On 12/12/06, Marc Riddell <michaeldavid86(a)comcast.net> wrote: What do you mean by "diluted"? If you can give examples, and maybe even some tentative solutions, we would be more than happy to listen to your bitching and moaning.[1] Steve [1] Well, I would anyway. I think our categories fall well short of what they can and should do.

On 12/12/06, Marc Riddell <michaeldavid86(a)comcast.net> wrote: And yes, this makes sense. We've had other writers pulling the same nonsense in other topics in the past: fictional vampires listed as fictional vampires but not mythological creatures, articles regarding aspects and criticism of the Koran being included in the Koran category but with editors insisting that including it in the "Islam" category as well being redundant, various grades of conspiracy theory put into one group but not another, and so on. A square is a rectangle, is also a regular polyhedron, is also a polyhedron, and so on, and each of these categories applies in its own way. I think the same thing is true for categories, the more categories something is listed under (as long as they are relevant), the better and redundancy (the idea that one category's listings are also completely encompassed by another) is not a bad thing, it only indicates a tighter degree of search. Parker

On 12/12/06, The Cunctator <cunctator(a)gmail.com> wrote: No, articles should belong to whichever categories they reasonably fit into. If an article fits into Mythology, and Greek Mythology, and Pre-Homeric Greek Mythology, and Athenian Mythology all at once, then it should be in ALL of those categories, because each is a different search. Yes, Mythology might encompass all of Greek Mythology, but So What? Someone looking for various mythological figures might grab Mythology first, and we're well served by having them do so and have an easier time migrating to the more specific area from there. Parker

The Cunctator wrote: Why? I'm not following your reasoning here. It seems to me a good point has been made - if you want all suicides, then having to traverse all subcats is tedious, to say the least.

Steve, But thrashed out to what conclusion? Marc

I wrote: By which I mean, the argument usually boils down to trying to decide precisely whether category membership is supposed to denote an "is-a", "has-a" or "is-related-to" relationship. But that can't be answered, so the arguments can never really be resolved, and people have to fall back to using categories not to implement rigid OO-like inheritable hierarchies, but rather, looser collections where the only semantic attached to category membership is "is-kinda-related-to". Some categories and the editors who maintain them will reach some kind of loose, relatively informal consensus that they're categories which embody a web of loose relationships (e.g. "Topics relating to Paris"), and some will similarly reach some kind of loose, relatively informal consensus that they're categories which embody a tighter taxonomy (e.g. "Counties in California"). Now, despite what I said about the problem not being solveable without additional and potentially more-complicated technical mechanisms which aren't likely to happen soon, it seems to me that one loose, relatively informal, "soft" solution to the part of the problem would be to try to reflect a category's semantic in its name, e.g. "Category:Counties in California" and "Category:Arrondissements in Paris" and "Category:Topics relating to Paris", rather than just "Category:California" and "Category:Paris". And we're probably doing a lot of that today. But there are still (and will always be) lots of problems when categories contain other categories, and we'll always be wondering whether category membership is or isn't or should or shouldn't be transitive, and it's these larger-scale questions which we can't (under the current architecture) ever fully satisfactorily resolve.

The important thing is to campaign for what you want -- an easy way to list all of the people who have committed suicide, say rather than for a particular implementation. On 12/12/06, Marc Riddell <michaeldavid86(a)comcast.net> wrote:

On 12/13/06, Steve Summit <scs(a)eskimo.com> wrote: Well, yeah, unfortunately this is basically true. We should have a well-defined category structure, or use really loose and flexible keywords...we sort of have the worst of both worlds here. The downside is it's much easier to guess "Paris" as a category name than "Topics relating to Paris". What would actually be good would be stronger, semi-automated systems for funnelling people's guesses into the right names. There's nothing wrong with someone guessing "Paris" if a bot can then channel that into the correct name - or at least highlight it for human recategorisation. Just like we don't "object" to people adding {{stub}} - we have people that come along and do the stage two recategorisation. Wikis work very well when you tolerate this kind of behaviour - don't ask for perfection up front, but instead appreciate every micro-improvement. Well we've discussed this at length, and I think we can go some way towards resolving it through either naming conventions or some other way of formally describing categories. Good rules might be "This category should not directly contain pages.", "This category is a geographical hierarchy.", "This category is thematic" etc. Steve

On 12/13/06, Marc Riddell <michaeldavid86(a)comcast.net> wrote: I don't think there are any simple dichotomies in this. Different kinds of stuff have different kinds of categorisational needs. We have a current trend towards splitting big categories. We prefer a single category "Dutch authors" rather than adding two massive categories to an article: "Dutch people" and "Authors". However this approach is pretty clearly a dead-end: how many times can we split? What do we split on next, "Dead Dutch authors"? Good category intersections would really help that one... I guess you're right in that people are just made up of attributes: where/when they're born/died, what field they were primarily in etc, all of which could be much more complex trees. Steve

On 12/13/06, cohesion <cohesion(a)sleepyhead.org> wrote: Yeah, I know about it. The widespread assumption is that Semantic MediaWiki will never happen at Wikipedia. Which is a pity, and possibly a self-fulfilling prophesy. Is there some small part of the semantic stuff that we could adopt, without taking the whole thing? Steve

On 12/13/06, Marc Riddell <michaeldavid86(a)comcast.net> wrote: It didn't evolve into an unfocused mess -- it started as an unfocused mess. As did Wikipedia. We don't necessarily need a small committee of experts to "fix" the "problems" of Wikipedia. One of the things that makes Wikipedia great is its flexibility. Don't.

Chaos? On 12/13/06, Marc Riddell <michaeldavid86(a)comcast.net> wrote:

David, I read it. A crowd can never come to a decision about anything; they can only make enough noise to call attention to the fact that a decision needs to be made. At some point, very early in its history, there was no Wikipedia crowd. How, and by whom, were the decisions made that evolved into what Wikipedia is now? Marc

"For what do you hunger, Lord?" Moneo ventured.         "For a humankind which can make truly long-term decisions.  Do you know the key to that ability, Moneo?"         "You have said it many times, Lord.   It is the ability to change your mind." Frank Herbert - "God Emperor of Dune" (p.55) I have read everything that has been written here, as well as the links detailing Wikipedia's history on the subject. A lot of thought - some excellent ideas. But change can happen only if enough people believe it needs to. I believe the Wiki Category system needs some serious study & re-evaluation. Does anyone else believe it's worth a try? Marc

The Cunctator wrote: I disagree. The answer is "it depends", and what it depends on is utility in navigation. There are a number of cases where people have managed to do this and it was the right decision. But, in every case you will have to win a small war over it. In general if all of the non-sub-category members of a category can be placed in a sub-category then they should, and you should not try and do both. I think the special case of putting an article in both is when there is a compelling reason for navigation purposes. While not common right now because of the way we do it and because few people have put time into making the Category system an interface to the site, this could become common that someone would use the category tree to "browse". In such a case they should be able to find articles with basic knowledge of the subject. Take for example [[Chess]], did you know it is in the [[Chaturanga|Chaturanga game family]] and therefor [[Category:Chaturanga game family]]. This category being on the article us useful for people already at the article, they learn something new. But for the person who is navigating the category tree this is almost with out value unless he happens to know that the chaturanga game family is and that Chess is a part of it. The person using the Category system (our veriosn of an index) will get as far as [[Category:Abstract strategy games]] and not know where to go. To solve this problem [[Category: Chess]] is a subcategory of both of these even though one is a subcategory of the other. And the article [[Chess]] has all three (though I am inclined to say that [[Category:Abstract strategy games]] should be removed form the article itself. There is another great example I think involving articles n the individual members of the Beatles and weather they should be included in a category as being members of the Beatles and also as being musicians. Somethign I wrote about this when I first became a wikipedian comes to ming so I dug it up. It was in the form of a proposal (I do not expect it to be accepted which is why I never formalized it) but I think it is worth thinking about on this issue. This is from: http://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Categorization&… Proposed Guidelines I have run into this issue a few times in my very short stay (so far) on Wikipedia and did some thinking about the situations where I think an article should be in a category as well as a sub-category. I came up with two guidelines and wanted to see if other people had similar ideas or if I am smoking crack. This seems to be the best place to get opinions on this issue so here I go .... Qualitative vs. Quantitative If the articles in a category are fundamentally the same type of article (i.e. about people or places etc.) and the difference between the category and sub-category is fundamentally qualitative then the article should only be listed in the sub-category. This is the example from the main page: Queen Elizabeth should not be listed directly under People, but Queens of England might be a good place for her. To say it a different way Queen Elizabeth would not be put in "Queens of England" just because she was a better/more famous/more notable person. The other side of this is when the difference is primarily quantitative. The example than I ran into that is good here, is the category "Chess grandmasters" and "Chess players". A person DOES go from being 'just' a chess player to being a Chess grandmaster by being a better chess player, so the difference is primarily quantitative and the article should be listed in both. A good way to tell if you are in this situation is: If the sub-category were deleted and all its contents imported into the super-category would it significantly reduce the usefulness of the super category? This brings me to the second guideline. [However I now think this should be modified by the notion that if all the articles of a specific type, in a category can be put into subcategories they should and the super-category should be dropped from the articles. Using the same example is a set of acceptable and inclusive sub-categories (grandmasters, International masters, masters, ... , unrated) could be agreed on so that no people articles remained then that should be done and the category [[Chess: chess players]] should be restricted to non-people articles. (a good example of this is [[Category:Magnet schools]] which has the schools themselves in sub-cats and only has a few articles on the topic itself in the category.)] <SNIP OUT MY NAVIGATIONAL UTILITY SECTION AS ITS THE SAME AS WHAT I WROTE AT THE TOP> Summary: I hope this is the right place for this, and as a clueless newbie I hope I am not re-hashing something that has already been discussed into the ground and resolved. If so and there has been some succinct treatment of the issue somewhere a pointer on my talk page would be awesome, even though I put "Proposed guidelines" above I am really just asking since I keep running into this and I am a but of a meta data freak. Dalf | Talk 02:18, 21 Apr 2005 (UTC) SKL

Marc Riddell wrote: Hate to be a downer, but this ain't ever going to happen. Just when you think it's all figured out, a new group comes along and undoes everything, all the while declaiming about how awful your system is. Categories are one of those things that editors can do without having any writing ability, or even much subject-matter knowledge, so the pool of categorizers is quite large. Although we don't have much consensus on categories, the "no supercats" rule is pretty broadly followed. It would be less work to change MediaWiki, even if you're not a programmer. :-) Stan

On 12/13/06, Marc Riddell &lt;michaeldavid86(a)comcast.net&gt; wrote: Yes, but our "small group of dedicated, focused people" is busy implementing single signon. ;-) Although the above statement is meant to be tongue-in-cheek, the kernel of truth in it is that making the category system sane will require major technical changes, which will require nearly the full attention of our core developers. Since they're already backlogged on at least two high-priority tech upgrades, there's little to be gained by asking them to devote time to the Category Question just now. Unlike other respondents, I think it can and will be fixed, but I agree that *it will not be fixed soon.* Until then, I would otherwise recommend the CatScan tool on the Toolserver, but the Toolserver hasn't updated from en since June; a situation which certainly hasn't escaped the notice of the community of Toolserver users on en. So until a functioning toolserver is restored to our lives, I'm afraid there's no good way to solve this problem.

On 12/13/06, Marc Riddell &lt;michaeldavid86(a)comcast.net&gt; wrote: Hehe, this really is a good welcome to wikipedia-policy thread! :D Judson [[:en:User:Cohesion]]

On 12/12/06, Marc Riddell &lt;michaeldavid86(a)comcast.net&gt; wrote: Yes. This is actually a very simple thing to do programmatically. What you are calling dilution is the exact opposite. Maybe we should add a feature that traverses category trees automatically, but it also seems like a nice outside-of-Wikipedia javascripty feature someone could build pretty easily.

On 12/12/06, Marc Riddell &lt;michaeldavid86(a)comcast.net&gt; wrote: It pretty much defeats the purpose of having subcategories. By putting an article into a subcategory, you are effectively putting it into the parent category. Maybe we should figure out a way to make that more obvious.

On 12/12/06, Marc Riddell &lt;michaeldavid86(a)comcast.net&gt; wrote: The suggestion is that some kind of tool -- as yet, I believe unwritten -- could generate a list of all members of a category and its subcategories. I imagine this would suit your purpose. -- Sam

Sam Korn wrote: Perhaps an option on the category page which allows a choice of "expand" upwards (category tree) and/or downwards (articles)? -kc-

I don't know that what I am asking for is as complex as your example. Cancer limiting the number (and specificity) of the subcategories would help to keep the process within manageable limits. Marc

On 12/12/06, Eugene van der Pijll &lt;eugene(a)vanderpijll.nl&gt; wrote: This is a real problem, which an extension, semantic mediawiki, (http://meta.wikimedia.org/wiki/Semantic_MediaWiki) was made to solve. It's definitely worth taking a look at for situations like this. It allows you to do things like say, "Paris is a [[type:=city]] in [[located in::France]], with a population of [[population:=1,000,000,000]]. You can then do fun things like say, show me all the cities with populations over 20 etc. For this original example, Fred is a [[profession::writer]] who committed [[type of death::suicide]]. [[Category:People]] etc. You could then query all the writers who's type of death was suicide without having to have one million categories for every possible union. I hope one day we can (maybe cross your fingers) use some of this on the main wikipedia. It's usefulness as a replacement for a lot of categories is one of the biggest benefits. Judson [[:en:User:Cohesion]]

The Cunctator wrote: I don't know about the "translation", but some of us are old enough to remember when articles were written in natural language. Some task-oriented people become so focused on what they believe would be easy that they often forget that there's a real world out there where people don't speak markup. Ec

On 12/14/06, Earle Martin &lt;wikipedia(a)downlode.org&gt; wrote: Yes, my example is inline, but it of course works anywhere. Also, people may want to look at this thread on wikitech-l right now http://mail.wikipedia.org/pipermail/wikitech-l/2006-December/040402.html Someone is working on a way to get category intersections, which would be helpful probably. Judson [[:en:User:Cohesion]]

On 12/14/06, The Cunctator &lt;cunctator(a)gmail.com&gt; wrote: Agreed. I don't see a major benefit in attempting to write "semantics" *in* the text. This would be fine: Paris is a city in France with a population of 10,000,000. ... [[Category:...]] [[Type:=City]] [[Location::France]] [[Population:=10000000]] Perhaps rather than attempting to fuse that stuff with the text, try fusing it with an infobox, which is inherently a way of formatting raw data? Steve

On 12/18/06, James Hare &lt;messedrocker(a)gmail.com&gt; wrote: Who are you mocking here? Steve

Eugene van der Pijll wrote: I agree that this often happens, but in my opinion some of these sorts of things are bad uses of the category system and can be avoided (which I do in my own categorization). In particular, categories that are very clearly collections of things, like [[Category:Cities in France]], should *not* contain "related-to" subcategories. The article [[Paris]] should be in [[Category:Cities in France]], but [[Category:Paris]] should not be. -Mark

Steve Bennett wrote: But it's by far the most logical construct to make, and there is very little reason to keep the software free from any support of that construct (even if it is to the detriment of other, more exotic constructs).

Steve Bennett wrote: Kind of like how ls has a recursive feature, but it's off by default, and how find is recursive by default, but you can limit it? -- Alphax - http://en.wikipedia.org/wiki/User:Alphax Contributor to Wikipedia, the Free Encyclopedia "We make the internet not suck" - Jimbo Wales Public key: http://en.wikipedia.org/wiki/User:Alphax/OpenPGP

Another example that recently occurred. In the Article on [[Sophia Yakovlevna Parnok]], the Categories, 'Poets', 'Russian poets', & 'Jewish poets' were each included. An editor came along and deleted the 'Poets' Category, saying it was a 'duplicate category'. If I now clicked on that Main Category, 'Poets' would Sophia Yakovlevna Parnok, the poet be included in that list? Marc

Steve Bennett wrote: That sounds like a good argument for uprading the software. Ec