http://www.readwriteweb.com/archives/wikipedia_is_a_mess_wikipedians_say_1_i...
Now, we have a lot of work to do, it's obviously encyclopedic and it would be hard to get really wrong.
What needs to be in place to make it possible to recruit newbies for the task of referencing things? (Alleviate the citation syntax problem. Make the results easily checkable by the experienced. Ban the use of Twinkle or similar semi-botlike mechanisms on the resulting edits, as nothing repels good-faith new users like instant reversion. What else?)
- d.
On 2 November 2011 21:07, David Gerard dgerard@gmail.com wrote:
What needs to be in place to make it possible to recruit newbies for the task of referencing things? (Alleviate the citation syntax problem.
I'm thinking that the problem here is inline references. An inline reference is one where you plonk the reference in the middle of the text <ref>lots of stuff</ref>. The problem with those is that they break up the flow of the text, making it very hard to maintain.
Perhaps the Wikipedia should have a bot to convert ALL inline references into "list defined references":
http://en.wikipedia.org/wiki/Wikipedia:LDR#List-defined_references
e.g.
Frobar is a beedub<ref>http://www.hello.com/</ref>and can duubdu<ref>Rank bad links by Mr. Somebody 2008</ref> and more stuff.
which is essentially unreadable, the bot would change this to:
Frobar is a beedub<ref name=ref1/>and can duubdu<ref name=ref2> and more stuff.
And at the end add:
==References==
<references> <ref name=ref1>http://www.hello.com/</ref> <ref name=ref2>Rank bad links by Mr. Somebody 2008</ref> </references>
In other words put and keep the actual references in the actual references section. It would certainly be a lot more readable; and I think that's a lot of the battle when editing. I think it's the inline references that are the main problem.
As in, you wouldn't force people to not use inline style to insert references, but it would just get automatically adjusted to the site-preferred format.
- d.
-Ian Woollard
On 2 November 2011 22:38, Ian Woollard ian.woollard@gmail.com wrote:
I'm thinking that the problem here is inline references. An inline reference is one where you plonk the reference in the middle of the text <ref>lots of stuff</ref>. The problem with those is that they break up the flow of the text, making it very hard to maintain.
Yep.
Perhaps the Wikipedia should have a bot to convert ALL inline references into "list defined references": http://en.wikipedia.org/wiki/Wikipedia:LDR#List-defined_references
I believe there was en:wp community resistance to this idea.
- d.
Good.
Do it then.
On 2 November 2011 22:42, David Gerard dgerard@gmail.com wrote:
On 2 November 2011 22:38, Ian Woollard ian.woollard@gmail.com wrote:
I'm thinking that the problem here is inline references. An inline reference is one where you plonk the reference in the middle of the text <ref>lots of stuff</ref>. The problem with those is that they break up
the
flow of the text, making it very hard to maintain.
Yep.
Perhaps the Wikipedia should have a bot to convert ALL inline references into "list defined references": http://en.wikipedia.org/wiki/Wikipedia:LDR#List-defined_references
I believe there was en:wp community resistance to this idea.
- d.
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: https://lists.wikimedia.org/mailman/listinfo/wikien-l
On 11/02/11 3:38 PM, Ian Woollard wrote:
I'm thinking that the problem here is inline references. An inline reference is one where you plonk the reference in the middle of the text <ref>lots of stuff</ref>. The problem with those is that they break up the flow of the text, making it very hard to maintain.
Inline references are a problem even for newbies wanting to make a simple correction. In a reference rich article the error may be easily visible in article space, but becomes difficult to find in edit space when one needs to wade through a lot of references.
Ray
On 3/11/2011 10:45 p.m., Ray Saintonge wrote:
On 11/02/11 3:38 PM, Ian Woollard wrote:
I'm thinking that the problem here is inline references. An inline reference is one where you plonk the reference in the middle of the text <ref>lots of stuff</ref>. The problem with those is that they break up the flow of the text, making it very hard to maintain.
Inline references are a problem even for newbies wanting to make a simple correction. In a reference rich article the error may be easily visible in article space, but becomes difficult to find in edit space when one needs to wade through a lot of references.
WikEd has syntax highlighting to make editing easier.
Alan
On 11/03/11 7:36 PM, Alan Liefting wrote:
On 3/11/2011 10:45 p.m., Ray Saintonge wrote:
On 11/02/11 3:38 PM, Ian Woollard wrote:
I'm thinking that the problem here is inline references. An inline reference is one where you plonk the reference in the middle of the text <ref>lots of stuff</ref>. The problem with those is that they break up the flow of the text, making it very hard to maintain.
Inline references are a problem even for newbies wanting to make a simple correction. In a reference rich article the error may be easily visible in article space, but becomes difficult to find in edit space when one needs to wade through a lot of references.
WikEd has syntax highlighting to make editing easier.
That just introduces another geekish piece of software. In a random page in need of help there is no link to this software, If I want to fix an obvious typo with one or two keystrokes I don't want to spend an hour tracking down and learning some new tool. That just makes the cure more onerous than leaving the error in place. Non-technical people are quickly turned off by the fairy world of additional software.
Ray
On 4 November 2011 07:07, Ray Saintonge saintonge@telus.net wrote:
On 11/03/11 7:36 PM, Alan Liefting wrote:
On 3/11/2011 10:45 p.m., Ray Saintonge wrote:
On 11/02/11 3:38 PM, Ian Woollard wrote:
I'm thinking that the problem here is inline references. An inline reference is one where you plonk the reference in the middle of the
text
<ref>lots of stuff</ref>. The problem with those is that they break up
the
flow of the text, making it very hard to maintain.
Inline references are a problem even for newbies wanting to make a simple correction. In a reference rich article the error may be easily visible in article space, but becomes difficult to find in edit space when one needs to wade through a lot of references.
WikEd has syntax highlighting to make editing easier.
That just introduces another geekish piece of software. In a random page in need of help there is no link to this software, If I want to fix an obvious typo with one or two keystrokes I don't want to spend an hour tracking down and learning some new tool. That just makes the cure more onerous than leaving the error in place. Non-technical people are quickly turned off by the fairy world of additional software.
I still think a bots the right answer. A bot that collects the references to the end of the article has no very major downsides I can find; the users can carry on doing more or less what they already do, and the bot just tidies up.
It could work something like, if a reference is named and there's a reference list which it's not in, then the bot moves it into the list and leaves a named tag that points to it, if it's unnamed or there's no list then leave it where it is.
Ray
On Wed, Nov 2, 2011 at 9:07 PM, David Gerard dgerard@gmail.com wrote:
http://www.readwriteweb.com/archives/wikipedia_is_a_mess_wikipedians_say_1_i...
Now, we have a lot of work to do, it's obviously encyclopedic and it would be hard to get really wrong.
What needs to be in place to make it possible to recruit newbies for the task of referencing things? (Alleviate the citation syntax problem. Make the results easily checkable by the experienced. Ban the use of Twinkle or similar semi-botlike mechanisms on the resulting edits, as nothing repels good-faith new users like instant reversion. What else?)
Responding more to the opinion piece published in the Signpost, than what you are saying, my experience of looking through such backlogs is large amounts of mis-labelling, or outdated labelling. Is it very discouraging to think you are working on a backlog to find that the article either never had the alleged problem, or that it was fixed but no-one bothered to remove the tag identifying the problem. So I think those numbers quoted in that opinion piece are worthless (i.e. over-inflated through poor tagging practices). Random sampling, tailored to specific areas, would give a better idea of the extent of any problems, IMO.
What I think often happens is that someone tagging stuff thinks: "there is a problem with this article, but rather than use the right tag, or look at the problem in any detail, I'll put a tag on to be safe and then move on". And then someone else, later, might fix the article during general editing without even looking at the tag, and not remove the tag, or might expect others to remove the tag (no, really, that is a common attitude among some people who prefer others to judge any remedial work they have done - you put the tag there, you should come back and assess whether it is still needed).
And this in general put me off tags. I've hardly ever tagged articles (preferring to fix them myself or point them out to someone who can fix them), and I tend to ignore tags on articles, preferring to form my own judgement over whether an article is reliable or not (i.e. why should I trust the judgement of a random Wikipedian over whether the article has problems, when all articles should be read with jaundiced eye towards potential problems?).
Carcharoth
On 3 November 2011 11:10, Carcharoth carcharothwp@googlemail.com wrote:
On Wed, Nov 2, 2011 at 9:07 PM, David Gerard dgerard@gmail.com wrote:
http://www.readwriteweb.com/archives/wikipedia_is_a_mess_wikipedians_say_1_i...
Now, we have a lot of work to do, it's obviously encyclopedic and it would be hard to get really wrong.
What needs to be in place to make it possible to recruit newbies for the task of referencing things? (Alleviate the citation syntax problem. Make the results easily checkable by the experienced. Ban the use of Twinkle or similar semi-botlike mechanisms on the resulting edits, as nothing repels good-faith new users like instant reversion. What else?)
Responding more to the opinion piece published in the Signpost, than what you are saying, my experience of looking through such backlogs is large amounts of mis-labelling, or outdated labelling. Is it very discouraging to think you are working on a backlog to find that the article either never had the alleged problem, or that it was fixed but no-one bothered to remove the tag identifying the problem. So I think those numbers quoted in that opinion piece are worthless (i.e. over-inflated through poor tagging practices). Random sampling, tailored to specific areas, would give a better idea of the extent of any problems, IMO.
My reaction was somewhat different. I went into the list of categories or
{{unreferenced}} tagging (by month) just to have a look. Well, it's pretty miscellaneous. I did a few, including some of my own articles (embarrassing, but except for one there was nothing that was really out of hand).
The normal reaction is to slice and dice. Doing it by oldest goes back five years, which is certainly not excellent; but the old ones didn't seem more worrying than others, really. How many are also tagged as orphans? This seems more likely to be where really mucky stuff might lurk. Articles of the type [[1853 in Canada]] are basically lists, and unreferenced lists are really another issue. Priorities seem clearer when you get involved. Small town in Slovakia: easy to check it exists.
The thing is that with a better classified backlog you'd get some easier progress. If you Google the topic of these older articles, you tend to get mirror material back, so I don't know that it is fair to ask newbies to sue their own unsupported initiative.
Charles
On Thu, Nov 3, 2011 at 5:02 PM, Charles Matthews charles.r.matthews@ntlworld.com wrote:
The thing is that with a better classified backlog you'd get some easier progress. If you Google the topic of these older articles, you tend to get mirror material back, so I don't know that it is fair to ask newbies to sue their own unsupported initiative.
Sue? Was that meant to be "use"? I agree, some backlogs are better dealt with by more experienced editors. How can such slicing and dicing be done? And if there were manageable chunks, I'd do bits as well.
Carcharoth
On 3 November 2011 17:56, Carcharoth carcharothwp@googlemail.com wrote:
On Thu, Nov 3, 2011 at 5:02 PM, Charles Matthews charles.r.matthews@ntlworld.com wrote:
The thing is that with a better classified backlog you'd get some easier progress. If you Google the topic of these older articles, you tend to
get
mirror material back, so I don't know that it is fair to ask newbies to
sue
their own unsupported initiative.
Sue? Was that meant to be "use"? I agree, some backlogs are better dealt with by more experienced editors. How can such slicing and dicing be done? And if there were manageable chunks, I'd do bits as well.
I think quite a lot could be done with Catscan 2.0, searching high-level categories for pages carrying the {{unreferenced}} template. Something like [[Category:Places]] or [[Category:American people stubs]] to some subcategory depth. It's a sophisticated gadget, and the toolserver is said to be sickly right now. But I imagine a determined operative could come up with useful listings that would be better for the purpose of chipping 1% off unreferenced articles.
Charles