Paul Everitt wrote:
> Back when we originally tried to get the locator going, I tried to
> follow the format specified by the IAFA working group (Internet
> Anonymous FTP archive). It had several classes of records that could
> point to each other. Classes included organizations, people, software,
> and documents.

Note to other Locator-SIG readers: more info on IAFA and
ALIWEB is at http://www.nexor.com/public/aliweb/ . Of particular
interest is the description of the IAFA format at
http://www.nexor.com/public/aliweb/doc/iafa.txt .

This is an interesting structure; we could write helpful
registration CGIs which ask for, say, your personal information. The
CGI then generates an IAFA form containing your data, and says "Put
this file *here*, naming it *this*; I'll remember it and check it
later tonight. Then I'll check it every week."

However, will people be more likely to maintain the IAFA
files? Perhaps... we could make things easy by registering people, so
they'd get a response like "Hi, Andrew! Here are all the packages you
have registered: 1 2 3..." When editing a package, the CGI could grab
the current IAFA file, let you edit it, and generate it as before.
(Of course, if everything is done via CGIs, what's the point of making
the user keep a copy on their machine? And we're back to a
centralized registry...)

Another option: take something like Ultraseek, and let people
register pages with Python software on them; those pages would be
crawled and indexed. Then users could do an Ultraseek search of those
pages. Not very structured, but it might be easy to implement.


Andrew Kuchling
amk@...
http://people.magnet.com/%7Eamk/

_______________
LOCATOR-SIG: Discussions about a Python Locator for resource discovery

send messages to: locator-sig@...
administrivia to: locator-sig-request@...
_______________