Updating the local Referring List involves collecting indexes from different servers. Collecting these
indexes is like collecting data for the database. Although feasible, conventional methods such
as "robots" or selective collection might not be appropriate.
The new method used here
is to have the local server collect indexes from those remote sites,
that its local browsers have
made access to.
Since browsers do not do any collecting, the job must be done by a
special collector
program. A browser could invoke this collector program
when it is requesting a document.
The advantage of the above fetching method is that it does
not require significant Internet resources
when these indexes are updated.