[pmwiki-users] Some more Catogory logistics(WAS: Re:onover-bundling)

[V.Krishn]

On Wednesday 09 March 2005 11:50, pmwiki-users-request at pmichaud.com wrote:
> OK, I had started to re-arrange the Cookbook by hand by clustering related
> entries. However, I realised there's a problem with this approach:
> maintaining it.
>
> Pro: A category-based list can be generated automatically, and it takes
> less effort to put a category on one page than it takes to compare the
> functionality of one page with that of many pages.
>
> Con: On second hand, my initial try (using pagelist) takes about 20s for
> two lists and 15s for Pm's Uploads page, while the much larger Cookbook
> page takes a mere 2s. It may just be too tough on the server if we make a
> whole lot of category pages.
>
> Con: Another issue I have with ontologies of any kind: they're artificial.
> The deeper these hierarchies go, the tougher to maintain them.
I think this is one of the Main reasons why most of the wikis have a feature 
called Orphan pages. This helps in generating pages that are not linked to 
any other page.
In the Category Logistics a small function could be included which checks 
every page for the occurence of [[!SomeWord]].... If it does not exist report 
it as Orphan. 
This function could be called when a special(None-Creatable) page is clicked.

Secondly, there can be another Special category page, which when clicked 
whould make a TREE like structure. Something which we get when using 
"pstree". Suggestion: "Category.All"  :)
.....Not sure.... I think "Category.Category" does that.
>
> I think I'll go back to my original idea (clustering).
>
> Comments?
>
> At 11:44 PM 3/8/2005, John Rankin wrote:
> >The first step in using categories is to develop a 'folksonomy' -- an
> >informal vocabulary suitable for categorising the pages in question.
> >These might be the names of the 'aggregator pages'.
>
> Cheers,
> Radu
> (www.monicsoft.net)

V.Krishn