Hi,
Some interesting ideas here!
It doesn't have to be strictly "Page of the day" ("Page of the browser
refresh" is fine for our purposes). It would just be leaves, not
categories, doesn't have to include every page of a certain document
type but rather a hand-picked selection and doesn't have to be
especially random.
I was thinking of a manually updated file of page URLs maintained by the
administrator with a bit of JavaScript(?) that picks a "random" URL from
the file and 'somehow' extracts the page title (for use as link text) and
the correct thumbnail filename (which should be easy as the primary image
filename is a subset of the page URL).
Regards,
James
P.S., You are aware that XSLT is Turing-complete aren't you ;-) Jesting
aside, thanks for your HOWTO pages, I refer to them frequently.
On Fri, 21 Sep 2007, [EMAIL PROTECTED] wrote:
Date: Fri, 21 Sep 2007 14:08:27 -0400
From: [EMAIL PROTECTED]
Reply-To: [email protected]
To: [email protected]
Subject: Re: Random page of the day script
"Page of the Day" implies daily update so the update function should
be scheduled once each day. Cache the pagelet. Delete the homepage
from cache so it updates to include the new material..
The selection should exclude the homepage. Only include leaves or
should category pages be included? Should editors set a flag on pages
that should be included? Maintain list of previously chosen pages to
prevent frequent repeats. How many items on recently used list?
Use XMAP and XSL to create list of possible pages. Add a random
number to each item. Use the page with the lowest number.
The main design decision is how to create the random numbers. A
completely XSL solution would be fun, but a custom transformer that
adds a "random" attribute to every element (excluding root) would be
simple.
solprovider
On 9/21/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
We want to provide a thumbnail on the home-page which represents and links
to a random page of the day under Lenya. The random pages don't
necessarily have to selected "live" (I was thinking of dumping the URLs
to a file periodically and making some sort of script to select one).
So, has anyone done something similar? What's the easiet way to do it?
Cheers,
James
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
--
"Hey Microsoft, where do you want to 'SCO' today?" :-)
[EMAIL PROTECTED]
SDF-EU Public Access UNIX System - http://sdf-eu.org
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
