US patent 8,086,604 is the one Apple used to temporarily get sales of Samsung's Galaxy Nexus blocked. It covers using a single UI to search multiple data sources, including both local and remote ones. Here is the first claim:

1. A method for locating information in a network using a computer, comprising:

  • receiving by the computer an inputted information descriptor from a user-input device;
  • providing said information descriptor received from the user-input device to a plurality of heuristic modules, wherein:
    • each heuristic module corresponds to a respective area of search and employs a different, predetermined heuristic algorithm corresponding to said respective area to search the area for information that corresponds to the received information descriptor, and the search areas include storage media accessible by the computer;
    • searching by the heuristic modules, based on the received information descriptor, the respective areas of search using the predetermined heuristic algorithms corresponding to each respective area of search;
    • providing at least one candidate item of information located by the heuristic modules as a result of said searching;
    • and displaying by the computer a representation of said candidate item of information on a display device.

It's a continuation of another patent application filed on Jan. 5, 2000. There must be tons of prior art from before that. One very obvious piece is Apple's own Sherlock, first released as part of Mac OS 8.5 on Oct. 17, 1998. What else is there?

  • Unified search is Siri just fyi.
    – ihtkwot
    Commented Oct 15, 2012 at 4:09

7 Answers 7


AltaVista released "AltaVista Discovery" in 1998. This was an application that ran on your local PC, and maintained a locally-stored index of your documents and emails.

It also provided a browser toolbar via which you could launch a search, specifying whether you wanted to search your local files or the web. Regardless of whether you performed a web search or local search, the results would be displayed in the one format within your browser.

Here's some quick links:


A forgotten hero in Desktop search


Palm OS had a system-wide search that would call each program's main function with an argument to do a search (typed in a text field in a search dialog). The call to main was on the stack of the search system without even fully launching each application. Each program could display the results that it chose based on its own search.

(Mykland, Robert. Palm OS Programming from the Groud Up. Osboure/McGraw-Hill: Berkeley, 2000, pgs. 313 to 321)


Would running linux executing the find command to search for a song across multiple data-sources (local disks, virtual file-systems, and a remote mounted NFS partition) count?

  • a user-input device (physical keyboard) providing an inputted information descriptor (null-terminated byte stream, aka a string representing the filename arg of the find command)
  • said information descriptor (string) is provided to a plurality of heuristic modules (various functions/file-system modules) wherein:
    • each module corresponds to a unique file-system (nfs, ext2, proc, ramdisk, etc) employing different predefined heuristic algorithms (various file-system kernel modules/function calls), including various storage media accessible by the computer.
    • search happens
    • at least one candidate item of information (matching files) located by the heuristic modules as a result of searching is returned
    • the result is then be printed out to your console
  • This is missing a “different, predetermined heuristic algorithm”: filesystems aren't heuristic. This is more like a search engine aggregator, which I'm pretty sure I used back before 2000. Commented Oct 13, 2012 at 14:32
  • Challenge Accepted. Commented Oct 15, 2012 at 17:40
  • 1
    I'm going to try and force the square peg to fit.. CDFS/NTFS/NFS/CIFS file-systems for searching files on local storage media (cdrom/sdcard/hdd), LAN-storage, and WAN-storage. ProcFS and DevFS for inspecting, modifying, copying, and searching data in block-devices, apps, and the kernel. GmailFS for indexed and searchable files, emails, and links to websites from your email account. Cascade FS for searching through recent programming activity (svn). And most importantly, YoutubeFS for finding your favorites videos on youtube. Commented Oct 15, 2012 at 21:22
  • Can you point to ones that existed prior to Jan. 5, 2000?
    – peastman
    Commented Oct 17, 2012 at 19:08
  • CDFS/ISO9660/NFS/NTFS/Proc/Dev and find have been around since before the mid 90s. But YoutubeFS/Cascade/GmailFS are recent! Commented Oct 19, 2012 at 1:46

Windows NT since at least Windows 2000 has had search system built in that includes a plugin system so new types of documents can be indexed.


It was not on by default and has an interface the looked like a web search results but it worked.


The patent doesn't specify a locally run application, just "a method..". So meta-search services on the web prior to 2000 would be prior art? They take a user input, and query multiple heuristic based search engines.

eg www.dogpile.com, metasearch.com etc etc

  • 1
    It must also search local (file) resources on the client computer, by my reading of the patent.
    – Ron J.
    Commented Oct 25, 2012 at 10:23

The PACSL: Philadelphia Area Consortium of Special Collections Libraries (http://pacscl.org/about) implemented something very similar starting in 1998:

From their web page:

  1. Finding Aids Project including EAD training, digitization of existing finding aids, and establishment of a unified search interface (1998-ongoing);
  2. the creation of a Collaborative Online Public Access Catalog for those member libraries without individual OPACs (1997-ongoing);

  3. its Initiative for the 1990s I and II cataloging and access projects (1991-96); the PACSCL Name Authority Project (1991-94);

  4. Museum Loan Program for auction and booksellers catalogs (1990-93);

The PACSL software allows through an unified interface to locate a wide variety of documents based on user-defined input.

They provide the user-input to a wide variety of heuristics (one per type of document and area of research)


NEPOMUK, From 2006-01-01

Networked environment for personal ontology-based management of unified knowledge

NEPOMUK is designed to search files, calendar, contacts, messages, ID3 tags from MP3's (music meta data), and EXIF-data from graphic images (photos).


See also: http://www.semanticdesktop.org/ontologies/

From the file specification part at http://www.semanticdesktop.org/ontologies/nfo/:

NEPOMUK File Ontology (NFO) intends to provide vocabulary to express information extracted from various sources. They include files, pieces of sotware and remote hosts.

Since 4.9.1, the KDE project uses Heuristics in their NEPOMUK implementation:


I think the next book also mentions NEPOMUK and its use of heuristics:


At least, some of the researches listed on the EU-page can tell all about prior art.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .