Prefetching
Our goal in implementing prefetching was to enable the client software
to fetch documents that it predicted will be accessed by the user in
the near future. It is also required that this prefetching does not
interfere with the normal operation of the browser from the user's
perspective, except in the area of response time, which it is our goal
to reduce. This problem was broken up into two separate parts: the
creation of the prefetching tables so the client knows what needs to
be prefetched and when, and the implementation of the code that
actually does the fetching without disrupting the normal operations of
the main (parent) client process.
Back |
Index |
Forward