[pmwiki-users] Great pagelist question...

[Dominique Faure]

On 8/30/06, Patrick R. Michaud <pmichaud at pobox.com> wrote:
> On Wed, Aug 30, 2006 at 06:17:07PM +0200, Dominique Faure wrote:
> >
> > Practically speaking, how do you want the PageVariableExtensions
> > recipe to be integrated into this scheme?
>
> Just as a general markup comment (and not to be taken as a criticism),
> somehow the {$Name:function} syntax just doesn't work for me.  In many
> ways I'd prefer us to be able to do something like function({$Name})
> instead.
>
> In particular, I'd really like to be able to write things like
>
>     url(Attach:somefile.jpg)
>
> which gives me back the url for somefile.jpg, instead of the
> <a href='.../pmwiki/uploads/somefile.jpg'>Attach:somefile.jpg</a>
> that a simple "Attach:" returns.
>
> The syntax being used by PageVariableExtensions (PVE) doesn't really
> give us a way to do functional things with other markups.  What this
> really says to me is that the function syntax of PVE may be too
> tightly coupled to page variables, and that we might want some more
> orthogonality between them.
>
> That said, given that I don't have a functional syntax that I *do*
> like yet, I can't complain too loudly about PVE's current syntax.
> I can say that I don't think PmWiki will introduce anything in the
> core that will cause PVE to stop working for existing markups, but
> we may add some features with page variables (or page-variable-like-things)
> that PVE won't be able to immediately take advantage of without some
> sort of update.
>

Well, I should confess that designing PVE as chains of postfixed calls
was far easier than having to deal with imbricated function calls that
we might have with a "classic" functional syntax... I've some ideas
about this, but I should still keep them mature before I could have
something working.

> Creating markup is easy, designing good and extensible markup is
> hard.  :-)
>

This is indeed (very) often a problem of specification, or in other
word, this come from the specification of the problem ;-)

Dom