Talk:Incremental search

Excremental

The link to "Excremental" redirects to "Feces." Can it be removed, or is there a better article to link to? Hac13 (talk) 05:49, 15 November 2008 (UTC)

Page rename

I disagree with calling this feature "Non-modal incremental find". It's quite modal, and the biggest drawback is that it's modal without being a dialog, leading to mode error. --User:Jesse Ruderman

I removed the "non-modal" bits as the name is too lengthy. --minghong 09:03, 20 Apr 2005 (UTC)

Origins

What are the origins of this type of feature? I remember hearing once that Emacsen pioneered this, but I'm not at all sure. --maru (talk) contribs 04:47, 3 April 2006 (UTC)

I'm flabbergasted that WordPerfect gets the credit for this. GNU Emacs has AFAIK always had this, although a quick search only confirms it was present in Emacs 15 (the NEWS for Emacs 16 lists an enhancement for low baud usage), which was released in 1985. I do not know if this was new with GNU Emacs or if Gosling Emacs had this as well. Kjetilho (talk) 12:41, 10 June 2009 (UTC)

I did some research and added some earlier history. I think the WordPerfect story could be less chatty and more encyclopedic, but I'll leave that for others to edit. Kjetilho (talk) 08:32, 11 June 2009 (UTC)

The first WordPerfect draft was 1979, the first full release was launched in late 1982, I was using it myself in the spring of 1983. — Preceding unsigned comment added by 2.123.173.109 (talk) 08:47, 29 May 2018 (UTC)

EMACS had this in the 1970s; it was called "incremental search". It was modal: you type ctrl-S to begin, the prompt says "I-Search:" and you start typing. As you type (or delete), it scrolls through the buffer to the matching string; you exit by typing ESC or any editing command. This version of EMACS was written in TECO on the PDP-10, not sure who created this particular feature. The (much fancier) version of emacs on the Lisp Machine also had this, about 1977. I think EMACS may have coined the term "incremental search", by the way. —Preceding unsigned comment added by 173.13.229.185 (talk) 19:10, 8 September 2010 (UTC)

It's far older than that, and wider - you're talking of instances, this is an entire class of processing.

The parent class is bubble searching/sorting, and the most notable example of that is at the heart of most multitasking operating systems: every action bubbles up through the list of routines asking each one "are you my Daddy?" until either it finds the one which can handle it, or it pops out of the end of the list as an error (which are usually let go of). Incremental searching is a similar subclass, where a user's query bubbles through the list of all data collecting a class of possibles, from which the user either selects one or sub-classes their search further, filtering the list they just came up with iteratively until they learn enough to meet their need, with the options of reverting if they've gone thwe wrong way, or abandonning the search in bafflement!

A random example of that might be if they've seen a barrel helm and wanted to know more about it. A first search would probably be on "helmet", then incremented with "old", and so on and on until they're satisfied.

Text searching in the way you describe is an instantiated subclass of true Incremental searching, applying it to the specific instance of text, and you can see the same iterative reduction of the data set until what is sought is found.

I was part of the team in and around Imperial College, London, in the 1960s working on bubble searching techniques, and discovering what a revolutionary effect they have. The first level of processing is sequential, the kind if thing you perform on a calculator, which is step-wise. The second level adds logical branching, typically through an If... or When... test. Bubble sorting is the third level, and there are many more, right up to intuitive patterning and induction, which would result in human-indistinguishable transaction. And yes, my IQ at the time set the top limit of 163, I now know - I had just joined the baseline reference group, I have a share in the 2012 Nobel Peace Prize to prove it, for my work in Peacemaking, which passes through my mother to Ghandi, and is why I wish to remain anonymous, because we worked as a team, and it's impossible to say now who did what, not that it matters - the recognition goes to the team, not the player, because without the support of his peers, he might not have managed it alone. — Preceding unsigned comment added by 2.123.173.109 (talk) 10:25, 29 May 2018 (UTC)

The oldest documented mention of Emacs having an incremental search is in a 1978 document, linked from this very article. But as with most ideas and inventions, not just in software but in any endeavor, Emacs most surely built upon similar precursor ideas. Here's an example of an incremental search of the autocompletion variety from the DSD console driver program for Control Data Corporation (CDC) Cyber mainframe computers dating from the mid-1960s. This particular quoted section is from a 1975 printing of a KRONOS 2.1 manual, linked here: . Quote: "DSD is an interpretive display driver. When a console operator is typing a command, DSD completes the command as soon as it recognizes enough characters to establish the uniqueness of the command." 73.164.110.30 (talk) 03:30, 8 May 2019 (UTC)

less?

The article claims that incremental find is available in less, but it doesn't seem to be available in version 406 of less. Does anyone know if there exists a version of less that supports incremental find?

word wheeling

A similar article called word wheeling exists. I would like you guys to have a look at it. --Kushal^t 19:47, 26 October 2007 (UTC)

Fair use rationale for Image:Faytie7.jpg

Image:Faytie7.jpg is being used on this article. I notice the image page specifies that the image is being used under fair use but there is no explanation or rationale as to why its use in this Wikipedia article constitutes fair use. In addition to the boilerplate fair use template, you must also write out on the image description page a specific explanation or rationale for why using this image in each article is consistent with fair use.

Please go to the image description page and edit it to include a fair use rationale. Using one of the templates at Wikipedia:Fair use rationale guideline is an easy way to insure that your image is in compliance with Wikipedia policy, but remember that you must complete the template. Do not simply insert a blank template on an image page.

If there is other fair use media, consider checking that you have specified the fair use rationale on the other images used on this page. Note that any fair use images uploaded after 4 May, 2006, and lacking such an explanation will be deleted one week after they have been uploaded, as described on criteria for speedy deletion. If you have any questions please ask them at the Media copyright questions page. Thank you.

BetacommandBot 11:09, 27 October 2007 (UTC)

Linux

Non-modal incremental find is found in:

Modern operating systems:

• Apple Computer's Spotlight (which searches the entire computer);
• Microsoft's Windows Vista;

Is Linux just omitted is does it really not have it? ʄ!•¿talk? 17:25, 20 June 2008 (UTC)

All or most of the applications listed run on linux, so technically linux has this. The more detailed question is whether and where this feature appears in KDE, gnome and the like. Surely it's there; just a matter of identifying specific examples. dr.ef.tymac (talk) 23:57, 20 June 2008 (UTC)

Incremental search vs. incremental find

The term "incremental search" seems far more common than "incremental find": Google ["incremental search" -wikipedia] 90,500 vs. ["incremental find" -wikipedia] 34,900. In Google Books: 671/40. Though there are admittedly other meanings for "incremental search", scanning the results seems to show that "incremental search" is still more common in this sense than "incremental find" is; moreover, it is the name in Emacs, which originated it and is still in wide use. I suggest the article be moved. --macrakis (talk) 21:44, 20 December 2009 (UTC)

Incremental search - needs section on significance of incremental search on Web

My notes - that described (briefly) the fact that web based searches used non-local resources as opposed to local resources - have been edited out by Macrakis. The casual observer will now be left wondering why web searches use more resources. It is a very important point and has been covered in numerous blogs and criticisms of Google Instant for example.

This is what I had added:-

---

local scope v. internet scope

While offline (or private network) incremental searches utilize mainly "local" resources (such as a single personal computer's processor and its attached peripherals such as hard disks), an incremental web searches represents a paradigm shift in the scope of such a search.

An incremental search on the "Google web space" involves enlisting a global network of resources (including worldwide communications, online databases etc) and consumes bandwidth even during the time a user is entering individual keystrokes. The implications for efficiency are significant since, at the level of keystrokes (as opposed to whole words or phrases), the number of false positives is intrinsically quite high until more of the entire search key is eventually entered or selected. This will inevitably consume more resources (and use more energy) than would have occured on a full search key

A further disadvantage to web based incremental searches is the insurmountable problem of varying response times in relation to a users latent typing speed and, to some extent at least, "thinking" speed (since searches are often typed "on-the-fly"). If the response time is slowed (due to poor connection or other delays), the user may experience some disorientation effecting the overall "user experience" of the interface.

---

I believe my changes should be put back in. (unsigned comment by User:86.142.127.235 2011-09-16T08:15:18)

Under Wikipedia policy on reliable sources, blogs and Q&A forums are not reliable sources. It it certainly true that incremental web search requires additional communications resources and some incremental server resources, but how much more? Is it significant?

As for the quality of results, you are guessing that there is a large number of false positives. But, though Web search has a long tail of unique queries, it also has a big peak of common queries, for which a few keystrokes are enough. What proportion is an interesting empirical question, where a priori statements are not very useful. --Macrakis (talk) 15:54, 16 September 2011 (UTC)

You have been in the (academic?) computing sector for 30 years and you ask if requiring additional communications resources "and some incremental server resources" is significant? What planet have you been living on? I have been in the commercial computing industry for more than 40 years and I learned very early just how significant transaction rates and transmission speed really are. In fact, 99% of my time in the industry has been concerned with high transaction rate, on-line systems requiring < sub-second response times. Think bank teller machines, airline reservations systems etc. I have always tried to get the most out of available resources with the minimum of overhead and it is only recently that this has been on the green agenda relating to computing resources.

I do not think there is anything in my edit that is in any way controversial and rather is largely common sense to commercial IT people in the real world (but not necessarily to the casual obsever). In your condensed edit, by the way, you left out the varying "user experience" issue caused by uneven response times (which is very, very annoying to put it mildly). I could have raised the related issue of annoying repeated "flashing" of overlaying new data onto the screen - that may have significance to people with Photosensitive epilepsy (the World Wide Web Consortium - Web Content Accessibility Guidelines (WCAG) Version 2.0, produced in 2008, specifies that content should not flash more than 3 times in any 1 second period.). Clearly, someone who types rapidly (like myself) gets flashes much more frequently than that (when the internet is not running slowly that is). — Preceding unsigned comment added by 86.142.127.235 (talk) 07:34, 17 September 2011 (UTC)

These are all interesting issues, but Wikipedia requires reliable sources, not personal experience and observation. (And by the way, I have worked on commercial interactive systems for some time now.)

I certainly agree that both speed and efficiency are good things. But singling out one particular interaction technique for its effect on resource usage is not "common sense". Every aspect of every interactive system has some effect on resource usage. Google could use fewer server resources by (say) restricting search to the top 10,000,000 pages. It could reduce server and bandwidth usage by crawling Web sites less frequently. It could reduce bandwidth usage by only supporting low-resolution video on YouTube, or not allowing large gmail attachments, or (indirectly) by not searching video. It could only show 10 image results at a time. etc. etc. These are all tradeoffs between user experience and resource usage. If we can find a Reliable Source for the incremental server and bandwidth usage of these various techniques, then sure, let's refer to it. But it's not helpful to speculate that the resource usage of Google Instant is "significant" -- I suspect that video streaming swamps everything else.

As for the uneven response time issue, it would certainly be interesting to see a study of user reactions to the Google Instant experience. But just noting that you personally find it "annoying" is what we call WP:Original research. For the epilepsy issue, Google Instant is easy to turn off, which I believe satisfies WCAG.

If you learn of (or publish!) solid findings in these areas in reputable publications, then by all means let's see how we can incorporate the information into the article. --Macrakis (talk) 13:19, 17 September 2011 (UTC)

References

External links modified

Hello fellow Wikipedians,

I have just modified one external link on Incremental search. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://web.archive.org/web/20070423072215/http://www.ieforge.com/InlineSearch/HomePage to http://www.ieforge.com/InlineSearch/HomePage

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}).

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 07:37, 21 July 2016 (UTC)