I have a suggestion: While our main page has a high pagerank, individual articles probably don't (depending on how well they're linked to). So, it would help to link to Special:Allpages from the main page (the static cached version is fine). That way, it would take at most 3 links from the main page to any article. Further, the second level pages should also be made static and updated once in a while, because crawlers don't request dynamic pages (I think). Arvind On Mon, Apr 12, 2004 at 03:49:37AM +0200, Erik Zachte wrote: -- Its all GNU to me

Erik Zachte schrieb: [...] Last year, I played a lot with google's pagerank and the way new pages find their way into the index. One of the monst important factors in the PageRank formula is freshness. If I copy a page from the source web site which is already in the google index, I have a certain chance to get ahead of if - for some time. After a while, the ratio of age will balance and other factors (the global pagerank of a site, and it's update frequency) will catch up. So the answer is time. Most pages won't stay ahead of wikipedia.org for long time - If the other comply to the license, they have to link to us, which is boosting wikipedia.org. So in the long term (and that's what wikipedia is certainly good at: deep breath) we gain from all the parasits (which appears to me as a POV term) By the way: Noone is searching for "wikipedia rembrand" if he doesn't know wikipedia yet. I use to use google as a full text search engine for wikipedia as long as we can't provide this yet by ourself. Our aim is to get to position 1 for the search term "rembrand" and all the other lemmata which is a project for ages. Under these condiditions we can surely afford to let other sites to provide the wikipedia content in a GFDL compliant way. Mathias -- nach uns der synflood.