Active
Project:
Links Package
Version:
5.x-1.7
Component:
Code: links_related.module
Priority:
Normal
Category:
Support request
Assigned:
Reporter:
Created:
10 Jul 2007 at 08:43 UTC
Updated:
8 Aug 2008 at 06:18 UTC
When some SE bot comes to my site, it scans a page and finds a link that looks like "http://www.mysite.com/links/goto/5/9/links_related" and not the real link like "http://www.some-other-site.com" . Is this a problem or SEs will recognize what it is about by following a link.... Will SEs still notice that mysite is linking to some-other-site?
If this has been mentioned somewhere already, sorry.
k.
Comments
Comment #1
syscrusher commentedIt should work the same for search engines as for your browser. The Links Package handles outbound link tallying by sending the browser to the "goto" page. The goto page never issues any content at all, only a redirection header to the real page (and it internally tallies one page visit for the link in the Drupal database).
Most search engines catalog the destination of redirected URLs, not the source, so it should be okay.
I have a pending feature request from someone who wants to be able to have the Links Package skip the redirection step, under site admin control. This would make the URLs "real" but would prevent Links from tallying the outbound page visit. So this will be implemented as a config option rather than just a forced change, to let the admin decide whether tallying or real-looking URLs are more important to their individual web site.
Kind regards,
Syscrusher
Comment #2
WiseTeck commentedAnother option will be to keep the normal href link and add JS onClick event that will do the tracking. (so in the end the user will be able to chose form backend one of the 3 variants: link redirection, js tracking, no tracking)
something like this:
In this example 534 is the link_id for website.com/ and 913 is the node_id
and the countOutgoingLink JS script will log the click
Comment #3
smoothstr@drupal.org commentedSome observations:
When you use Google, if you are logged in, then Google tracks your visits using link redirection.
They are doing it in quite a clever way though: the link itself is normal, but if you have javascript on, then when you click the link, it's rewritten as a redirect. That's the best of both worlds.
Secondly, rather than tracking all clicks. If you have enough visitors, then you can derive statistically sound results by sampling. That means that you don't have to have redirects on all links all the time, only on some, some of the time. By knowing the number of visitors you have to a page, what the page contains, and knowing what a random subset of them click on, you can derive results for all the visitors.
At this stage, I don't think that this is an option that's sensible for the Links Package because this is a lot more complicated than tracking all links, but it would mean that most visitors to the site see clean links but that you are still able to track. The problem with this would be managing the Drupal cache, so that you get a representative sample: you don't want to just track logged-in users because that might bias results.
Maybe keep this in mind for the future.
Comment #4
goosemoose commentedHas this option been implemented. I'm hosting a network of several hundred sites and many of them link to each other in the directory. I want normal links to each other to make sure they are getting the rank benefit.