Comments on: Yahoo and Google take different evil baby steps

By: adam

adam — Sat, 19 Feb 2005 08:42:28 +0000

Assuming you’re referring to rewriting outbound links, Google does not do this. A link on Google’s results page is a direct link to the site, not a link through a redirector page.

Unless they’re capturing the results with a javascript hack, which I guess is possible, Google does not seem to be tracking outbound link clicks.

By: Jeremy Zawodny

Jeremy Zawodny — Sat, 19 Feb 2005 06:50:10 +0000

Google and Yahoo both do this as of the last time I looked.

By: adam

adam — Thu, 17 Feb 2005 21:12:30 +0000

Maybe. In the case of Yahoo, I’d like to see a little bit of public discussion about whether they’re keeping this data, for how long, whether it’s being correlated to other data they’re collecting, what’s it’s being used for. Not accepting cookies doesn’t mean you can’t be tracked. Using an anonymizer helps, but that’s still a fair pain to use. And I’m sometimes compelled by third parties to use Yahoo services (like groups) that require logging in. From a social and business standpoint, it’s bad form for me to say “Sorry, I can’t use Yahoo groups” – should it be all or nothing? I’m not saying that Yahoo shouldn’t be tracking outbound links – but how about a little transparency here?

For Google – I don’t know, it’s a sticky question. I think that the end user should be able to format content as they like. But if Google’s going to be blithely inserting their own content into random web pages, I think that’s a problem. What happens if the address is already a link to Mapquest or Yahoo Maps? What happens if it’s a link to Amazon or some other third party that results in an affiliate payment, and the Google substitution mangles it? I think we should be careful, as end users, with granting third parties the ability to make unmonitored modifications to web content.

Maybe it’s better for the end user to have all sorts of stuff automatically linked to Google, but maybe it’s just better for Google. I can see both sides.

By: Emmanuel Pirsch

Emmanuel Pirsch — Thu, 17 Feb 2005 20:51:20 +0000

I don’t really see any of this as evil baby steps… Using a redirect for search results give the search engine a way to make subsequent search more relevent. If I search for XYZ and two site are found : ABC and DEF. If a majority of people click on the DEF link after searching for XYZ, then the DEF site is probably more relevent than ABC… I know that this has a potential for abuse but then you can disable cookies so the search engine will not be able to track you. BTW, you are already giving up privacy by using a search engine. They can already track your search query… Tracking in addition the exact link your interested will only give them some refinement.

As for the creating links to google maps for something that looks like an address… I think it’s a good idea… It helps bring more information to the web user.

It would be a good idea for a site operator to be able to say that it don’t want anything like this to happen… Like when they provide robot.txt files so their site is not indexed. Let’s call this file semantics.txt. This file could describes, like robot.txt, rules for allowing or disallowing semantics interpretation for specified pages. This way, the googlebar could read the file and see if it is allowed to add links for something that looks like an address.