Comments | justben: Apache and hardcore content negotiation

justben

Apache and hardcore content negotiation

Oct 02, 2006 23:00

Today at work, near the end of the day, I was spending a bit of brainpower figuring out how sehellenes.org should serve up its content. The site news lives in XML files. Because of the nature of the data, it's sensible to view it a particular as raw XML Atom entries or transform it into XHTML. I may even want to add a way to transform it into RSS ( Read more... )

tech, geek, web

Comments 9

elgecko October 3 2006, 04:44:59 UTC

This brings to mind a very simplistic question. Since when are URL's URI's?

justben October 3 2006, 04:58:14 UTC

http://www.w3.org/TR/uri-clarification/

elgecko October 3 2006, 05:02:32 UTC

Oy gevalt isimer. What a mess!

justben October 3 2006, 05:16:16 UTC

Nah, it's just technical terms. Well-defined terms for slightly different subjects. They just happen to sound similar to confuse the newbies :-)

(The comment has been removed)

justben October 3 2006, 12:34:30 UTC

It is, actually; it's just massive overkill for many applications. Certainly for sehellenes.org. For high-end content management systems, though, it could be quite reasonable.

(The comment has been removed)

justben October 3 2006, 12:52:14 UTC

Yeah, that kind of interaction is going to be a hard requirement for sehellenes.org pretty soon. Imagine: most of the people who will want to add events and site news and stuff aren't going to want to log in to the server and write an XML file from scratch. SIlly nontechies :-)

I'm experimenting with just storing the XML in the filesystem instead of in a SQL database. My plan is to have CGI scripts read and write that filesystem-based XML directly until I really need a database. We'll see how that works out...

Thread 6

streamweaver October 3 2006, 12:20:12 UTC

I work quiet a bit with XML data mostly through hooks to external systems we need data from. In most cases I have to develop an xml interface for that resoruce we can query or hook to. For the most part we've been going to AJAX to work with data there, but we do that from a reasonable backbone of permanent URIs so they tie together conceptual content instead of physical data returns.

I didn't get a chance to look at your links so no idea if that will help but GWT is a useful AJAX framework for us here and is worth a look if you take that route.

justben October 3 2006, 12:44:03 UTC

Thanks for the input! I'd love to chat with you about website architecture stuff sometime. I'm definitely looking forward to setting up some AJAXy goodness for sehellenes.org, but I want to get basic site architecture, content, and layut stable before breaking into that experiment. I'll definitely check out GWT when that time comes, though.

In your XML serving, do you ever use HTTP content negotiation to serve up different data representations to different applications for a single conceptual data URI?