XSLT and search engines

sde · 02-12-2006, 12:28 AM

I was working with XSLT and wondering what the difference would be for search engine ranking between .xml and .html files.

Both pages could render the same to a user's browser, but would the engines favor a .html page? I can't think of many times I've got .xml pages in search results.

It seems like it would be easier for an engine to read through. Of course you could make it a .php page and set the content-type to text/xml, but I'm pretty sure they judge the content type by that and not the extension.

I'm just wondering if I were to launch a new site, if it would be smart to use xml/xsl or not.

DJMaze · 02-12-2006, 05:19 AM

There are not many websites which run on xml+xslt.
Even XML usergroups use plain html

, for example see http://www.sgml-ug.nl/

I only use XML for DocBook and DOM to generate other formats.

sde · 02-12-2006, 08:41 AM

seemed like a fun project to make a site purely xml/xslt with ajax calls for server stuff. i suppose you're banking on newer browsers though for everything to work right and would be taking a chance if hoping for good SEO.

DJMaze · 02-12-2006, 12:29 PM

Yeah it would be fun though to see what happens. Not only the server load but also the client side of things. After all the client has to process the XML+XSLT.

But my main question is: Why should you, we have XHTML+CSS now?

sde · 02-12-2006, 06:50 PM

to answer your question, .. because i can.

i've always learned a lot by finding a project to apply something i'm learning about to.

my question was about seo, .. not efficiency or good design. you're server load comment doesn't make sense to me though. what's the difference between serving an xhtml/css page vs xml/xslt to the server. nothing if the same amount of data is in the files.

teknomage1 · 02-12-2006, 07:28 PM

I think google, for instance, would go ahead and index it, but your rank would still be dependent upon the number of links pointing to you, so if people are scared to link to your .xml files, you MIGHT lose out.

Do a google search for 'gentoo' and the first hit is www.gentoo.org which is actually www.gentoo.org/index.xml .

sde · 02-12-2006, 08:56 PM

that's interesting, .. it's an html4 dtd when you view the source. i guess it's all xml if it's well formed enough.

you could get around the hesiation of people not wanting to click on a .xml file by just making it .php and forcing the header content type to text/xml.

it might be cool to launch 2 brand new similar domains with the same content except 1 xml and see what the comparison is between the 2 in the different search engines. any ranking takes time though and i don't have that kind of patients