and which is partly the reason why I'm here proposing this, so that we exract the decent content from other wikipedias and have rleiable sources to back them up. Dr. Blofeld ^{White cat} 14:46, 27 July 2009 (UTC)

I wasn't meaning to propose that we were. The problem would be in determining whether the content referenced in those articles is in fact really supported by the indicated reliable source, and, possibly, whether those sources are such as would meet WP:RS here. They may be only theoretical problems, but I have no doubt they would appear in at least limited numbers, probably sooner rather than later. I still like the idea, but would want to see how to address those concerns. The comments here regarding the Danish wikipedia in particular and others presumably as well might also be problematic. Particularly in countries/areas where there are major international or interregional differences, what might be considered a reliable source in, for instance, Karnataka might not be considered a reliable source in Tamil Nadu, and the other way around. John Carter (talk) 14:55, 27 July 2009 (UTC)

Well yes of course. If we were to find information about places in Greece and southern Cyrpus we wouldn't exactly go running straight to Turkish wikipedia. Dr. Blofeld ^{White cat} 15:46, 27 July 2009 (UTC)

I think in principal the idea is good. But i think other wikis may find this useful for themselves. Ive seen many times on french wikipedia where an english article appears untranslated. It could benefit inter wiki coordinations and bring english wikipedia some high quality editors which normally would only work on one wiki, and vice versa. That said. I dont know how a bot or any automation process can be used easily for this or should (i do not know much about bots). I like the idea of generating lists and then manually translating the articles and filling in gaps. But we do have an articles for creation section already and maybe coordination is due there (would we be stepping on toes here?). Anyway i think the project has potential. It could grow,translation of new and even existing articles is useful. Theres also concern of using a number of non english sources in the articles which can be hard to verify for a non english user. but care is going to be needed as vyvan says above. but i think it can work and grow so long as the foundation is strong Ottawa4ever (talk) 14:23, 27 July 2009 (UTC)

As long as there's a manual review proces to catch these things then I think we'll be fine. This has been tried before on a smaller scale, where people could request that an article be translated from another wiki, but without coordination such as what is proposed here these efforts either die due to lack of interest or people just use Google Translate and do whatever they want. This is an opportunity to coordinate these efforts and ensure that quality and efficiency go hand in hand. Vyvyan Basterd (talk) 14:41, 27 July 2009 (UTC)

This is a fine example of what is happening now because we don't have any coordinated effort to import these articles. People just use Google Translate and do as they please. I can find more examples if needed. Vyvyan Basterd (talk) 14:46, 27 July 2009 (UTC)

That is exactly how I see it. Agreed with Ottawa too that such a project would be perhaps evne more useful for the other wikis, particularly the lesser devleoped ones. However i believe there is already something which helps them with outgoing content. People do forget that similarly we could greatly benefit from content on other wikis. The ultimate ideal would be a translation bot which produces articles on different wikipedias in any language fully and accurately first time but that is obviously a dream, nothing more, at leats not yet anyway. I don't really think bots should write articles, I believe it should be done by the hard work and research of people but I do think that bots will be of a massive benefit to starting content and remove the inadequacies of us clumsy manual editors in trying to start them by starting them more efficiently, faster, and more resourcefully. There should equally be bots running on here which do more than just fix minor errors. Content should be of uttermost importance to us and I think a project which seriously realises the differences between articles on different wikipedias and does something to work towards equalising it is greayly needed and I mean that not only on English wikipedia but on all the other wikis too. In the end it will always come down to hard work and involvement of many people to develop articles fully on wiki but a bot and project of editors who work on putting articles on the plate consistently and in coordination is definately the start we need. Dr. Blofeld ^{White cat} 14:47, 27 July 2009 (UTC)

This is a very strong point. Bots are very good at doing repetitive tasks quickly. However, there is a common belief that bot can't make judgments. This is false. A computer program can make a judgment based on a serious of rules. The accuracy of its judgment will be related to the accuracy of its rules, but it will at least be consistent. It is difficult, but not impossible for computer programs to make complex judgments. --ThaddeusB (talk) 00:36, 29 July 2009 (UTC)

• Using other language Wikipedia as a way to bring high quality content to Wikipedia English is a good idea. I think that every article that is a feature or good article on another Wikipedia should be brought to Wikipedia-English almost without exception. (And I can't image the reason for any exceptions.)

But I'm not convinced that bringing the full list of articles from categories topics from another language Wikipedia is a good idea. Some topics like geological formations (lakes, rivers, mountains) it is okay to have short stubs. Other topics, having a short stub is much more problematic because it might introduce biases into the content. For example, some might have an article started on them because they are a member of a parliament, and someone is writing an article on each member of parliament. But the person might actually be much more well known for other accomplishments that would be unmentioned while they had a dismal record in parliament. If no one is really has an interest in the topic, but is bringing the article to WP-en as part of a process, then a follow up article about the person may never be written.

So I think that the process needs to deliberately figure out ways to prevent introducing a large number of stale or incomplete articles that could have the potential to introduce biases about the subject of the article. --FloNight♥♥♥ 15:23, 27 July 2009 (UTC)

This is a valid point. Peter Garrett is one article which might be similar, if the article dealt only with his political or musical career. Someone like Ilona Staller might have even more problematic articles. We would definitely want to ensure that any articles directly relating to BLP concerns would be gone over very very thoroughly. John Carter (talk) 15:31, 27 July 2009 (UTC)

If you really think the project is dedicated to creating unreferenced sub stubs about Italian porn stars.... Dr. Blofeld ^{White cat} 15:40, 27 July 2009 (UTC)

IMO an incomplete or stale article is better than no article. (though perhaps this doesn't apply for BLPs.) We don't go deleting articles that are incomplete or stale, and in the same way there is no reason that they shouldn't be started to begin with. Some information is better than no information. Calliopejen1 (talk) 15:33, 27 July 2009 (UTC)

BLPs should not be translated without a thorough review by an experienced Wikipedia who understands policy. Ever. However, I prefer a good referenced stub class article on a 1911 movie rather than text for the sake of having more text. Vyvyan Basterd (talk) 15:37, 27 July 2009 (UTC)

I made the point in my original comment that for some categories to have stub is is helpful. For example, having a stub for every river, lake, mountain in a country is good information. But other topics such as people and organizations need to have review to make sure that the topic is presented in an unbiased manner. FloNight♥♥♥ 15:49, 27 July 2009 (UTC)

If Jenna and the others concerned about BLP are interested, the project could have a department which reviews any articles to be started on living people. Sources would be put forward and together an agremeent would be made as to whether or not that source or bit of information is adequate to start the article. As I said this should be a large scale project and adhering to BLP and having a group of editors who authorise banks of articles to be needed and discuss it with the rest of the group is what I am proposing. Dr. Blofeld ^{White cat} 15:44, 27 July 2009 (UTC)

@Vyvyan - BLPs are started by inexperienced, clueless users all the time. Obviously bot creation is a special concern and at this point I don't feel one way or the other strongly about it. This would have to be discussed before a bot were approved in any event, so there's really no need to worry about this special case now. And "text for the sake of more text"? Text=information. Yes information for the sake of more information is obviously good, that's what this whole project is about. Calliopejen1 (talk) 15:46, 27 July 2009 (UTC)

BLPs might be started by noobs regularly, but that doesn't mean that those started articles may not violate BLP, copyright, or something else. And somehow I think porn stars who have gone on to other things might be among the most problematic cases. I do think that having some sort of BLP section would be useful. One other concern, although this one might be a small one. Would there be any copyright concerns if, for instance, someone in Ugarit translated an English copyrighted source into Ugarit which we then translated back into the language where that content is copyrighted? I really don't know much about copyrights, so I don't know how real a threat if any that might be, but would there be any chance of getting into "back-door" copyright problems like this? John Carter (talk) 15:50, 27 July 2009 (UTC)

Any article that is a copyright violation in another wiki that is translated here will be a copyright violation as well. I have seen maybe one or two cases of this before. It might be worth having a template we stick in the foreign-language wiki talk page (would require discussion there obviously) after an article is translated to the point that copyvio could arise. ("XXX (1925-1987) was a German politician" obviously does not raise copyvio concerns.) Then if the foreign language article were detected as a copyvio we could be notified to delete the article as well. This might also be helpful for when articles are translated out of English by other wikipedias, so that en.wiki copyvios aren't infecting other languages. Calliopejen1 (talk) 16:12, 27 July 2009 (UTC)

These are worse case scenarios. Nobody has any intentions at least not until 2020 to auto generate articles about Italian adult film stars from Ugarit. We will be focusing on the higher quality articles or at least verifiable articles from the bigger wikipedias like German, Dutch, French and Spanish and on rather less controversial subjects mostly. In answer to Lady's comment about self referencing above. The idea is to find Reliable Sources used on other wikipedias to reference the information given. I think that tag may be acceptable for articles about places but seeimgly not for people, they must be verified by external sources and given. The best thing would be to start articles from other wikipedias and find external sources to support them. The problem as John said, might be to ensure that external sources support what is given in the foreign wiki article. Dr. Blofeld ^{White cat} 15:53, 27 July 2009 (UTC)

Yes, we can use the reliable sources from the other-wiki to reference it, but if we also import the structure, and translate the prose, we do need to attribute it somehow. I would prefer attribution by import, or in the edit summary, or on the talk page, but iw-ref can be used on the article page itself as well. –xeno^talk 16:00, 27 July 2009 (UTC)

Dr. Blofeld, I think you misunderstand me. I'm not suggesting any of that as a reference for the content of the article (I think we all know Wikipedias are not WP:RS), but to preserve the editing history. If we are talking about bringing work here that editors have done at other Wikipedias, those editors must be credited for their work. That's not to say that work is a reference for the article here. LadyofShalott 16:06, 27 July 2009 (UTC)

Ahh I see what you both mean, I agree! Legally if any text is translated in bulk from another wikipedia it must be attributed anyway, isn't that correct Jen? Dr. Blofeld ^{White cat} 16:07, 27 July 2009 (UTC)

yes, that's right, to comply with cc-by-sa. I prefer edit-summary/talk-page attribution to article-space attribution because other writers are not credited in article-space. I think it might be worth putting in a bug report about a uniform way to attribute translations/merges/splices throughout wikipedia, so it's machine readable. There was a discussion about this on the village pump a while back but it never got anywhere. Calliopejen1 (talk) 16:15, 27 July 2009 (UTC)

Yes... In part because the ghastly article-space translation templates were kept in the TFD... =) –xeno^talk 18:16, 27 July 2009 (UTC)

Gah, I know!!! Calliopejen1 (talk) 18:21, 27 July 2009 (UTC)

Editors here might be interested in a little page I wrote as I was working on the translation system here, Wikipedia:Translation/Overhaul - it describes some changes made, remaining problems, and some suggestions for improvement. It's obviously not coextensive with this effort but might be helpful background. Calliopejen1 (talk) 18:31, 27 July 2009 (UTC)

For the most part, I agree with a lot of what's been said here. And I tend to fall on Calliopejen's side, in that I agree with the sentiment that more is better. (I know I'm generalizing...it's a long thread, and I'm too lazy to go back and find the exact words. :-) ) There are two concerns that I wanted to address, however, as I haven't seen them mentioned.

• One: we need to give special consideration to articles in other languages. This will certainly be an issue when dealing with Wikipedias such as the Russian, Greek, Macedonian, Japanese, and Chinese. So...we need to find out what the transliteration standards are. I know that for some languages they have been codified. I would also recommend setting those articles aside in a special location, separate from the things with Western orthography. (Because a list of French, or German, or Spanish names, in the Western alphabet, can be created as-is, whereas a list of Russian, or Greek, or Japanese ones cannot.)
• Two: something that's a little more peripheral to the issue at hand...something that FloNight mentioned. I once, some while ago, advocated the creation of a series of bots that would cover some of those areas you mention that would be ripe for stubbing. For instance...one would create articles about landforms like rivers, mountains, lakes, and the like. Another would fill in some of the genus and species stubs that need to be created. (Both sourced, of course.) And there was talk of a bot for creating settlement articles...I'm sorry that never got off the ground. Anyhow, my point is that those sorts of things, I think, could be covered in other ways; this project would be ideal for starting missing biographies and things of that ilk.

These things would certainly be on the top of the list for "phase two" of this project. (Phase one being list generation, phase two being automated content creation.) --ThaddeusB (talk) 00:40, 29 July 2009 (UTC)

That's my two cents', anyhow. --User:Ser Amantio di Nicolao ^{Che dicono a Signa?}_{Lo dicono a Signa.} 15:47, 28 July 2009 (UTC)

Name discussion

WP:Wikiproject Importing articles from other Wikipedias? That might be a little unwieldy. Other ideas? LadyofShalott 18:13, 27 July 2009 (UTC)

was there somewhere WikiProject Transwiki was shotdown? –xeno^talk 18:15, 27 July 2009 (UTC)

We already have a translation page (without a corresponding wikiproject) and wikiproject echo (supposed to translate FAs, now all but dead. I'd like to see this (at least eventually) be part of a WikiProject translation with an FA department, a bot department, etc. Calliopejen1 (talk) 18:23, 27 July 2009 (UTC)

At the very top of this page, Dr. B. shows that Wikipedia:WikiProject Transwiki is already in use for transwikifying articles from here to other WPs. Of course, maybe this should be merged with that project???? LadyofShalott 18:24, 27 July 2009 (UTC)

Yes, the scope of that project is quite small. I would suggest embiggening it. –xeno^talk 19:24, 27 July 2009 (UTC)

Yes I agree. We'd turn that one into a full blown project likely. Dr. Blofeld ^{White cat} 19:35, 27 July 2009 (UTC)

Yes, seems like two sides of the same coin. It makes perfect sense to have both directions of transfer covered by the same project (and the project should be an inter-wiki one).--Kotniski (talk) 08:15, 28 July 2009 (UTC)

Except the current WikiProject Transwiki is about sending things outside of Wikimedia (i.e. stuff that isn't appropriate here, fancruft, etc). It wouldn't really make sense to merge this project with that one. Calliopejen1 (talk) 11:59, 28 July 2009 (UTC)

Transwiki could be the umbrella title for a number of things: 1. exporting inappropriate Wikipedia content to Wiktionary and other sister projects, or to other non-Wikimedia wikis; 2. exporting articles to other language Wikipedias; 3. importing articles from other language Wikipedias - this project; 4. Importing articles from other wikis, e.g. Wikipedia: WikiProject Citizendium Porting. Having them all under one roof would help with standardising practice and with awareness of each sub-project. Fences&Windows 16:39, 28 July 2009 (UTC)

I like it. F&W's suggestion makes sense to me. LadyofShalott 16:44, 28 July 2009 (UTC)

Posted to Blofeld earlier: Yes, I like the idea of WikiProject TRANSWIKI very much. I think another way of expanding the project and making it of real educative value would be to get schools, colleges, etc. involved too. I used to get so bored having to translate passages from Caesar and the like that had been translated and re-translated over and over again by many generations of school kids. It really turned me off both Latin and French. What I would have loved was a project working with others to translate something that had not already been done to death - something NEW and EXCITING! My proposal would be to try to get teachers to assign a translation project to create new (or better) pages for Wikipedia. A group or a class taking French could, for instance, translate some article from the French Wikipedia. To make it even more interesting and educative - it could be to translate something of historical, geographical, cultural, or some other interest - so it would be even more than just learning how to translate (and to use computers to edit something like the Wikipedia). Maybe they could get extra credits if they could get an article to good article stage??? Anyway, this is just an idea of mine - I am not sure how to get anyone or any group inspired enough to try it. Whadaya think? Cheers, John Hill (talk) 10:57, 29 July 2009 (UTC)

What should we call it then? TRANSWIKI is fine but I would suggest renaming the other project which deals with outgoing material and we occupy that space. Other than that a task force under Wikipedia:WikiProject Missing encyclopedic articles might be a good idea. Dr. Blofeld ^{White cat} 09:53, 31 July 2009 (UTC)

Major concerns

I realise this proposal is fundamentally flawed. If I write a summary of a source (using inline citation) which I have read, then I can vouch for that source personally, but if I transwiki another editor's summary, then I am blind copying. Since Wikipedia is a teriary source (a summary of reliable source), then blind copying of tertiary source undermines the whole concept or reliable secondary sources. It seems to me that we cannot assume that summarising a summary is anything more than plagarism. We cannot substitute research for reguritation. To write an article and comply with the spirit of WP:RS makes us duty bound to go back to the original source. If we are to stand on the shoulders of giants, then we have to use our own feet, not stand in another editors shoes. --Gavin Collins (talk|contribs) 12:56, 28 July 2009 (UTC)

Uh, do you know what plagiarism is? Reliable sourcing is one issue, but it is completely unrelated to plagiarism, which is using someone else's work without attribution. Calliopejen1 (talk) 13:10, 28 July 2009 (UTC)

Here's my one comment in response to this, to which I won't add any more. (Engaging you is futile, I believe.) Wikipedia is a work in progress. It would be preferable to have looked at sources firsthand, but it isn't required - all that is required is that something is sourceable. Translating some information is far superior than not having articles at all. There may be some errors, yes. (There are many errors elsewhere too, as I'm sure you know!) But you would rather not have this information at all? Especially if a translated article is sourced, a reader may be able to verify it himself. (And even if a translated article is unsourced, it is easier to verify whether something is true than to research from scratch.) I think you have a very narrow-minded conception of what Wikipedia is to be and how it is to grow. If we had put strict limitations on how articles could start, we would have far fewer articles than we do today and we--and our readership (i.e. the world)--would be much poorer for it. Calliopejen1 (talk) 13:18, 28 July 2009 (UTC)

I think I do know what plagiarism is, and if I understand correctly, transwiking articles wholesale is precisely that. You say that plagarism is using someone else's work without attribution. If I contribute to the article (say Enron scandal), then it is I who have written that contribution, even if I cite reliable secondary sources - I made those editorial decisions. Transwiking what I have written is plagarising what I have written. But setting this arguement aside for a moment, this process is still flawed because what is proposed is the use one tertiary source to write another, and we get into the realms of circularity sooner or later. --Gavin Collins (talk|contribs) 13:24, 28 July 2009 (UTC)

You are incorrect. Wikipedia content is freely available given proper attribution. As such, all that is required to avoid it being plagiarism is acknowledge of the original source (i.e. foreign language Wikipedia) --ThaddeusB (talk) 00:44, 29 July 2009 (UTC)

Wikipedia can't plagiarize itself. If done right, the only connection between the original article and the en.w article will be that the original was used to tell us that the topic existed. Abductive (talk) 00:53, 29 July 2009 (UTC)

Gah, how am I letting myself be drawn into this??? There is no plagiarism because the other wikipedias are attributed using {{Translated}} or similar. The whole POINT of free licensing is to be able to use others' work with attribution. Using another wiki introduces distance from the original source but unless the foreign-language article originally used en.wiki as a source (doubtful considering this process is designed to fix the problem of missing articles here) there would be no circularity. Calliopejen1 (talk) 13:43, 28 July 2009 (UTC)

One editor's vouching for a source is as good as another's, even if that other editor wrote on a different WP in a different language. If I take something from fr.wp, localize it and put it on en.wp, then it's just as if the original French editor had done the same. Unless you think that English-speakers are more trustworthy than foreigners, I don't see any logic in what you're saying.--Kotniski (talk) 13:54, 28 July 2009 (UTC)

Gavin, your point is well, pointless. As Jen said "The whole POINT of free licensing is to be able to use others' work with attribution". We should be working as one team translating each others work into a different language. Dr. Blofeld ^{White cat} 14:21, 28 July 2009 (UTC)

I think the process wouldnt be a direct word for word translation. If youve done translation before you realize that senatnces become something entirely different after theyve been transalted. But i can see the concern that others arent being credited. I think a way to deal with this, is of course a wiki proect template on the talk page, identifying that the article was orginially translated from another wiki. And additionally below that the source page kinda like a DYK template that says the date at which it was translated (a possibility perhaps). At that point the transaltion will make it a different sounding and written text anyway, no one transaltes word for word, otherwise you have a google translation document which is hard to read and understand. You need to consider the sources you have and write along those. My worry is about the sources within those articles being in other languages and being difficult to verify, but i dont think thats a huge problem. Anyway thats my two sense. Its free licensing on wiki. Thats part of the appeal to write here. Ottawa4ever (talk) 14:46, 28 July 2009 (UTC)

In answer to the Dr. Blofeld, I agree that copyright is not an issue here as the article plagarism explains. In answer to Ottawa4ever, even if we assume that we can obtain a perfect translation (which I think is a stretch of the imagination, but set this aside for a moment), the key issue here is that transwiking involves blind copying of another editors work. Perhaps the intent to plagarise is not present, but the result is the same, even if there is no obligation to give the original editor acknowldegement. However, to say that we don't have to provide attribution for another editors work is a doubtful arguement from an intellectual perspective, and is not one that is accepted, say, for images. If I were to upload a library of images which are freely licenced, I still obliged to say what the origin of each image is.
Having said this attribution is not my main concern. For me the big issue is blind copying, which I think is wrong, no matter how perfect the translation. I think editors can use another non-English Wiki for research guidance, but I can't see how you can comply with WP:NONENG if an article does not provide a translation of the the original source itself. --Gavin Collins (talk|contribs) 15:29, 28 July 2009 (UTC)

WP:NONENG only requires translations for direct quotes. Obviously if we translate a quote, the original quote will be there in the original language in the foreign-language wikipedia. I don't know why you're going on about plagiarism. Other editors will be attributed. Plagiarism is negated by attribution. One contributes to wikipedia knowing it will be copied, because that is the point of wikipedia. (See the edit screen: "If you don't want your writing to be edited mercilessly or redistributed for profit by others, do not submit it.") Calliopejen1 (talk) 15:51, 28 July 2009 (UTC)

(EC) @Gavin, How on earth did you arrive at that conclusion from what is proposed here? We're talking about translating articles, which includes reading and evaluating the sources provided in the foreign language article. There's no blind copying at all. How can you possibly claim that WP:NONENG should be in the way of such an effort? Vyvyan Basterd (talk) 15:54, 28 July 2009 (UTC)

Piling on because it seems a major point is being missed: it is absolutely and totally in the plans of this proposal to attribute the editors who created the articles on other Wikipedias. If you read above, we've been discussing exactly how to do that (not whether to do it, but the details of how). There is total, unanimous agreement that prior editors must be credited for their work. LadyofShalott 16:12, 28 July 2009 (UTC)

Gavin, nobody has even said every article is going to be fully translated and copied!! On a lot of articles they will be started translating a paragraph or two and will probably be written following this by whatever editor in their own words and using reliable sources. I would imagine a high proportion of cases will involve the developer afterwards seeking out sources across the web and in books to verify the information and make the article their own. Who've lost what the proposal is about. It is about starting missing articles from other wikipedias and then allowing people to develop them afterwards in their own right. A lot of them time this will be outside of the projects parameters as traffic come across the solid starter articles the project will start and then add what ever information they think is needed. It really wouldn't anyway if entire articles for copied providing they are supported by reliable sources. that's the whole point of the licensing system on here to allow work to be represented, certainly within the wikimedia group. All of our 12 million articles are everybody's articles, they were just started in a different language, I can't emphasise enough that we are really one project. Anyway, hundreds of articles are "lplagiarised" by other wikipedias everyday and you know what? The information is making a major difference to education and knowledge in developing countries as people gradually gain access to the Internet. They benefit immensely from having our good articles translated into their language, the same for us and any project that is missing good content. I think this particular discussion is again irrelevant and should be discontinued as it is not a valid concern. Dr. Blofeld ^{White cat} 16:12, 28 July 2009 (UTC)

Thank you for refocusing us here. It is my understanding that the idea of the project was to start articles on notable topics that don't exist yet on en.wikipedia.org. It is much less intimidating to expand a stub than start from scratch, IMO. Thus the idea is if a stub exists with basic facts translated from another Wikipedia editors are much more likely to expand the article than if the information didn't exist at all. This is where the bot comes in. With present technology, a computer program cannot reliably translate text. However, what it can do is identify key information that follows a specific pattern and reliably translate that. Most articles start with a formulaic sentence that gives the reader some basic information about the topic. Infoboxes have a set format. Other types of facts are always presented in the same way. It would be these areas that the bot would serve. (See fuller explanation below.) --ThaddeusB (talk) 00:52, 29 July 2009 (UTC)

In answer to Calliopejen1, I think your interpretation of WP:NONENG is somewhat broader than mine. And I quote:

Where editors translate a direct quotation, they should quote the relevant portion of the original text in a footnote or in the article. Translations published by reliable sources are preferred over translations made by Wikipedia editors.

I think the key to understanding this is that, where you cite a source in a foreign language, you are supposed to cite the orginal source as well, so that other editors can establish the context and that the translation is correct. What is being proposed does not meet this requirement; a translation of the article itself is nto the same as quote the relevant portion of the original text. --Gavin Collins (talk|contribs) 18:05, 28 July 2009 (UTC)

What are you talking about? WP:NONENG is about use of sources in foreign languages and quotes in foreign languages. This is an attempt to have humans translate and adapt articles on other Wikipedias and review the sources to ensure that everything is up to standard. This has nothing to do with what WP:NONENG was invented to deal with. You're grabbing that piece of WP:V out of context. Vyvyan Basterd (talk) 18:24, 28 July 2009 (UTC)

My reading differs from yours in that case. My reading is that WP:NONENG presumes that you are oblidged to go back to the original coverage from the reliable secondary sources and provide the text in its original language. "An attempt to have humans translate and adapt articles on other Wikipedias" without the original sources is simply a cheap way of circumventing this requirement, i.e. blind copying. There is no way that you can provide evidence that a review of the sources "to ensure that everything is up to standard" has taken place if you don't comply with WP:NONENG. --Gavin Collins (talk|contribs) 18:40, 28 July 2009 (UTC)

Sorry, this is not a reasonable reading of the policy at all. How has WP:TRANSLATION managed to exist for basically the entire history of wikipedia if what you're saying is true? I can't argue with blatant misunderstanding. Calliopejen1 (talk) 18:43, 28 July 2009 (UTC)

I'm sorry Gavin but that's not the established consensus at all. I suggest you post on WP:RS/N if you're unsure about this because you've misunderstood something fundamentally important here. Vyvyan Basterd (talk) 18:44, 28 July 2009 (UTC)

I am not sure how you have arrived at that conclusion because the text of WP:NONENG is pretty clear - you have to provide the text of any foreign language sources cited in an article, not just a translation of the article itself. Since the blind copying of Wikipedia articles only involves the copying of the text of the article (which is a summary of the sources) and not the text of the sources themsleves, then you will have to propose an amendment to WP:NONENG to be fully compliant with WP:V. --Gavin Collins (talk|contribs) 08:54, 29 July 2009 (UTC)

Either you don't read what I write or your understanding of policy is appalling. I've told you there's no blind copying. The article is translated, the translator checks and verifies the sources/references provided in the foreign language version of the article and reuses them in the English version of the article. There's no blind copying and the sources are cited in their original version. Please either familiarize yourself with what is being discussed and the relevant policies and guidelines or disengage. This isn't the first time this week you refuse to get the point, should I take this behaviour of yours back to AN/I for review or are you going to take the time to familiarize yourself with the topic before wasting anymore time on this page? Vyvyan Basterd (talk) 09:57, 29 July 2009 (UTC)

I would like to agree with you, but the problem still remains that there is no verifiable evidence that the sources have been transwikied properly unless you provide the text of the original source, so from a Wikipedia perspective, it is still blind copying. For example, if an article is transcribed from another Wiki, with edit history to provide attribution, what use is that to the editor who may be reviewing the provenance of the original sources? What you are proposing still does not meet the requirements of WP:NONENG, because the full text of the original source is not being transcribed in the process is being proposed. Note to Vyvyan: Ad hominen attacks are not really necessary here, as this is not a discussion about personal matters. --Gavin Collins (talk|contribs) 11:29, 29 July 2009 (UTC)

Example: Hero Corp

Gavin, please check if this article I just created, Hero Corp from the French Wikipedia is in any way deficient. Abductive (talk) 11:35, 29 July 2009 (UTC)

There are several problems. Firstly, your inline citations are not working correctly - there is no way to match the references sections with the text of the article. Secondly, the link to the first citation is broken. Thirdly, but not least of all, is that notability has not been established in accordance with WP:GNG. --Gavin Collins (talk|contribs) 11:47, 29 July 2009 (UTC)

Gavin, I am implacable in my belief in reliable sources, so rest assured; Hero Corp has 32 Google News hits. I'll remove the dead link. Abductive (talk) 11:54, 29 July 2009 (UTC)

@Gavin, what are you talking about now? What "evidence"? The ref will be there, just as it would on any other article. What evidence? Have you read the reliable sources guideline? I don't need to provide "evidence" of what is in the source, in fact I could cite sources that aren't publicly available to wikilawyers such as yourself and it would still be valid. You don't have a case and it's time you realize why everyone else disagrees with you. Vyvyan Basterd (talk) 12:02, 29 July 2009 (UTC)

No, Gavin was right, a ref was stale because the link was dynamic. I replaced it with a working ref, and I am sure the article is on a notable topic. As a matter of fact, it was the most-viewed article on the French Wikipedia for which no English equivalent existed. But Gavin is wrong if he thinks that a flood of non-notable articles will be unleashed on en.wiki. Abductive (talk) 12:27, 29 July 2009 (UTC)

That's why the refs should be manually checked when translating articles. I've pointed this out several times to Gavin now. If we do that then the language of the ref doesn't matter. He repeatedly refuses to listen to those who've tried to explain why he's wrong about this. Instead he's achieving what he desired which is to stall this proposal by repeating the same counter arguments over and over. Vyvyan Basterd (talk) 12:36, 29 July 2009 (UTC)

I recomend that the ad hominen attacks cease, as they reflect badly. If you disagree with me, that is just fine, and if you can provide reasons for the disagreement, then I can respond them. In answer to Abductive, notability is not supported by the sources cited in the article, and google hits are not admissible as evidence. For instance the citation from tvmag.com is not a reliable source, and in any case the person who wrote it is not named. --Gavin Collins (talk|contribs) 13:35, 29 July 2009 (UTC)

I doubt that a broadcast TV show in a G7 nation is not notable. The main point of the example was to show that these articles will behave as any other. For example, you could tag the article for sources, and people would add sources. You could take the article to AfD, and it would be debated fairly. In no way are these articles inscrutable. Just to demonstrate, I added a reference to the show in Première, an impeccable source. Abductive (talk) 14:28, 29 July 2009 (UTC)

I would tend to agree with you, but you have not provided evidence that Hero Corp is notable, and if this is the best that transcription can achieve, then I can't endorse this proposal. Both the producers of the show and publishers of Premiere magazine are owned by Lagardère Group, so they affiliated if not under the same mangement, and are not independent. Unfortunetly, you have picked on a bad example, which confirms my concerns about blind copying of articles, namely that there is no basic quality control as was the case in Wikipedia:Articles for deletion/Claus Peter Poppe, which set a bad precedent. Clearly there is going to have to be better quality control, otherwise transwiking will result in the importation of articles with content and notability issues in the enWP. Hopefully, you will be able to come up with a better example of a successful transcription. --Gavin Collins (talk|contribs) 17:35, 29 July 2009 (UTC)

Please, if you can find out this sort of information about the show, haven't you demonstrated that it is notable? I ended up with Premiere because this source wouldn't let me copy and paste. I made no claim that this is the best the project can achieve. As for the problems with the Claus Peter Poppe et al., it was BLP concerns that sank that ship, not notability. Had those stubs been on 17th century musicians, they'd be sitting there until Wikipedia folded. Abductive (talk) 19:03, 29 July 2009 (UTC)

I've added your information about the production company's ownership by Lagardère Group, with a {{fact}} tag, could you please provide the source(s) for that? Abductive (talk) 19:34, 29 July 2009 (UTC)

You added it, that is down to you. --Gavin Collins (talk|contribs) 08:21, 30 July 2009 (UTC)

A case example and the potential problem that does remain

Gavin, here's an example for you (going the other direction, but the principle is identical):Faith Hunter is a (stubby) article that I have written here on enwp. Someone over at frwp decided they needed it. My words were translated directly (which may or may not be what happens with any given article under this proposal). There is a template at the top of the page which says it was translated from enwp. This is not plagiarism of my work. Why not? It says exactly where it came from, and anyone who cares to do so can come over to enwp and look at the editing history here. It is preserved. LadyofShalott 16:35, 28 July 2009 (UTC)

Now the potential problem I see: Someone at enwp could look at Faith Hunter and decide she is not notable enough (this is not a suggestion, folks! ;-D); the user nominates the article for deletion and the consensus at AfD is to delete it. The article remains at frwp, but now the editing history is lost - in this scenario, no one who saw the frwp article would be able to see all the work I put into the article. Without importing article histories, how can we prevent this scenario from happening? LadyofShalott 16:35, 28 July 2009 (UTC)

One idea is some sort of uniform template (like I suggested above) to record that the article has been transwikied, that would be placed on the talk page of the source article. It might be worth encouraging all wikis to do this together. Calliopejen1 (talk) 16:51, 28 July 2009 (UTC)

But would the template prevent deletion from the originating project to ensure preservation of history? LadyofShalott 16:55, 28 July 2009 (UTC)

It might ensure that the history would be copy-pasted to a subarticle or something like that, before deletion. (e.g. fr.wiki history pasted to en.wiki subarticle, linked from en.wiki history or talk page, before fr.wiki deleted) This obviously would require interwiki coordination though. Calliopejen1 (talk) 17:09, 28 July 2009 (UTC)

What a refreshing change it is to have some more female input on this site! There are too many guys here, not enough female editors. Makes a welcoming change! A neutral encyclopedia needs articles written equally by both, it is a shame that this site doesn't attract more female editors. Sorry, it was just very rare to see a section with two women sharing their views that's all! I agree 110% with you about encouraging this project on other wikipedias. Dr. Blofeld ^{White cat} 17:06, 28 July 2009 (UTC)

In answer to LadyofShallot, you make a valid point, but I am not sure how notable Faith Hunter is in France, and just because she has an article in enWP does not establish this. Your article suggests that she might be notable in the English speaking world, but has her work been published or even read in France? I have voiced my concerns about blind copying in the previous thread, but here I will limit myself to a discussion about notability. I can't say I have all the answers, as this is unexplored territory in terms of the notability guideline, but if French sources have not noted her, I am not sure she is notable within the the context of French Wikipedia. Therfore is one article in enWP is enough perhaps to fulfill the entitlements for a standalone article? In the AlbertHerring case, I would ask the same thing - how many German politicans in local government are notable outside the context of Germany? I can't say, but if there is no published source in English, then I would argue that foreign sources, no matter how reliable, aren't a really a substitute. I am not saying that we should not have transwiking at all, I am saying that maybe this is unnecessary duplication based on the mistaken idea that notability can be inhertited from another tertiary source, such as enWP. --Gavin Collins (talk|contribs) 18:32, 28 July 2009 (UTC)

You really don't seem to understand a lot of policies here. Now it's notability. Just like notability is not temporary, it is not geographically or linguistically bounded. There is no such thing as someone who is notable in French or France who is not notable in English or the US. If she is notable in English, she is notable in French, and vice versa. (Though a discussion on fr.wiki and en.wiki might not necessarily lead to the same conclusion, mostly by chance.) No one is saying that a French person is notable by virtue of being in French Wikipedia. In fact, it is the reverse! Someone is in French wikipedia because they are notable. (If there is a wikipedia that has markedly differnet notability standards call it to our attention so it can be excluded from this translation drive.) Yes there will be a couple mistakes here and there, and they will be deleted as non-notable in due course. Essentially, translation draws the reasonable assumption that nearly all of the article in fr.wiki meet en.wiki's notability standards, because nearly all of them meet fr.wiki's notability standards. Calliopejen1 (talk) 18:39, 28 July 2009 (UTC)

By the way this is nowhere near "unexplored territory" but has been conclusively established in countless deletion and policy discussions. Calliopejen1 (talk) 18:41, 28 July 2009 (UTC) See e.g. Wikipedia talk:Notability/Archive_21#Multilingual_notability. Calliopejen1 (talk) 18:46, 28 July 2009 (UTC)

(EC) @Gavin, as far as notability goes there's no requirement that the sources must be in English. Have you read the relevant guidelines before you posted this or are you just asking because some of the things you're posting here baffle me. Vyvyan Basterd (talk) 18:39, 28 July 2009 (UTC)

As long as the foreign article is properly sourced, notability is assured. There might be some concerns about WP:V if no English sources are available. Abductive (talk) 21:58, 28 July 2009 (UTC)

I think verifiablility can be done with some of the editors in a manual process when they go through these articles for creating in english wiki. Obviously people will need to be semi familliar with the other language if they plan to translate, but they can get help from other members of other language wikis. This is why i like this idea so much. It will encourage alot of communication between different language wikipedias. Some foreign sources could be checked by members of those projects for us, and vice versa for them. I think the key though is a matter of coordination with the other wikis still. Just some random thoughts hereOttawa4ever (talk) 22:13, 28 July 2009 (UTC)

Agree with Abductive. As long as the other WP projects source the article and meets the general notability criteria in their respective language, the subject is notable. Sources don't have to be in English on the EWP. If some obscure (to EWP) person from x country is notable in x country, they are notable full stop. Lugnuts (talk) 08:24, 29 July 2009 (UTC)

I think I would have to accept the argument that it is not geographically or linguistically bounded. But whether or not Faith Hunter is a notable author, it still seems strange to me why you would want to create an article about her in French Wikipedia if none of her work had been published or read there, but I accept that despite this, any reader might still want to know about her. --Gavin Collins (talk|contribs) 09:25, 29 July 2009 (UTC)

I think it has. Nevertheless, there are French people who read English-language books. And the why doesn't matter; I chose this example precisely because someone at frwiki did want to create the article, for whatever reason. If you think it matters why they want it, you'll have to go there and ask. However, are there people whose work English-only readers would like to be able to read about even if they have never been translated to English? Sure, Pluk van de Petteflet is an article I quite enjoyed reading, even though that book and the movie based upon it have never been translated to English. I don't read Dutch, but I'm glad someone wrote the article in English for me to read. LadyofShalott 12:38, 29 July 2009 (UTC)

How I envision this proposed process working

Since I am offering my programming skills, I think I should make it pretty clear how I envision this process working...

Phase 1

In phase 1 areas that might produce good results are identified by human editors. This might be "German politicians" and "Cities in Germany" on de.wikipedia and "French painters" and "Geographic features of France" on fr.wikipedia. Eventually all relevant categories (i.e. proper nouns) would be "raided", but we could start with the most promising.

Once target areas are identified, the bot would go and find all articles on xx.wikipedia that don't exist yet on en.wikipedia. It would create a list of these article for project members to review.

Phase 2

In phase two human editors work to determine common elements of a given category of articles and prune out anything that clearly doesn't belong. With help of translators, I then write a set of rules to accurately copy and translate key information from a foreign wiki. For example, for a city this would include things like location, population, date founded, elected officials, etc. There would be no requirement that every piece of info be in every article, so the list of capturable information could theoretically grow quite large.

The bot would then use the captured information (which would include the reference section of the foreign article) to generate a basic stub. The resulting stub would include both attribution and expand-via-translation templates. Since it would only be copying basic facts, things like potential copyright concerns wouldn't matter. Ideally it would put the resulting article into mainspace to maximize its chance for expansion. (Although, it would probably start by putting article into Wikipedia space until people grew comfortable with the idea.)

Phase 3

Lists of articles created in phase two would be created for those interested in expanding and/or watchlisting them. Expansion would then proceed as "normal" with editors adding content (either through direct translation or via research on their part) as they see fit.

Obviously getting the bot "right" is a crucial aspect here. In that regards, I would like to assure the community that I always proceed with great caution when making bots. I view accuracy as the number one priority of any bot, as bot edits are rarely scrutinized and can be exhausting to undue (since they can make so many edits so fast.) Of course, if this is to happen the bot will have to go through the normal approval process which will allow both technically minding people and average users to express their concerns and strengthen the code.

I look forward to working with interested parties to make this happen. Please do ask any questions you may have. Thank you, ThaddeusB (talk) 01:21, 29 July 2009 (UTC)

• Can the bot grab information such as length of article, number of citations in article, number of editors of article, number of page views of article or any similar metric that would help us assess how well-developed and reliable the article is or how popular the topic is? For example, I went through the German Wikipedia's top 1000 page views list and found three articles out of the first 250 there; Sturm der Liebe, Triftbrücke, and Concordiasee (Dr B made one for Concordiasee today) that don't have articles here. Abductive (talk) 03:02, 29 July 2009 (UTC)
  • Yes, it could certainly grab one or more statistical measures of an article's "quality"/interest level. Where did you get the top 1000 list? http://stats.grok.se/de/ is almost a year out of date and Triftbrücke isn't listed, so I'm guessing it wasn't from there. :) --ThaddeusB (talk) 05:29, 29 July 2009 (UTC)
    • http://wikistics.falsikon.de/latest/wikipedia/de has the last month. Abductive (talk) 05:45, 29 July 2009 (UTC)
    • You know, some of the other Wikipedias have an article rating system like our FA/GA/A/B/C/Start/Stub system. Would it be possible to collect that info somehow? Abductive (talk) 06:06, 29 July 2009 (UTC)
      • It certainly can. --ThaddeusB (talk) 03:28, 31 July 2009 (UTC)

WikiProject TRANSWIKI: Possible involvement of schools and colleges - providing an educational tool

Here is a note I posted on Dr. Blofeld's discussion page which he thought might be of interest here:

Hi again! Yes, I like the idea of WikiProject TRANSWIKI very much. I think another way of expanding the project and making it of real educative value would be to get schools, colleges, etc. involved too. I used to get so bored having to translate passages from Caesar and the like that had been translated and re-translated over and over again by many generations of school kids. It really turned me off both Latin and French. What I would have loved was a project working with others to translate something that had not already been done to death - something NEW and EXCITING! My proposal would be to try to get teachers to assign a translation project to create new (or better) pages for Wikipedia. A group or a class taking French could, for instance, translate some article from the French Wikipedia. To make it even more interesting and educative - it could be to translate something of historical, geographical, cultural, or some other interest - so it would be even more than just learning how to translate (and to use computers to edit something like the Wikipedia). Maybe they could get extra credits if they could get an article to good article stage??? Anyway, this is just an idea of mine - I am not sure how to get anyone or any group inspired enough to try it. Whadaya think? Cheers, John Hill (talk) 21:20, 29 July 2009 (UTC)

I agree strongly. I had tried to do this by including an "involve your language class" link on the {{Expand French}}-type templates but I took it off because maybe it was a little too in your face... (Thoughts?) One problem I see is that generally students learning French would benefit more by translating into French than out of it, so even better targets might be native French speakers who are learning English. (Maybe an exchange between classrooms could be done where one is translating into French and another out of it?) I'm not sure what the best way to do outreach for this would be, but I agree that language classes is where we could get our needed army of translators. Calliopejen1 (talk) 21:25, 29 July 2009 (UTC)

This would be wonderful - if done well. The best example of of combining editing with a classroom experience is WP:MMM. I'll drop a note on JBMurray's talkpage, and ask him to come comment here. LadyofShalott 22:16, 29 July 2009 (UTC)

Wikipedia:School and university projects would be the main project with which this one would coordinate any such activity. I did write Jbmurray a note, but his page says he's busy with RL and may be slow to respond. LadyofShalott 16:39, 30 July 2009 (UTC)

Some ideas on how to get it going

I just replied to a note from Dr. Blofeld on my Discussion page and thought I should post it here as well:

There is a massive number of schools and universities and tutors teaching languages in the world - and I think if we can get some good early success stories, it may just pick up enough momentum to keep spreading on its own. This is a really important idea - I really believe in it. I think the potential is endless. I think as soon as people see what a powerful (and free!) teaching tool it could be - it could catch on in a big way. The Wikipedia is the perfect tool for this sort of international and interlingual collaboration. Wow! Wouldn't it, couldn't it be great!

Getting it up and started, on a firm grounding and reaching critical mass is the trick. Surely we have some educators amongst the editors who would run with the idea with enthusiasm? I'm too old and sore myself to do much running these days ;^) but, hopefully some of you energetic younger sparks can provide the energy and direction needed????? Here's hoping! John Hill (talk) 05:43, 30 July 2009 (UTC)

Good idea, it is like a ghost town on wikipedia today. Dr. Blofeld ^{White cat} 16:25, 30 July 2009 (UTC)

School's out for summer. The recruitment of translators should be addressed, but let's get the project going and they will come. Abductive (talk) 17:39, 30 July 2009 (UTC)

I would suggest someone go ahead and start setting up the page. This is a WikiProject proposal after all and doesn't need formal community approval, just 5-10 interested editors. Clearly that level of commitment has been reached. I suggest we pick a name, get the basic page structure set up, and then decide what our initial targeted areas will be. There is no need to debate endlessly the potential problems of this endeavor. Only actually getting started will reveal any real problems and those can be addressed as they come up.

It will be easier to recruit translators and such once the project is going. --ThaddeusB (talk) 03:37, 31 July 2009 (UTC)

Extended conversation moved to Wikipedia talk:WikiProject TRANSWIKI - We're all set up now! As such, conversation about how to proceed should take place there. --ThaddeusB (talk) 02:24, 1 August 2009 (UTC)

Wikipedia:WikiProject Echo

Isn't this what Wikipedia:WikiProject Echo used to do? I don't know if they're active anymore, but it might be a place to get some ideas. -- Ned Scott 11:37, 31 July 2009 (UTC)