From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me X-Spam-Level: X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00 autolearn=ham autolearn_force=no version=3.4.4 X-Google-Thread: 103376,703c4f68db81387d X-Google-Attributes: gid103376,public X-Google-Language: ENGLISH,ASCII-7-bit Path: g2news1.google.com!news3.google.com!news.glorb.com!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail From: Vinzent 'Gadget' Hoefler Newsgroups: comp.lang.ada Subject: Re: Teaching new tricks to an old dog (C++ -->Ada) Date: Wed, 23 Mar 2005 13:00:42 +0000 Message-ID: <2312933.PZFk6lQT9j@jellix.jlfencey.com> References: <4229bad9$0$1019$afc38c87@news.optusnet.com.au> <1110032222.447846.167060@g14g2000cwa.googlegroups.com> <871xau9nlh.fsf@insalien.org> <3SjWd.103128$Vf.3969241@news000.worldonline.dk> <87r7iu85lf.fsf@insalien.org> <1110052142.832650@athnrd02> <1110284070.410136.205090@o13g2000cwo.googlegroups.com> <395uqaF5rhu2mU1@individual.net> <1110329098.642196@athnrd02> <1110361741.551255@athnrd02> <422edaec$0$26554$9b4e6d93@newsread4.arcor-online.net> <1111464133.508323@athnrd02> <423fe9df$0$11476$9b4e6d93@newsread2.arcor-online.net> <1111524623.705533@athnrd02> <4240937b$0$9215$9b4e6d93@newsread4.arcor-online.net> <1111568536.640808@athnrd02> <1948665.ii7ipjq4RT@jellix.jlfencey.com> <1111575196.993136@athnrd02> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7Bit X-Trace: individual.net ZH0hq4YKsPS3QtXCB37cHAO6G5wjgYMsnjVRpwCAgwQft9J6q3 X-Phone: +41 62 961 13 52 X-Mood: Beautiful day to take over the world. Xref: g2news1.google.com comp.lang.ada:9793 Date: 2005-03-23T13:00:42+00:00 List-Id: Ioannis Vranos wrote: > Vinzent 'Gadget' Hoefler wrote: > >> Yeah, right. Binary searches are not much slower. Only about 10 or 20 >> times. Seem, you just rendered hashing algorithms obsolete. > > OK, one can use a third-party hashed container then, like a hash_map: Which usually has it's worst case behaviour when it is close to full (unlike the array, where a time complexity of O(1) is guaranteed, hashing algorithm define this as average case at best). If that's ok, it is ok, sure. But sometimes it's just not what is needed. Sometimes hashes are appropriate, sometimes tree-like structures, sometimes vectors (or arrays, how I'd like to call them). They /all/ have their uses. _But_: To defy an O(1) example with an O(log n) one and then stating, even for big data sets, the time difference isn't that large (usually I do call an order of magnitude a large difference) is either just plain ignorant or plain stupid. Vinzent. -- worst case: The wrong assumption there actually is one.