next »

[Phaëton]

okay i have a funny idea..

make a site that has a bunch of links that makes google go in circles forever..
but be the same info, randomly replaced on pages... it will think there are infinite
new links to follow on your site to crawl, but you keep outputting the same words
with new links to follow which just run the same php script ... lol

how would the architecture look.... im still foggy on the whole setup of this would it be a
mod_rewrite in the htaccess file to call my php script, then grab the request_uri
..

i guess my question is how would i get to a line of code in php from no matter what
url is requested... i hope this makes sense... trying to spit out what _i want_ no matter
what link was requested on my site....

[perkiset]

Well first off, G is already on to you and they won't spend infinite time on your site. Others bots might, but G is pretty good at that stuff now.

But the way to do it is have all requests for any page (or even graphics, if you chose) get rewritten into a single PHP script, where you can build the page you want to. The caller will think it's /thisdir/thatdir/anotherdir/apage.HTML but it's just one script.

This is actually how my web framework works. All requests come into a single script and the "page" is only spit out at the very last moment, giving me the ability to do just about anything I want before returning the HTML.

Can't do it just this moment, but the Rewrite is not tough and Ill post an example shortly. 

<edit: typo>

[nutballs]

called a spider trap, and like perk says, they be on to you.

however...

that is only if the links do not persist. Google will follow a few links down, but will then backtrack to verify authenticity. Or at least that what it seems to me in my evil empire.
So if the page you create is stored and re-presented when asked for, then G is a little more likely to keep it, and all the sublinks as valid. obviously there are some other requirements as well, like content uniqueness for example.

[lamontagne]

don't change existing content.. append to existing content..
google doesn't like when you delete information.. they thrive on information... instead of replacing the information, just add more information...

[perkiset]

Very true - G likes wider and deeper, not shifting sand.

But for G to completely index a 200K page site, you'll need some inbound links (at a minimum) for G to think that it's worth the time to scrape.

[lamontagne]

Make every page no more than 3 clicks away from any other page .. Max 80 links per page internally. Split these up into good navigation, related links, in context link, and either sub category/category listings.

use good seo pagination , 1|2|3|4|5|6|Next-> style..

use inner content links (on descriptions for page/category listings)

The rest is just basic navigation stuff....There are a few other tricks but that should get someone started. A few really heavy links could get the site indexed very quickly... IF the conditions are right. There is one major part missing from the puzzle here, but I can't give it all away. Look at the serps.

If you need to go higher than the conditions above (there is a page limit) max out the links per page, not the distance of each page from one another (every page should always be no more than 3 clicks from any other, increase links per page to make this happen if you must). Put these new links at the top of the page to let google know they are new (common sense). The key here is that if you can decrease the distance google is more likely to arrive at it faster, even if you must break a very general rule of 80 links per page, to google it is fine because at this point you have established yourself). Build this slowly but at an exponential rate (starting at 2 pages on day 1 you will get there.)... link to a few pages per day each day for a week or two and you should have it indexed (at 17 days you'll have 262144.. formula link for this below)

http://www.wolframalpha.com/input/?i=sum+series+2%5En+from+n%3D1+to+n%3D17