Date: Mon, 07 Oct 1996 17:36:42 -0400 From: "Kevin P. Neal" <kpneal@pobox.com> To: "=?KOI8-R?Q?=E1=CE=C4=D2=C5=CA_=FE=C5=D2=CE=CF=D7?=" (Andrey A. Chernov) <ache@nagual.ru> Cc: terry@lambert.org (Terry Lambert), joerg_wunsch@uriah.heep.sax.de, freebsd-hackers@FreeBSD.org, current@FreeBSD.org, bde@zeta.org.au Subject: Re: I plan to change random() for -current (was Re: rand() and random()) Message-ID: <1.5.4.32.19961007213642.006c157c@mindspring.com>
next in thread | raw e-mail | index | archive | help
At 11:26 PM 10/7/96 +0400, Andrey A. Chernov wrote: >> Unless you are a mathematical programmer, you are unlikely to be >> able to aprehend the consequences of even a trivial change away from >> mathematical standards will have. There are verifiable standards >> of correctness, and each standard dictates issues of precision to >> which one can trust the code. Obviously, differences after the >> significant digits can be ignored for comparison -- and are, in fact, >> stripped from results as the "noise" that they are. > >FYI, I am applied mathematic, B.S. degree. > >> I suggest strict adherence to standards -- mathematical standards, >> not ANSI or ISO C standards -- with regard to maintaining precision >> and historical implementation, as required to ensure repeatability >> and trust. > >Current random() code is joke from mathematical point of view (but not from >ANSI/ISO standards). It is why it needs fixing. Wait. I feel like I'm missing something here. The pseudo-random calls are documented. They have been for a long time. They give repeatable results, cross platform, from the desk machine to the supercomputer. Physics experiments, for example, rely on the results being predictable and repeatable. Some number of people (small or large, depending on your point of view) have programs to do, for example, physics experiments. These things take *way* too long to run, and produce *way* too much raw data, so in effect the experiments are not re-runable. You keep a couple of graphs to store the results. The proposed change would break all of the above. No extremely serious need to change the system has been shown, but an extremely serious need to *NOT* change the system has been shown. An alternate system has been suggested, that would give numbers that were more random, and still not break anything. (/dev/random?) Is that about the gist of it? And how much hardware isn't supported yet, while this argument about changing something minor goes on? How many features does, for example, Linux have, while a debate about pseudo-random numbers go on? What did I miss? -- XCOMM Kevin P. Neal, Sophomore, Comp. Sci. \ kpneal@pobox.com XCOMM "Corrected!" -- Old Amiga tips file \ kpneal@eos.ncsu.edu XCOMM Visit the House of Retrocomputing: / Perm. Email: XCOMM http://www.pobox.com/~kpn/ / kevinneal@bix.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1.5.4.32.19961007213642.006c157c>