Search via a URL means this:
Search the full-text from a list of webpages via a URL they are listed on
eg. a user or general del.icio.us tag is a URL with a list of webpages…so the idea is to put the URL of the del.icio.us tag into the search engine, and it will search the full-text of every bookmark in that tag.
Now imagine this when OPML URL’s are thrown into the equation.
I could make an OPML with items such as text, HTML links, and RSS feeds…then I could enter this OPML URL into the search engine to search the full-text of the text in the actual outline, the full-text of the HTML links (webpages) in the outline, and the full-text of the RSS feeds in the outline (although it is better to search the HTML of the homepage as some RSS feeds are only excerpts and not full-text).
Now imagine if this OPML URL (Outline) also has other OPML URL’s as items…now you could also have an option to search everything in your outline plus the outlines of your OPML inclusions.
What I mean by this is to say to the search engine, search the full-text of the items in my outline (OPML URL), and if my outline has an OPML inclusion/s include or don’t include that in the search.
These OPML inclusions in your outline may have OPML inclusions of their own, and so on…you could say to the search engine, search the full-text of the items in my outline (OPML URL) and search all the OPML inclusions, plus all their inclusions
…or more basically put, please search the full-text of my OPML Directory.
What if we took a heap of OPML Directories and made one massive root OPML Directory…then we could search a directory of directories.
Just say I create an OPML URL, and have 2 OPML inclusions (inclusion A and inclusion B), then I click on inclusion A (this could be one of my other OPML’s or someone’s elses OPML URL that I’m pointing to).
So, just say it is someone elses OPML URL, this OPML URL may have an OPML inclusion that is the same as inclusion B on my OPML URL.
So, when we make an OPML URL, and point to other OPML URL’s, which point to other OPML URL’s, and so on, we are bound to come across the same OPML URL’s in our travels within our personal OPML Directories…so it acts like a directory that doesn’t have a strict taxonomy structure (it is allowed to have multiple occurences of an OPML) and multiple occurences of items with these OPML’s.
By an OPML Directory I mean that every item in your OPML URL is basically an OPML inclusion
…now if an OPML inclusion in your OPML Directory happens to be the type of OPML being used as an OPML Directory as well, then we are starting to see an intentional Directory of Directories.
We could think of it as a tree, with a root OPML, and branch OPML’s, then imagine a forest of trees.
This is starting to feel like the infiniteness of the mandelbrot set
…one of person A’s branch OPML’s could be pointing to one of person B’s branch OPML’s, or it could be pointing to person B’s root OPML, etc…
…person B’s branch OPML could within it be pointing (via an OPML inclusion) to person A’s branch OPML
It very much is a web that is created from the tree branches in the forest.