Desktop bot

Desktop bots are small applications that run in the background on personal computers, performing a basic indexing or remote retrieval function.
Taxonomy of desktop bots
Desktop Bots are small applications that act as a background service to perform a utility such as indexing, media distribution, geolocation, or other functions. They run as a background process intended to consume relatively little processing power.
Search engine bots
The growth of content on desktop computers has resulted in a number of background software agents known as bots that crawl and index desktop content in a manner that resembles Internet bots.
* Operating systems often include search agents, such as Apple's Spotlight tool or Microsoft's search system. Apple also assimilates searches across many online services through their Sherlock product.
* More recently, search companies such as Google have implemented their own bots that run in the background on personal computers, indexing content and merging personal data with online information. While initially available only for Apple Computer and Microsoft operating systems, Google announced Linux availability in June, 2007.
* Yahoo offers a desktop bot in conjunction with X1. And Ask Jeeves bought desktop search firm Tukaroo in June, 2004 to complete their offering.
* But not all desktop bots work in concert with online tools. Some, such as Copernic, offer a desktop-only search model as well as a combined desktop/online search.
Geolocation bots
Companies such as Plazes use geographic hints to determine the location of a computer system and publish positional information. While these bots may have access to GPS information, they may also combine observed networking information such as the MAC address and public IP on their LAN to determine position using a geolocation tool such as Maxmind or Quova.
While there are a wide range of geolocation tools, they have so far enjoyed limited adoption by consumers. Some possible reasons for limited adoption include privacy concerns, a difficulty reaching the critical user mass needed for a network effect, difficult interface, and lack of pervasive connectivity regardless of location. [http://tecfa.unige.ch/~nova/img/geoware_nova.pdf]
Media bots
Desktop bots that index content and stream it to external destinations are now appearing to make locally stored media accessible on remote desktops or portable devices. Such bots work in concert with a service of some sort to authenticate remote users.
Some bots (such as Webot's Mediabot) publish all forms of media, while others target specific applications (for example, Nutsie, affiliated with Sony/BMG and several phone carriers, streams iTunes to cellphones.)
Remote file bots
A variant on media bots is a system for remotely accessing files. While some agents such as GoToMyPC are aimed at complete remote control of a computer -- and only allow a single operator to control the machine at a time -- more lightweight bots such as SoonR are targeted at remote file retrieval. In some cases, a media bot may also serve as a remote file system -- Avvenu, for example, shares files but also streams iTunes media to cellphones.
uPnP extensions
While today's desktop bots are primarily focused on content and services within the machine for which they are installed, several vendors have indicated their intent to aggregate and distribute locally attached data from other sources such as upnp feeds on home media systems. In this respect they may perform some of the functions of a Slingbox device.
Privacy Concerns
Since many of the recent desktop search bots blend locally stored content with online content, privacy groups such as the Electronic Frontier Foundation have raised concerns that consumers should be wary of such tools.
Vendors such as X1 offer their hybrid online/desktop bot model for free, but charge for the desktop-only model. With the free product, the company makes its money from online search and advertising revenues; with the paid product the price pays for increased privacy.
Remote file system and media bots allow access to a system, but do not behave as servers. Instead, the bot connects to a trusted service that acts as an authority and relays access requests between remote clients and server agents. While this approach offers a degree of security, it is primarily done to avoid turning the desktop machine into a true server and to traverse firewalls with private IP addresses or IPSEC VPNs[https://www.gotomypc.com/help4.tmpl#firewalls].
Bandwidth and performance
Real-time access to remote content will depend the type of media being retrieved and the quality of intervening networks.
* Retrieval from a bot over a cellular network (using the Edge protocol, for example) will require considerably more buffering to deal with jitter than data retrieved over a broadband network such as a Wifi LAN.
* Streaming data will require buffering in order to deal with packet loss and jitter, as well as sufficient excess bandwidth to recover from such losses.
* Since bots work in the background, they are generally not used for isochronous communication in which the sender and receiver must interact, so round-trip-time is less of a consideration. ()

Main Menu

See Also

Main Menu