|
Saturday, July 16, 2005
Previous topic
|
Next topic
|
|
 Saturday, July 16, 2005
started 7/16/2005; 2:52:08 PM - last post 7/19/2005; 6:10:17 AM
|
|
Doc Searls - Saturday, July 16, 2005 
7/16/2005; 6:52:08 PM (reads: 7309, responses: 3)
|
|
He's It 
| | Since we're busy cleaning up one house Cliff Hickman designed, before selling it so we can move into another, I thought I'd pause in the midst to get Cliff into the tag search mill. |
| | Not only is Cliff an outstanding architect in stylistic tradition of Frank Lloyd Wright, but a veteran builder as well. And very much a hands-on dude. (He's over on the roof of the new place right now, talking to the roofer, who should begin shortly) |
Hmm
Wanted: objective evaluations of RSS search engines
| | So why doesn't any industry analyst compare them in detail? Why don't we see objective, reliable, authoritative, first-source studies comparing them? |
| | This question came up for me after Robert Scoble posted this and this. I thought the numbers he gave in the second post were wrong and misleading, but that we were all in a fog to some degree. The number of links yielded by a search is just one kind of result. How many sources were there for those links? How fast did the engine's index archive the sources? Was the search for a URL or for keywords? Was the URL for a whole site, a particular day or a particular post — and how do the different engines compare for each? How good was the keyword search? Could it search for word strings? How about searches for tags? Could the engine sort by authority, relevance or freshness? How did they define those, and what do they reveal about those factors in search results? How about memorizing multiple searches and making RSS feeds out of any search? How well do those feeds work? |
| | We get lots of junk about how different search brands market themselve, sell advertising or otherwise make (or plan to make) money, and differ in their emphases. But nothing objective, yet, about their actual performance. |
| | Where are the Gartners, Forresters and IDCs here? |
| | Sure, there's Alexa, but it only deals with site rank, and its source is only IE browsers on Windows. |
| | We need to do better than that. |
| | One last thought. Bloglines has already been bought by AskJeeves. Others have been calling for Technorati to be bought by Google or Yahoo. And Yahoo's beta efforts at RSS search have already been revealed. |
| | Here's something to keep in mind: approximately one hundred percent of this whole category, and all the inventions and innovations in it, have come from small independent developers. Not from Google, Microsoft or Yahoo. |
| | How much invention or innovation to you expect to come from those companies after they're no longer independent? Some, sure. But as much as we're getting now, even as we complain about how much it sometimes sucks? Don't bet on it. |
discuss
|
|
Steven Kaye - Re: Saturday, July 16, 2005
7/16/2005; 10:37:32 PM (reads: 551, responses: 1)
|
|
|
>Where are the Gartners, Forresters and IDCs here?
It's also worth noting that librarians have training in evaluating sources, couldn't hurt reaching out to Online, InfoToday, etc. and suggesting they do an article on the topic.
discuss
|
|
Doc Searls - Re: Saturday, July 16, 2005
7/17/2005; 5:00:22 AM (reads: 677, responses: 0)
|
|
|
Yes, I saw the same idea on a Librarian blog, and now I can't find it. Still, a good idea.
Not a bad student report challenge, either.
discuss
|
|
Jim Moore - Re: Saturday, July 16, 2005
7/19/2005; 10:10:17 AM (reads: 700, responses: 0)
|
|
|
Hi Doc, This is a terrific post. I agree whole heartedly that objective evaluation of search engines is needed.
I would go farther and suggest that there are other classes of open, public, infrastructure services that need to be evaluated, such as "pingers" like bl.gs and weblogs.com. Also, blog production software already gets a lot of comparisons, but not particularly comprehensively.
Overall I see a web world populated by "web superservices" in the hundreds and perhaps thousands--"large pieces loosely joined" and a need for evaluation of these services and also for other meta-tools for managing their discovery, evaluation, selection, subscription, linking together, testing, deployment and monitoring. You may find a post today of interest in this regard. http://blogs.law.harvard.edu/jim/2005/07/18#a991
Best, and thanks for your continuing perspective,
Jim
discuss
|
|
|
Copyright 2009 The Doc Searls Weblog
|
