It was thus said that the Great Richard once stated:
In article <003f01c66d7b$a4ba0fe0$21fe54a6 at ibm23xhr06>,
"John Allain" <allain at panix.com> writes:
How I used to publish simple databases.
Yeah, but what's evolving on my web space is no longer just a simple
inventory list. I want to include interpretive histories, links to
online documentation and software, hardware serial numbers, photos,
etc. Its becoming a big pile of metadata and although I can still
manage it manually, I can also see that this will no longer be the
case in a little while.
There was a similar "do it yourself" set of advice when it came to
cataloging my books, but instead <http://www.librarything.com> was the
perfect answer to my problems. I don't want to write my own PHP/MySQL
collection/database management system. (Hell, if I didn't want to be
a database monkey in 1978 when learning RSTS/E on the PDP-11/70, I
sure as hell don't want to be a database monkey in 2006!)
If noone is using a content management system that they like and would
care to recommend, that's fine, but I thought I'd ask first.
I converted my website [1] from hand crafted HTML pages to XML, then use
XSLT to convert it to HTML. A typical fragment would look like:
<page filename="blah">
<title>The title of the page</title>
<publicationDate year="2000">March 3, 2000</publicationDate>
<keywords>sample, keywords, example</keywords>
<description>A sample page that to show the structure I use</description>
<about>
<p>Some verbiage in HTML format about the page.</p>
</about>
<body>
<p>The actual contents of the page, in HTML.</p>
</body>
</page>
Another fragment (an actual fragment from my photo gallery which generates
http://www.conman.org/people/spc/photos/top10/car.html):
<page photo="yes" id="Chevy Nova" filename="car">
<title>A Long Lost Chevy Nova</title>
<tag>Vroom vroom</tag>
<keywords>car, junk yard</keywords>
<description>“A moving car gathers no
moss.”</description>
<location>Junk Yard, Powerline Rd south of Copans, Pompano, Florida, United
States</location>
<img type="jpg" width="678" height="532" />
<thumb type="jpg" width="100" height="100" />
<publicationDate year="1991">1991</publicationDate>
<body>
<p>Okay, so it's technically not moss on the car, but this picture
does get several comments, mostly “How did you do that?”
and “That's not real, is it?” Yes, it is real. And that
is the real picture; it hasn't been doctored at all.</p>
<p>So, how did I get that picture? Well, I went with my friend <a
href="http://www.geocities.com/CapeCanaveral/Launchpad/4891/">…
to a local junk yard and happened to have my camera along. This car
was perched up on a rack, covered in ivy.</p>
</body>
</page>
The XSLT files not only convert everything to HTML, but also generate all
the navigation links as well. For the XML file, I just created whatever
tags I needed so I can include as much meta data as I want within each
<page> segment. Writing the XSLT was a bitch, but now that I have that
done, I don't have to worry about it (and the "look" of the site is
provided
by CSS anyway). Adding new contents is quite easy now.
-spc (Once the HTML files are generated, I then use rsync to upload only
thost pages that actually changed to the server)
[1]
http://www.conman.org/people/spc/