I'm working on a new site, a tool where users can set various parameters, then generate an output then continue to further mess around with the various parameters and further change the output. Stand alone the tool provides a certain level of interest, but it would be great for the user to save a page with the parameters and then send a link to a friend, or post it in a forum, then the two or more user's can actively discuss the results.
My intention is to use the jsfiddle.net pattern, where one fills in the fields (for jsfiddle one enters the js, html and css code) and then clicks a save button that in turn creates an unique url by appending a '/' + random generated string. This saves the page such that anyone can easily return to.
If this feature is used, it will then create many links pointing back to many pages on the same domain. What is the best way to handle this from an SEO perspective? Does one simply leave all pages alone and let Google sort it out? Or, does one take control and manage what Google can and can't crawl, index and follow?
Intuitively I would guess that there is very little value in Google indexing the user generated pages. New users should be going to a blank page, not one that has someone else's random inputs. Also, there will likely be a lot of duplication or near duplication, which again brings no value.
On the other hand, one wants to capture the link juice of the shared links. I'm thinking in this case that the canonical tag is the best solution. The tag will point all the user generated pages back to the blank start page.
Also, what about crawling? and what about storage and the life of the page?
As a side note: I went to jsfiddle to see what hints they might provide on how they handle this issue. They explicitly call the robots meta tag as "index" and "follow", there is no canonical tag. Oh! they also include the "keywords" meta tag (LOL).
