Hacker News Reader

Comment by ajg1977 | Hacker News Reader

ajg1977

I'm curious as to whether you considered if this would have any effect on PageRank for content you host?

This type of behavior - using javascript to present the user with links different than the googlebot found - seems like the type of thing they frown on.

View on HN

Unless googlebot runs a javascript engine, I don't think it will ever be able to find out... unless a human over there reads this and adds this information to their PageRank database.

Google does this themselves for their search engine pages, in a way... if you have JavaScript running, there is an onmousedown pingback added for every link in the search results so they know what you click on. The link itself remains apparently unaffected. Check it out if you have Firebug/Web Inspector, each link has e.g. onmousedown="return clk(this.href,'','','res','1','','012394AFC')".

Google does execute javascript, and even some flash, but not on all its crawlers (which is very understandable; building a DOM for every page on the web is expensive, even for Google. I say "build a DOM" because abstract interpretation is still far off, specially for the web and its cruft of languages, standards and quirks.)

http://www.seomoz.org/ugc/new-reality-google-follows-links-i...

http://www.labnol.org/internet/search/googlebot-executes-jav...

This is interesting, but it leads me to ask, what can you expect a JavaScript-enabled web crawler do? Load the page into the DOM, wait a few seconds for the dust to settle, start firing click events on everything and recording what requests are generated? It would seem a crawler could cause some really undesirable behavior on your site if that were the case. Imagine click events that post comments, delete things, downvote them, or generate emails. Since they are hidden behind JavaScript a web developer might expect a crawler not to try them...

There have been a few discussions about this topic in the past, and in the end it boils down to the developer being smart about the way he does requests.

I'm quite sure (wish I could remember where I read it) the crawler only actually performs GETs (which makes total sense because GET was always meant for retrieving and listing data), so a smart developer would use other HTTP methods (PUT, POST, DELETE) for actions that have modify existing application data.

Reply on news.ycombinator.com