Talk:Wikicite (2006 proposal)

From Meta, a Wikimedia project coordination wiki

number of things[edit]

I see a number of things that could need sorting out right now:

  • Remove the information from en: now that it has moved to meta:
  • Citations or bibliographic references can serve two purposes, and I think this should be pointed out:
    • 1) Show which sources the article text was based on. This is the kind of citations found in academic papers, but seldom in encyclopedias. The talk about "a fact is only as reliable as..." seems to indicate that this would be the goal of Wikicite. Is this correct? I think it would be hard to keep such lists of sources up to date, as different people edit the article.
    • 2) Help the interested reader to find additional reading on the subject. This is the kind of literature reference normally found in an encyclopedia, at least the ones I own. The total of referenced literature might be a small core of "canonical" university textbooks, or a indefinitely large number of books, close to the entire Library of Congress.
  • List a few Wikipedia articles that already have good citations, to indicate what we want to achieve.
  • Discuss why the work of adding citations is hard, and how it could be aided.
  • Do not start by saying "we need a database", because that shouldn't be the mission statement.

--LA2 02:20, 11 Feb 2005 (UTC)

Here is something, perhaps of interest: Open Access Bibliography, http://info.lib.uh.edu/cwb/oab.pdf This bibliography has been published as a printed book (ISBN 1-59407-670-7) by the Association of Research Libraries (ARL). http://www.arl.org/pubscat/pubs/openaccess/ ARL and the author have made the above PDF version of the bibliography freely available. It is licensed under the Creative Commons Attribution-NonCommercial License. --LA2 20:08, 5 Mar 2005 (UTC)

Wikicite has been mentioned positively in this very interesting article: http://www.fla.fi/frbr05/2005_Jarvenpaa_LeBoeufw.pdf

Format of an ISBN number[edit]

I see something in this subsection which appears to be an error (although I am not an expert on this subject). ISBNs only allow the character X in their last digit; all other digits only allow 0 to 9 as possibilites.

This means that the only base-eleven digit is the last one, and that all other digits in the ISBN are base-ten. S. Neuman 14:56, 30 August 2005 (UTC)[reply]

Status of proposal?[edit]

What is the current status of this proposal? Is it slated for implementation or still in the planning/discussion phase? Thanks for the information. Courtland 23:01, 2 January 2006 (UTC) P.S. I came here via a reference at http://en.wikipedia.org/w/index.php?title=Wikipedia:Village_pump_%28policy%29&oldid=33650659 (the "Titl(ing) books" section)[reply]

I have been updating this project with a new, more detailed functional design. Once this is finalized I will begin implementing the necessary enhancements to Mediawiki. At this point useability feedback and additional functional requirements would be very helpful. Would also appreciate help in publicizing this project among the community to get as much feedback as possible. Thanks.
Jleybov 18:54, 12 January 2006 (UTC)[reply]
I just want to be express my deep apprecation for someone actually going ahead and writing some code on this. I think it's a wonderful a idea, and I strongly look forward to using it, but before we can the code needs to be written. Thanks again! en:User:JesseW (not logged in) 07:54, 20 January 2006 (UTC)[reply]
Thanks:) And again, feedback would be very valuable at this point since I don't want to build something that cannot or will not be used because of major functional or usability flaws. Jleybov 22:27, 25 January 2006 (UTC)[reply]
The ability to overlap article text to be cited would be nice, for e.g. the birth and death dates at en:R._C._Sherriff/Sources: (<f38><f46><f31><f40>1896</f40></f31>-06-06</46></38> – <f38><f46><f31>1975</f31>-11-13</f46></f38>)
f40 is only for the birth date year, f46 is for the complete birth and death dates. -- Jeandré, 2006-01-22t15:59z
Interesting case. However, you are thinking somewhat backwards from the stand-point of the source rather than the fact. Born: 1896 and Died: 1975 are two distinct facts, and should each have their own citations. Thus both f40 and f46 attach to the birth year assertion, while only f46 attaches to the year of death assertion. The fact that fact-points can be somewhat sloppily used will I think help towards their adoption. Though I guess allowing such imprecision in the citation data from the start could be a problem in using it for other types of analysis later on... 22:27, 25 January 2006 (UTC)

This is fantastic and I also greatly appreciate someone being willing and able to actually put up the code. The most exciting part to me is the checking that the references are used correctly. If done right this can take us out of the current state of unreliability. I think it would need to note what editor did the checking, and probably record the oldid. There will probably have to be some way to disregard bad faith checks. That could either come from only allowing trusted editors and/or experts to do it in the first place or just from a system of tossing out bad ones. My final question is how will this integrate with Cite/Cite.php? That system seems to be gaining momentum as the one to standardize on. To get this one to work it will probably have to integrate with cite.php. - Taxman 17:03, 23 February 2006 (UTC)[reply]

Sorry for such a late response; this page doesn't get much traffic, so it took me a while to notice your comments. Regarding reference checking, this is intended to occur in hand with article validation and some form of stable version designation. So in this scheme, a particular revision of an article is nominated as a stable candidate, and editors then vote for or against its receiving stable designation. This project adds the idea of explicit fact-checking, but this will still occur in the context of a vote, and so each checker/voter's id is indeed recorded, and hopefully aggregate fact-check results will be enough to overcome a few bad faith checks without either limiting this function to experts, or having to use complicated "webs-of-trust".
Thanks for bringing up the Cite extension; it will probably make sense to extend its functionality instead of creating new citation mark-up. I will definitely look into this.
Jleybov 17:29, 28 April 2006 (UTC)[reply]

Toy Examples[edit]

The best way to get people to start citing (and then citing in more sophisticated ways) is to make a small cookbook of simple toy examples.

Immitating idioms like this is the best way to learn how to program or cite. I teach writing and inline citation plus bibliography entry (Harvard System) is so simple that people do it.

If you start making examples, other people, like me will be right on your heels, adding to those examples [Jonfernquest] 6:47 July 13

Architecture[edit]

As requested, here's my feedback.

You'll notice that I've normalized the actual sources out of the citations. This may overlap a bit with WikiTextrose, but it's a necessary change for a clean schema.

Here's how I would lay out the schema: (Arrows point from foreign keys to primary ones.)

Articles <--- Citations ---> Sources
                  ^             ^
                  |             |
            Citation Evals  Source Evals

A source would have to be registed in the sources system before it could be used to cite information in an article. When adding a citation, there would be the option of searching existing sources or adding a new one. By normalizing the source data, sources can be cleaned up and evaluated more easily.

Users could comment on the validity of a source or an individual citation. It's important to be able to comment on the source itself, not just the citation. Some sources are unreliable as a whole. Some sources are great but cited for invalid reasons. Furthermore, indictments and support of a source's reliability often apply to other places the source is used. Thus, clicking a citation would not only show source information, but could also show other places it's used and its evaluations, both citation and source level.

Tables:

  • Articles
    • Current fields
  • Citations
    • Article ID (foreign key)
    • Source ID (foreign key)
    • Location in source (using location type specified in source record)
  • Sources
    • Source ID
    • Various source fields (ISBN, URL, access time, etc.)
    • Location type (page numbers, paragraph numbers, etc.)
  • Citation Evaluations
    • Citation ID
    • Type (Accurate, Misleading, etc.)
    • Note
  • Source Evaluations
    • Source ID
    • Type (Good, Biased, etc.)
    • Note

--Davidstrauss 23:24, 9 September 2006 (UTC)[reply]

cite template for streaming media[edit]

Hi, can you tell me what specific cite template can be used to link to streaming media on en.wikipedia? Thanks Lostintherush 14:39, 25 September 2006 (UTC)[reply]

Related idea[edit]

I had a similar idea a while ago. Not very similar, but similar in concept.  :-) The bibliographic information would be kept in a database separate from articles (which could be edited collaboratively? or generated from a combination of many library entries, like isbndb.com does), and summoned whenever needed, like to auto-generate a References section. Facts would be cited with Wikipedia's original web citation format: [http://www.example.com/factsource.html] (Cite.php is horrible). Except it would use URNs instead of web-only URLs, so it could reference books [1], scientific articles <891:ZEBK>2.0.TX;2-B , and so on. So each reference would have a unique identifier that could be called from within the article text. My original idea had users searching for a reference and then copying and pasting the URN into the article source. I brought it up on Wikipedia and everyone hated it, but maybe it gives you some ideas for this project? — Omegatron 16:25, 4 November 2006 (UTC)[reply]

Questions[edit]

  1. This seems to be a replacement for Cite.php (which is a very good thing), but, if so, how does it handle footnotes? I'd say footnotes aren't really appropriate on a web encyclopedia; wiki is not paper and all that, but I've been consistently opposed on that.
  2. I'm not sure I understand the fact point concept. You would just type Columbus was most likely Genoese++fn,, or would you type the numbers, too: ++fn2,? I see both in the examples and I'm not sure I understand the relationship between numbered and unnumbered fns.
  3. How do you handle overlapping facts? It seems that the fact point concept divides everything up into serialized facts, but often a chunk of text will have some parts that are verified by only one place, some parts that are verified from multiple places, etc.
  4. Can you assign multiple references to a single fact? — Omegatron 16:37, 4 November 2006 (UTC)[reply]

Complexity[edit]

This seems like a user/consistency nightmare. I'd rather just stick with cite.php and some sort of article validation software. Voice-of-All 00:43, 28 April 2007 (UTC)[reply]

Wikidata[edit]

Does anyone know where Wikidata is being discussed currently? All those links appear to be untouched since June 2006, and the wikidata-l mailing list is utterly silent. Omegawiki is not an official Wikimedia project (? so why does OmegaWiki list all our sister projects at the bottom of its mainpage?), but Wikidata isn't even mentioned at the foundation site... Thanks for any help; and updates to relevant project pages would be even better :) Quiddity 19:25, 16 June 2007 (UTC)[reply]

Inactive?[edit]

Shall we just say that this page is inactive? Is there an inactive template here? Richard001 07:24, 6 November 2008 (UTC)[reply]

It only took 2+ years, but the project is now marked inactive. - Hydroxonium 17:53, 8 February 2011 (UTC)[reply]