Will google crawl CFM Dynamic pages?

[Dave_T]

Hi..

I am maintaining a web site that pulls articles from a database using coldfusion. Do you know if this articles will be indexed by google?

Thanks

[atadams]

I use Coldfusion and have had good success getting Google to crawl my pages. The only issues I have encountered are:

• don't use more than 2 URL variables
  
• don't use "id" as a URL variable name
  

[Dave_T]

atadams,

thanks for your respond. I only use 1 variable, but I do use the variable name as "TopicID". Would that consider "id"?

thanks in advance.

david

[atadams]

I don't think so, but only Google knows for sure.

[Marcia]

id= isn't good, but pages are in there with affiliateID= all right. Until they catch it, I guess.

[anoryu]

The company that I work for developes almost exclusivly in cold fusion. We have never had a problem with google or any other engine crawling or displaying any of our sites, and we have thousands of them out there. No worries...

[gilli]

Hi Dave_T

I've been using CF for a while & have been through all the ups & downs related to Google.

The most important thing is never append CFID and/or CFTOKEN values to the urls. These are the equivalent of sessionId in php and will put a stop to any crawling of your site.

I've got sites that use a url structure of www.domain.com/index.cfm?pageId=123,456 and these seem fine with Google, however I prefer to remove the? and go for more "logical directory" style urls like www.domain.com/index.cfm/section/about_us/page/contact_us.

To do this & get everything after the index.cfm to behave like url variables you need to do a little trickery. The variables cgi.path_info and/or cgi.script_name will, in this case, return index.cfm/section/about_us/page/contact_us. From here you can loop over this string and pull your variables into the URL scope.

Here is the script I use (place somewhere where it will be run early in every request):

<cfset qStr = cgi.path_info> <!--- or <cfset qStr = cgi.script_name> ---> 
<cfloop from="2" to="#listlen(qStr,'/')#" index="i" step="2"> 
<cfset "url.#listgetat(qStr,i,'/')#" = listgetat(qStr,i+1,'/')> 
</cfloop> 

After this script has run I would have url.page = contact_us and url.section = about_us.

I originally set this up when I was under the impression that search engines hated "?". From my observations of sites with and without this set up I think it makes bugger all difference in that regard, but it looks much better from a users point of view.

One other note... if & when you put this in place you must use absolute paths for all your links and images/css/js references because although your web server will understand that its running index.cfm in the root directory your browser will think its calling a file in the "contact_us" directory.

Oh and one more... this will work on apache but it may take a little tinkering with your apache config.

[Dave_T]

hi Gilli...

Thanks for your detailed respond... I will definetely try it.

David

[biggles]

don't use more than 2 URL variables

don't use "id" as a URL variable name

A client of mine has a ColdFusion site with a CMS which generates page URLs along these lines

[abc.com...]

Whilst pages are indexed by Google and have PR4 a link check shows Google recognises no internal links. Also pages can't be found by searching on their URL.

Would that be because of the multiple variables & use of "id"?