Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.2 9/18/84; site brl-tgr.ARPA Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!think!harvard!seismo!brl-tgr!tgr!cottrell@nbs-vms.ARPA From: cottrell@nbs-vms.ARPA (COTTRELL, JAMES) Newsgroups: net.lang.c Subject: Forever Message-ID: <1825@brl-tgr.ARPA> Date: Tue, 1-Oct-85 14:47:51 EDT Article-I.D.: brl-tgr.1825 Posted: Tue Oct 1 14:47:51 1985 Date-Received: Thu, 3-Oct-85 06:28:34 EDT Sender: news@brl-tgr.ARPA Lines: 28 /* Old business department. Doug Gwyn sez: > I think this illustrates one of the potential problems with > hiding the actual language underneath definitions. If the > loop REALLY continues "for ever", then it is probably not a > good algorithm. There are *lots* of `forever' type loops. How about process control applications, or communication boxes? These processes *never* exit (they are forcibly shutdown by pulling the plug, rebooting, or `exit' somewhere in a called funxion). > Possibly #define repeatedly for ( ; ; ) > would be a better fit to English. `Repeatedly' vs `forever'? Sounds like quibbling to me. > But (unlike the Boolean > data type) the extra effort of becoming familiar with this > usage does not seem to be adequately repaid in clearer code. > (Of course that's a judgement call..) I agree. My `handy.h' usually include a forever define which is forever unused. I just never remember to type it instead of `for(;;)'. jim cottrell@nbs */ ------