Crawling constantly would overload the blog servers and make Google unpopular. This way they get access to the backend and can index and load links straight into the crawler as people post.
Should Google crawl every possible weblog constantly? Most of the popular blogs have in common that they update at least once a day or more. Google crawls those sites already more then once a day without problems, catching Zeitgeist.
Trackbacks can be used to provide realtime ratings of a blogger's ranking within the community and thus ratings of the pages they link to. With blogs, Google can harness a huge distributed realtime relevancy filter.
But Google has a big relevancy filter, PageRanking.
Your remarks make me think though. Google could use Pyra's Blogger for a dedicated search engine like Daypop, but with faster updates and perhaps better filters (although the PR in combination with keyword density and other factors does a good job). Those results can also be integrated in the normal engine.
But I'm wondering if they do this at once, or wait till Blogger has more then active 200.000 users. What do you think?
I'm puzzled. I can't see instant synergies. Let me explain why.
a. Google News
Dan Gillmor, who broke this story, mentioned in an update the possibility, that the weblog links can be used to improve Google News.
But Google doesn't need to buy Pyra for that. Google can spider any leading weblog they want. Yes, there was this problem of interlinked weblogs resulting in a high PR (PageRank) for certain logs, but Google fixed that problem by giving more value to outgoing links then incoming links. They don't need to buy Blogger for indexing of weblogs.
b. Portal
Another suggestion that has been made: Google is moving to a portal.
I refuse to believe that Google is getting megalomanic. Besides, we all know what happened to AltaVista.
c. Direct access
Jshare suggested Google bought Blogger to get direct access to blog data.
But crawling the 200.000 active Blogs doesn't cost much resources. It's only a few gig of data. Why bother to buy a whole firm for that?
d. Journal with ads
Mateub suggests that Google could make a magazine out of the blogs, complete with ads.
But they can do that already. Have a close look at news.google.com. Search for, hmm, Google
At the right side, there's enough space for ads.
Google could index just the weblogs, like Daypop, and make a new product out of it (without buying Pyra).
Whatever the reason is behind the buy, it will have a huge impact. The simple fact that one of the hottest internet companies buys Pyra's Blogger will make the product main stream in months.
Slashdot Mirror
I've put the Flash-animation of eval.google.com on a second server.
Dear Hoggy,
Crawling constantly would overload the blog servers and make Google unpopular. This way they get access to the backend and can index and load links straight into the crawler as people post.
Should Google crawl every possible weblog constantly? Most of the popular blogs have in common that they update at least once a day or more. Google crawls those sites already more then once a day without problems, catching Zeitgeist.
Trackbacks can be used to provide realtime ratings of a blogger's ranking within the community and thus ratings of the pages they link to. With blogs, Google can harness a huge distributed realtime relevancy filter.
But Google has a big relevancy filter, PageRanking.
Your remarks make me think though. Google could use Pyra's Blogger for a dedicated search engine like Daypop, but with faster updates and perhaps better filters (although the PR in combination with keyword density and other factors does a good job). Those results can also be integrated in the normal engine.
But I'm wondering if they do this at once, or wait till Blogger has more then active 200.000 users. What do you think?
a. Google News
Dan Gillmor, who broke this story, mentioned in an update the possibility, that the weblog links can be used to improve Google News.
But Google doesn't need to buy Pyra for that. Google can spider any leading weblog they want. Yes, there was this problem of interlinked weblogs resulting in a high PR (PageRank) for certain logs, but Google fixed that problem by giving more value to outgoing links then incoming links. They don't need to buy Blogger for indexing of weblogs.
b. Portal
Another suggestion that has been made: Google is moving to a portal.
I refuse to believe that Google is getting megalomanic. Besides, we all know what happened to AltaVista.
c. Direct access
Jshare suggested Google bought Blogger to get direct access to blog data.
But crawling the 200.000 active Blogs doesn't cost much resources. It's only a few gig of data. Why bother to buy a whole firm for that?
d. Journal with ads
Mateub suggests that Google could make a magazine out of the blogs, complete with ads.
But they can do that already. Have a close look at news.google.com. Search for, hmm, Google At the right side, there's enough space for ads. Google could index just the weblogs, like Daypop, and make a new product out of it (without buying Pyra).
Whatever the reason is behind the buy, it will have a huge impact. The simple fact that one of the hottest internet companies buys Pyra's Blogger will make the product main stream in months.
Henk van Ess editor of Voelspriet
TIP: Check Ovidiu Predescu site now and then. He started working at Google's on January 22 and writes about it in his ...weblog.