[CLUG-tech] Google Re-Indexing
roland at giesler.za.net
Mon Feb 13 21:40:30 SAST 2006
Jonathan Hitchcock wrote:
> Do you know anything about HTTP, Roland?
I do, but not as much as I could.
> When you go to Google, and they send you an HTML page, it
> gets rendered in your browser.
> When you click on a link on this page, do you think you're
> somehow magically asking Google to send you on to that page?
Their webserver (generically speaking) does know.
> And that if that page doesn't exist, Google will notice?
Yes, not due to HTTP per se, but due to the server that serves the page. If
I were google, I'd use the server logs than one reason, which is not really
the point here.
> When you click on links on pages, you're asking your browser
> to go and fetch that page. Google isn't going to give a damn.
I beg to differ slightly. When I click a link, the webserver that served
the page is informed of the link that I follow, which I suggested could be
used to re-index the page. Unless it's a very specific IE/IIS function, I
know from work I did with MCIS that the logs could give you a HTTP 404 error
on the server that served the link, if the browser got a 404 back from that
site. I may be wrong about this (memory failure?), but we specifically used
that feature to suggest corrections to sites for clients. I have not
checked whether this is in the HTTP specification, so if may be something M$
did in MCIS/IE at the time.
If I'm completely wrong about this, please shoot quickly and from the hip...
;-) I promised on won't bleed on your carpet!
More information about the Clug-tech