Library clips

OpenSearch in a way is kind of similar to what Technorati tag search has done by incorporating results from not only the Technorati index but also aggregating results from Furl, del.icio.us and Flickr.

From the website:

“OpenSearch is a collection of technologies, all built on top of popular open standards, to allow content providers to publish their search results in a format suitable for syndication.”

Many sites have a search box where you can search the sites contents and view the results page. If this search mechanism is done according to an open system then the search results of this website can be displayed on other websites, just like RSS feed content from one site can be displayed on another site.

Now any website can enable search results of their content to be syndicated elsewhere by adding OpenSearch RSS to their code.

At the moment A9 works as an OpenSearch Aggregator, in that it displays search results from various websites.

There is a running list of content providers who have made their website search results compatible with OpenSearch.

For users: if you are not registered you can only add one website to search, otherwise you can add multiple sites.

So try searching A9 web plus PubMed at the same time.

This technology is amazing, I would love to search all my favourite websites at the same time, and see the results in individual columns

I wonder if it’s possible for people to hack or scrape sites to be included in OpenSearch.

Meta-searching

• The results are returned side by side and not in one list
• The actual indexes you are searching don’t have to be general search engines, they can be individual websites…many meta-search tools search several search engines similtaneously, but not many can search several websites similtaneously (you know what I mean)
• Google can do “site:” search for websites, but this is using the Google index…you can probably do multiple site searches in the one query but the results are in one collated page.
• A9 uses the native search mechanism which can index the site far better than a general search engine index process of visits from bots and spiders.
• Gigablast is an alternative, as you can make your own personal search engine by entering up to 200 sites to search, but then search results are collated
• The beauty of A9 OpenSearch is you can choose from the list of websites you want to include in your search at the time of searching

Here is a great article that explained it perfectly.

Also check out Micro Persuasion’s post.

On some previous posts I was commenting on searching within one RSS feed to generate a refined RSS feed.

I realised Feedster allows this on its Advanced Search…it enables you to search within just one feed and in turn generates a search RSS feed.

You can try do this on the main search screen by using the site prefix
eg. site:libraryclips.blogsome.com, but you can’t add a search term to this query.

Also I noticed on the search results screen, there is a TIP! that asks if you want to refine the search results to popular blog hosts or exclude these blog hosts

Eg. search for the term library

Here is the TIP!:

“Many of the results are from popular blog hosts.
Would you like to exclude them, search within them, or search only within news sources?”

So when I click on within them you will see the search string is:

library site:(livejournal.com | blogspot.com | deadjournal.com | xanga.com | msn.com | blog.com | aol.com | blogsome.com | myblogsite.com) just add a minus infront of the site prefix to exclude popular blog hosts.

So obviously you can set the string yourself to search just within one host if you like .
eg. cats site:(blogspot.com)

But if I do something like,
furl site:(libraryclips.blogsome.com), it won’t work.

So it seems that you can only search within a bloghost or domain at the general level and not within a particular blog within that host (unless you use the Advanced search - mind you this searches within the RSS feed URL and not the blog URL, so it will only work as long as the blog has a feed).

But as I mentioned at the start of this post, when using the site prefix, you can see all posts within an individual blog, but you just can’t add a search term to the mix…unless as just mentioned you use the advanced search.

I also noticed when you do a search in just news that your results suggest refining the search results to one provider, like Moreover.

Here are the field searching syntax, amongst the most comprehensive search capabilities I have ever seen…feed_id will be interesting to see in the future.
Notice it says on the site syntax you can only search to the sub-domain level, and not the 3rd level which would be the actual blog home page.

To recap:

You can do this:
site:libraryclips.blogsome.com

You can’t do this:
furl site:libraryclips.blogsome.com
(Unless you use advanced search: searches in the feed URL not the home page URL)

But you can do this:
furl site:blogsome.com

I wonder if there is a syntax coming to search within blog categories.
eg. subject:library

What ever search you do it always generates an RSS feed, that’s what we like!

ADDED 11/05/05: See 2nd comment for how to search within an individual feed.