Date: Sun, 2 Feb 2003 13:23:38 -0800 (PST) From: Doug Barton <DougB@FreeBSD.org> To: Bakul Shah <bakul@bitblocks.com> Cc: current@FreeBSD.org Subject: Re: rand() is broken Message-ID: <20030202131437.V79739@12-234-22-23.pyvrag.nggov.pbz> In-Reply-To: <200302022006.PAA25355@valiant.cnchost.com> References: <200302022006.PAA25355@valiant.cnchost.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 2 Feb 2003, Bakul Shah wrote:
> Yes, *I* can do it but I don't work at every place they do
> simulation!
Well the code is still going to be available in cvs. It's not like we're
going to magically make it disappear. :)
> If in the extreme you are suggesting that a
> portable application shouldn't rely on any OS features, you
> are of course right but that kind of makes mockery of any
> claims of compatibility. The point of compatibility is to
> not break interfaces unless there is a significant benefit.
I can think of one significant benefit... I had noticed that my perl
script to pick random windowmaker themes (which uses rand()) seemed to be
picking the same themes over and over again. Now I know why. :) I had to
create a "last seen" list to artificially increase the "randomness." Oh
wait, TWO benefits... xscreensaver is having the same problem.
Seriously though, I think the (vast?) majority of our userbase expects
random (or even pseudo-random) numbers to be, well, random. For a
sufficiently naive definition of random anyway. While I can certainly
sympathize with your needs here, with all due respect I think it's an edge
case, and one that could be handled with a little customization on your
part. Even if you didn't want a code solution, you could spool off BIGNUM
"random" numbers from whatever source you choose and save them to disk so
that you could get 100% repeatability. Your needs have more than one
solution, but I don't think that depending on rand[om]() not to be random
is a good one.
Doug
--
If it's moving, encrypt it. If it's not moving, encrypt
it till it moves, then encrypt it some more.
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030202131437.V79739>