From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search

Using Wikidata queries[edit]

Try using Wikidata queries to accomplish this. PiRSquared17 (talk) 18:02, 18 July 2013 (UTC)

This looks like a Wikidata-tool for me. --Goldzahn (talk) 16:01, 19 July 2013 (UTC)
But maybe its to much, when all the data necessary for identifying is beside the structural data on the species in Wikidata? Greetings, Conny (talk) 09:12, 20 July 2013 (UTC).
I agree that Wikidentify sounds more appropriate to pursue as a(n awesome) tool than a new Wikimedia "project". In other words, it doesn't sound like Wikidentify would need its own MediaWiki deployment, etc. This tool could get infrastructure support from Wikimedia Labs -- about which you can talk with knowledgeable people in real-time at #wikimedia-labsconnect.
I also agree that Wikidata queries would be ideal for this. In fact, Wikidata queries largely are this -- if anything Wikidentify sounds like it would be a compelling application / user interface for Wikidata queries. Emw (talk) 15:11, 20 July 2013 (UTC)


Als ich das las dachte ich eine Interessante Idee aber hier ist mal wieder wie üblich alles Englisch. dann nicht. Im Kurierartikel sollte darauf hingewiesen werden das das Projekt in Englisch ist. --Mauerquadrant (talk) 02:25, 19 July 2013 (UTC)

Sie können es übersetzen, wenn Sie wollen. Ich bin ein Übersetzungsadministrator auf Meta. Ich kann diese Seite zum Übersetzen freigeben. But it's probably easier to ask a de-N translation admin, e.g. User:Vogone, User:MF-Warburg. PiRSquared17 (talk) 02:43, 19 July 2013 (UTC)
Sowas sollte automatisch geschehen. Die WMF sollte für die Kommunikation mit sagen wir Usern aus den 10 größten Wikipedien Übersetzer einstellen. --Mauerquadrant (talk) 03:45, 19 July 2013 (UTC)
Sorry for replying in English. My German is not very good, so I hope my using "Sie" instead of "du" was not offensive. I just learned that "du" is normally used in Wiki[p|m]edia culture. Anyway, just click here to translate. I'm not sure about the auto-marking for translation thing. It would be difficult to implement, but it sounds like a great idea. PiRSquared17 (talk) 04:00, 19 July 2013 (UTC)

Done, thank you for advice and support. Um so viel wie möglich Interessenten anzusprechen, ist Englisch die Sprache der Wahl um Projekte bei Wikimedia zu kommunizieren. Jedoch ist die Software oftmals so ausgelegt, dass einer Nutzung und Entwicklung in der jeweiligen Sprache nichts im Weg steht. Grüße, Conny (talk) 06:14, 19 July 2013 (UTC).

other placed talkings[edit]


Nett wäre das auch wenn dies auf geografische Orte anwendbar wäre. Auf einem Bild von 1850 sieht man eine Kirche mit einem See in direkter Nähe, vermutet, dass es sich auf dem Gebiet des heutigen Baden-Württembergs befindet und kann mit diesen Angaben eine Auswahl von in Frage kommenden Objekten erhalten. --Friechtle (talk) 13:44, 20 July 2013 (UTC)

Sense of the project[edit]

What should this project do? The data is on Wikidata already. All we need is a tool to get this data in the way you described it. In addition, I think this interferes with Wikidata's phase 3 with the queries. Best regards -- Bene* (talk) 21:31, 20 July 2013 (UTC)

Basically right, but:
  • We need as one part of Wikidentify identification "ways", when there is need to describe key by key to get right object. These ways need to be Items - at this time policy gives no way to create them.
  • An other problem is, that we will have very much statements on Items (f. e. in biology), which maybe will overload normal view of items in the way of Wikidata. Nobody wants to scroll half a minute to find the taxonomy facts inside a big pool of identification details f. e.
  • A interface optimized on the task of identifing makes Wikidentify seperate (or there is a skin).
  • Maybe we need professionals who are helping Wikimedians to construct senseful identification ways (f. e. in botany) - Wikidata can not lift this.
Greetings, Conny (talk) 05:55, 21 July 2013 (UTC).


There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy

This seems to me very much like a project of the 19th century. Don't get me wrong, I don't want to criticise you for that, but the world is not made of concrete objects with a given location, volume and mass – however, it seems to me that this assumption is implied in this project. I guess it'll work fine with what's already fed into wikidata, that is towns, buildings, continents and so on: everything that's fixed in space or maybe also in time (e.g. an event that took place on one single day). However, there are limits to "universal" (that's why identification keys usually only cover a small range of subjects). How, for example, would you fit in a quark, a dinosaur or a specific mathematical function? What if a certain thing changed its location, volume or mass or traveled in time (and yes, there are some things that seem to travel backwards in time)? What about ideas, institutions, distribution networks, book translations, state borders, dances or international conventions?

I'm not asking this because I think your project won't work out. It's pretty much a reverse Wikipedia, so I see real chances. It will, however, much more complicated and twisted than the front page implies and there's a hell lot of work to it. If it is to function properly, you'll at least need as many people to run it as are currently needed for a medium sized Wikipedia, because all the entries and steps of the universal key will have to be maintained constantly. And if successful, prepare yourself for all the stuff that's already going on at Wikipedia: vandalism, a huge meta sector, political campaigns etc. --Toter Alter Mann (talk) 21:41, 21 July 2013 (UTC)

Wikidata Ways in my feeling can not fit all objects on earth. Changing location should be no problem, we can update the keys or mark them as outdated or give them a existtime, so the collapse automatically after 15 years f. e. Changings in volume and mass give us now the possibility to understand the way things are changing. In Wikipedia you can do this via the history, in Wikidentify hand in hand with Wikidata you could put it out sourced as a change in life. If there are two Ways or way get changed have to be discussed. Timetravel should no problem for defining. If we get so much dimensions in the keys we can handle at this time, it will work. Ideas have their times, book translations their specific habits. It looks like much work.
In my feelings there will be much discussion, interessting is that people living this dream understand on that way their lifes better and better. After long discussions maybe there is a small identification way as result. And then it's possible to improve this, step by step. Wikidata should not be masswork, it should be restictive manner and universally applicable for different peoples with different languages. So it must be well reasoned and very slow on some points. I think we need levels of trust, we need playgrounds and experts to identify the world. Reverse Wikipedia is a great definition :D .
The mechanisms in Wikidentify will set, how many people are necessary to manage it. Maybe reader gets only proofed ways and the playground is much much bigger. But an open concept like Wikipedia I would prefer. Shut down, if nonsense is overwhelming is always possible ;) . Greetings, Conny (talk) 06:38, 30 July 2013 (UTC).


Do I get it right that you don´t intend a querie, but a key? Have you read the article en:identification key? then you should have met some of the numerous problems to solve. First, if the key ist not about a class of relatively few, countable objects (e.g. continents), then you will have to deal with incomplete coverage. Most of species (and most other objects interesting enough to look in a key, instead of full-text search) don´t have an useful entry in wikipedia, and many of them never will. If you program a querie-like tool, it will give you any matches found - or no matches. It doesn´t give you any item not included in the database yet, or any hints about them or even their existence. A severe problem of every "simple" key, e.g. most picture keys, is wholly unwarranted overconfidence in the results - if I have found an entry which matches my criteria (e.g. a picture of a flowering plant that looks like my specimen) I´m quite happy - but there may be dozens of other species which are not yet included in the key. You can see some of the problems in the (not altogether bad) keys in "Offene Naturführer". If I use the snail-key, there is no safeguard into entering with an immature specimen - and if I´m no naturalist firsthand, how will I know? - leading to wrong results. Most keys refer to "common" species. If you enter with an "uncommon" one, you will often get a wrong answer.

And: even with simplicity as a main objective, the keys are riddled with specialized terminology not easy to comprehend. This is, because with too simple a querie (snail, 2.5 centimetres wide, somewhat off-white coloured") you get too many useless results and are not better off than with no key at all. And to make things worse: "my" species may not even be included. And we haven´t started talking on languages altogether: there is not even a German match for "snail" yet.

What´s all about? The real world is a difficult place with zillions of objects, most of them quite "untypical". If I try do describe them in an uncomplicated language,I will not get far - the technical terminology was not invented for fun or to look down on people who aren´t "us", but to solve real problems simple language wan´t able to cope with. And you can build an somewhat complicated, but exact key, of limited scope (e.g. of all Brassiceae growing in Britain) not easily, but well founded. But to many organisms there is no key, simple or complicated, at all, because scientists themselves don´t feel competent enough to produce one. And in biology, coverage tend to be somewhat better than in most other topics. To produce a key that covers everything you need a world model, including the gaps. If you want an useful introduction in problems you would have to cope with, look at about real problems real people tend to have.--Meloe (talk) 17:40, 29 July 2013 (UTC)

Maybe "have you read" is for me :) . In remembering - this prosposal should be the possibility for every user to give ideas, formulate problems and solve things. Wikidentify is not existing, if, it will be constructed here... I did not read the article completely, but this is not the point. I know how to handle a identifcation key in biology and I have visions for us. The problems of keys can maybe handled in an other way, past is not showing us...
Yes, there often can be many not documented objects in Wikidentify. Maybe its possible to calculate the probability of the distribution of the quantities to get better feeling how save the answer is. I would hide wrong results by adding a proof identification way only, if there is unambiguity in the each mated identification universe. It's much more better to get the genus of a species, than wrong details - right! But a genus identification system in f. e. biology is also a big thing!
To many results can be filtered down with timepoint, state of spread, object is living and f. e. geographical facts. When we get to many results, than we put out the category... We have to think, what the way should be - putting much data in it and only use excluding and so on is in my opinion not the way. People should care and maintain their identication ways and should decide the sense of adding one more now object in there ways... Not mass integration!
Untypical is a good point, than we have to focus down the sharpness of the keys. And than like told there will be not object, but a object class at the end. The final definitions could be made by profis. Wikidata should people give the possibility to understand what is arround, to question of wood my chair is, to understand what are the differences of metal or smells. If we take a look on food - the following thinkings after be able to differ the food I counsume and read about their contents will open thinkings for the world surrounding is more and more. Please don't see it just like a biologist.
If Wikidata can ensure, that people use identification ways created by other peoples to get some more informations about the things arround us, if there is nobody we can ask, than it will win. It's not a question of the detaildeep in the first step, its a question if people say, it is possible and worthful. I can't answer for specific questions - if there is no one here to do, than maybe it is not the right time jet for Wikidentify...
Thank you for the critics, Conny (talk) 06:25, 30 July 2013 (UTC).
I didn´t intend to be wholly destructive, but please keep the paramount difference beteeen queries (which are of Databank entries) and keys (which are of real world objects) in mind. Building meaningfull object classes is more than half the solution to query-like problems. The problem is linking real objects to the classes. To build a key which gives meaningfull intermediate resultates like genus or family (called a synoptic key) is much more difficult than building a workable key at all. For myself, I´m not able to use many existing biological keys, because I haven´t got enough knowledge about the included organisms, and the specialized terminology needed to describe them - and I think, I´m more familiar to keys than most users will be.-- 08:12, 31 July 2013 (UTC)

IRC come together[edit]

What do you think - maybe we can do an IRC hour or meet together in Hangout for creating more ideas and formulation sense? Greetings from Dresden, Conny (talk) 13:46, 1 August 2013 (UTC).


Are there people who can smalltalk idea in Hong Kong and give some feedback from the movement out there? Greetings, Conny (talk) 05:58, 6 August 2013 (UTC).

@Vogone: PiRSquared17 (talk) 20:23, 6 August 2013 (UTC)


Thank you very much for all the translations, this is great! Happy to think further with you in 2014. Greetings from Dresden, Conny (talk) 13:54, 19 December 2013 (UTC).