Browse Prior Art Database

A Smart Scheme For Bookmarking URLs Supporting Space and Ttime Optimizations Disclosure Number: IPCOM000031850D
Original Publication Date: 2004-Oct-14
Included in the Prior Art Database: 2004-Oct-14
Document File: 3 page(s) / 42K

Publishing Venue



A Smart Scheme for Bookmarking URLs Supporting Space and Time Optimizations

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 47% of the total text.

Page 1 of 3

A Smart Scheme For Bookmarking URLs Supporting Space and Ttime Optimizations

A URL contains information about the protocol and address information to locate and access a web page. The URL may also contain application state information to allow the application to display appropriate content. A bookmarked URL needs to store all of the information in the URL at the browser so that when the bookmarked URL is replayed, the correct page is displayed with the correct application state.

URLs which are generated by portal sites are usually quite long due to two reasons. One is that portal sites usually consist of lots of pages organized in deep hierarchies, so the URL often contains a long path identifying the point in the page hierarchy it is associated with. The second is that a portal page is composed of elements contributed by the portal server itself, as well as elements contributed by several applications (portlets) on the page. Thus the state information in the URL is usually large.

Devices such as mobile phones, owing to their smaller memories, also often impose size limitations on URLs. Thus, web servers in general and portal servers in particular have adopted several schemes for limiting URL size, while not sacrificing the bookmarkability of the URL, by retaining the ability to directly or indirectly retrieve all of the state associated with the URL when needed.

The idea discussed here improves on the existing approaches applied when URLs need to support bookmarking and the amount of state associated with the URL is large. Although these conditions are true in a portal environment, they may often be true in other, non-portal, web environments as well.

First, we discuss two common existing approaches which attempt to reduce the URL size while not sacrificing bookmarkability.

The first approach is to apply compression algorithms to compress the URL size as much as possible. Since the percentage by which the URL size is reduced cannot be predicted or controlled by the compression algorithm, it is possible that the compressed URL is still quite large in size. In particular, it cannot be guaranteed that any particular URL can be compressed enough that it fits within the size limitations imposed by a particular device.

A second approach is to store the entire URL on the server in a persistent store, and replace the long URL with a shorter one with a unique reference embedded within. When the URL is invoked, the unique reference may be used to locate the complete URL from the persistent store and display the appropriate page. However, this approach has the drawback that every long URL has to be stored persistently on the server, which adds both server memory and time overhead.

Obvious improvements are possible to both schemes above while still staying true to the essential idea of each. In this publication, we discuss a new idea which allows URL sizes to be bounded, while still preserving bookmarkability (i.e. it is possible to obtain