There are a number of communities in Japan whose members publish their
diaries on web.  In order to read the other person's diaries (and
sometimes to reply to them) in time, several programs called `antenna'
are created.

Antenna is basically a web page update detection software.  Based on
URLs listed in its configuration file, it periodically contacts web
servers, issues HEAD or GET requests, and produces web pages which
list sorted URLs with their modification times.

Several extensions are implemented to antennas.

In order to avoid overloading web servers, antennas share the metadata
they acquired.  By contacting another antenna that has many URLs in
common and getting metadata from it, an antenna can omit some of the
direct retrievals from web servers.

Some kind of `push' mechanism is also implemented.  Diaries are
usually written in a markup language specifically designed for
diaries.  These files are converted either statically (by a converter
on author's machine) or dynamically (by a program running as a part of
the web server) to HTML files.  These programs can recognize the
update of diary files, and notify the change to antennas.

At least six implementations of antenna software are used publicly (on
dozens of systems).  These implementations are mostly homegrown and
English documents (all the more papers) are rare.  I am thinking of
surveying them as they can be regarded as metadata scrapers and may
provide some clue for Roma extensions.