# Search function from hell

**URL:** <https://boards.straightdope.com/t/search-function-from-hell/102059>\
**Category:** The BBQ Pit\
**Created:** [April 5, 2002, 8:20pm UTC](https://boards.straightdope.com/t/search-function-from-hell/102059 "2002-04-05T20:20:02Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![Hinten](https://avatars.discourse-cdn.com/v4/letter/h/ecd19e/32.png) [@Hinten](https://boards.straightdope.com/u/Hinten)\
**Post date:** [April 5, 2002, 8:20pm UTC](https://boards.straightdope.com/t/search-function-from-hell/102059/1 "2002-04-05T20:20:02Z")

</div>

"The search term you specified (raw) is under the minimum word length (4) and therefore will not be found. Please make this term longer. "

What’s wrong with this board?  
I remember some postings about the pros and cons of feading raw food to pets. So, instead of reposting the same question over again, being the well mannered Doper that I am, I decide to search for the old thread first.  
This miracle new search functionality tells me that the word RAW is not long enough by exactly one letter. WTFFF? Since when do search engines require me to get my Thesaurus out? Are you telling me that my vocabulary consists of too many three letter words and I better go and get a literature degree before I dare to use the search function again? POFS software!  
How the hell do I make the word raw longer then three letters? Anybody can come up with other three letter words that are not: and, are, is, its, etc.?

Grrrrrrrrrrrrrrrrrrrr!

---

<div class="post-metadata">

**Author:** ![The\_Ryan](https://avatars.discourse-cdn.com/v4/letter/t/7feea3/32.png) [@The\_Ryan](https://boards.straightdope.com/u/The_Ryan)\
**Post date:** [April 5, 2002, 10:34pm UTC](https://boards.straightdope.com/t/search-function-from-hell/102059/2 "2002-04-05T22:34:01Z")

</div>

Uh, “The”, “BBQ”, “Pit”?  
How about “uncooked”?  
This is rather annyoning. It seems to me that it should be able to search for long strings such as “raw food”, even if it has three letter words in it.

---

<div class="post-metadata">

**Author:** ![LindyHopper](https://avatars.discourse-cdn.com/v4/letter/l/ebca7d/32.png) [@LindyHopper](https://boards.straightdope.com/u/LindyHopper)\
**Post date:** [April 5, 2002, 10:36pm UTC](https://boards.straightdope.com/t/search-function-from-hell/102059/3 "2002-04-05T22:36:38Z")

</div>

Yeah, this is a pain in the ass. I figured it was a way to enable using connectors like AND and OR. That way, the engine doesn’t have to distinguish between connectors and legitimate search terms.

It’s still annoying.

---

<div class="post-metadata">

**Author:** ![Esprix](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/esprix/32/7899_2.png) [@Esprix](https://boards.straightdope.com/u/Esprix)\
**Post date:** [April 5, 2002, 10:49pm UTC](https://boards.straightdope.com/t/search-function-from-hell/102059/4 "2002-04-05T22:49:14Z")

</div>

> [@](#):
>
> \*Originally posted by Hinten \*
> 
> WTFFF?

WTF? Stutter much? What the fuck fuckity fuck? Who Tamed Full Fathom Five? When To Fix Flat Furs?

😉

Esprix

---

<div class="post-metadata">

**Author:** ![bobo](https://avatars.discourse-cdn.com/v4/letter/b/f9ae1b/32.png) [@bobo](https://boards.straightdope.com/u/bobo)\
**Post date:** [April 6, 2002, 1:54am UTC](https://boards.straightdope.com/t/search-function-from-hell/102059/5 "2002-04-06T01:54:26Z")

</div>

think about how that search works… it’s going to rip through the database looking for strings that match “raw”, which will display a lot of hits that don’t mean anything

%grep -c raw /usr/dict/words

returns “59”, which means 59 words containing the string “raw” are in the words file for my linux box, for example:

brawl  
brawler  
brawling  
brawn… etc etc etc (and that’s just the B’s… how many posts would have the string “raw” in them?)  
like I said, it’s a lame way of limiting massive responses from searches… but it’s a fairly trivial way of keeping people from wasting CPU cycles on those searches…

if you have too few letters, you’re going to get too many hits that won’t pertain to what you’re looking for, therefore, don’t waste the compute time to find all those worthless hits…

it’s not the best way to do it, but it accomplishes the same thing… the server is spared all those returns, with a single if-then… they’re sanity checking your query for you

case in point

%grep -c a /usr/dict/words

returns “23183”… that’s 23,183 words in the /usr/dict/words file that contain the string “a”… do you really want the server to respond to that? with every thread that contains an “a”… I’m no statistician, but I’d bet that would be extremely close to the number of threads ever posted (maybe someone, somewhere, started a thread saying “lets not use the letter before B, hint hint”… though I doubt that would work)

if you want some more reading on regular expressions “sed & awk” by O’Reilly publishing will give you a basic primer… past that, the “camel book” also by O"Reilly touches on the subject… past that, there are text books dedicated to the subject… once you understand how regular expressions work, then you can complain about one of your searches being too ubiquitious…  
(normally I’d apologize for being this informative/cruel, but it’s the pit, so fuck it.)

read a damn book… learn how to use that “magic box” in front of you… then if you see a problem, you can complain… don’t complain about a web site that keeps you from doing something stupid.

(ps… try doing a search for “raw food” or atleast “uncooked”, raw occures too much in standard words to be a good search term… maybe it’s smart enough to let you try " raw "… the spaces on either end should really limit the hits, and might placate the string length requirement)

damn that got long… long week, long rant…

bye all… back to lurking

---

<div class="post-metadata">

**Author:** ![bibliophage](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/bibliophage/32/7615_2.png) [@bibliophage](https://boards.straightdope.com/u/bibliophage)\
**Post date:** [April 6, 2002, 4:33am UTC](https://boards.straightdope.com/t/search-function-from-hell/102059/6 "2002-04-06T04:33:04Z")

</div>

Jerry the tech god has explained it in this ATMB post: [http://boards.straightdope.com/sdmb/showthread.php?postid=1958155#post1928155](http://boards.straightdope.com/sdmb/showthread.php?postid=1958155#post1928155)

> [@](#):
>
> \*Originally posted by jdavis \*  
> \*\*
> 
> That value was chosen because that is what the software upgrade defaulted to. The other Administrators can choose different values if they wish. Seeing how the vBulletin search index is already 1.4 GBs in size and it took me 12 days to create it I’m not in much of a mood to change the minimum word length.
> 
> The search index is already larger than the total RAM of the server. That does not make for fast searches in the first place. Do we want even slower searches and all the whines and moans of how much slower the board becomes as it is impacted by those even slower searches? I certainly don’t.
> 
> Jerry \*\*
