[pmwiki-users] New Hierarchical Groups Recipe...

[Patrick R. Michaud]

On Fri, Feb 02, 2007 at 10:28:44AM +1100, Kathryn Andersen wrote:
> On Thu, Feb 01, 2007 at 04:21:00PM -0600, Patrick R. Michaud wrote:
> > On Wed, Jan 31, 2007 at 12:28:31PM -0500, The Editor wrote:
> > > On 1/31/07, Hans <design5 at softflow.co.uk> wrote:
> > > > Wednesday, January 31, 2007, 3:08:27 PM, The wrote:
> > But what does concern me a bit here is that people not find themselves
> > being misled as to what the recipe does and does not do.  It's been
> > my experience that when people think of "hierarchical groups",
> > what they're really expecting is to be able to write urls like:
> > 
> >     http://www.example.com/wiki/Kingdom/Main
> >     http://www.example.com/wiki/Kingdom/Animal/Main
> >     http://www.example.com/wiki/Kingdom/Animal/Canine/Main
> >     http://www.example.com/wiki/Kingdom/Plant/Main
> > 
> > As far as I can tell, this recipe doesn't really achieve this
> > goal (yet?), so calling it "hierarchical groups" feels a bit
> > like false advertising or something.
> 
> I agree.  And I love this recipe, so I'm not dissing the recipe, but
> since it doesn't provide a total full-blown "hierarchical groups"
> functionality, and probably won't, just a mostly-hg layer over the top
> of the existing stuff, it probably needs a different name.
> 
> A few suggestions:
> ...

FWIW, if we do come up with a good name for this method of
clustering groups, and there's sufficient demand, I'll be glad
to take over the recipe design and maintenance to make it 
better supported within PmWiki and future upgrades.

> (As an aside: I've found a problem with the $g1... variables and
> pagelists; they don't work.  I think I know what the problem is -- the
> variables are being set from the current pagename, they aren't taking
> a pagename and processing it -- but I won't be able to work on a fix
> until tonight or tomorrow)

You're correct, as written the $g1, $g2, etc. variables will
always acts as if someone had written {*$g1}, {*$g2}, etc.,
regardless of any other prefix that might or might not be present.

But I think we ought to be able to do much better than $g1, $g2, 
in the first place.  :-)

Pm