Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!newstop!sunaus.oz!sunchat!markd From: Markd@Aus.Sun.COM Newsgroups: comp.lang.c Subject: Re: Hash??? Not quite clear on what this is... Message-ID: Date: 1 Dec 90 09:10:54 GMT References: <17291@netcom.UUCP> <1990Nov22.235522.26487@ux1.cso.uiuc.edu> <885@compnect.UUCP> Sender: news@sunaus.oz Organization: Sun Microsystems - Australia Lines: 36 johncore@compnect.UUCP (John Core ) writes: >In article <1990Nov22.235522.26487@ux1.cso.uiuc.edu>, gordon@osiris.cso.uiuc.edu (John Gordon) writes: >> avery@netcom.UUCP (Avery Colter) writes: >> >> >I'm seeing all this talk of "Hash Functions". >... >College programming courses seem to never deal with the REAL (ie buisness) >... >Actually you have to know a lot, like the key's max length >and how many total records will be in your data base, since you have >to give the hash function a range to deal with. Example: >... >50% of new entries were going into collision. Took 2 days to rebuild >the index. If you use simplistic static hashing in REAL buisness (sic) you deserve the heat. In real life, you need to know neither "the key's max length", nor "how many total records will be in your data base". Dbm in the former, and Dynamic Hashing (and it's numerous variants such as Probing) in the latter, demonstrate workable techniques that obviate the need for such prescience. Both of which have only marginal relevance to C. >--just another old fart telling a war story... Quite. ------------ ----------------- -------------------- Mark Delany markd@Aus.Sun.COM ...!sun!sunaus!markd ------------ ----------------- -------------------- Brought to you by Super Global Mega Corp .com