Re: my dice roller thingy...

From: thomasSPAMGUARD@clark.net (Mark Thomas)

Date: Fri, 29 Sep 2000 00:00:00 +0000

In article <8r0e85$c06$1@borg.uits.indiana.edu>, mmeiss@borg.uits.indiana.edu wrote:
>In article <slrn8t753v.uji.bruno@cerberus.csd.uwm.edu>,
>Bruno Wolff III <bruno@cerberus.csd.uwm.edu> wrote:
>>On Thu, 28 Sep 2000 16:12:36 GMT, Chris Williams <williac6@rjrt.com> wrote:
>>>Hey as long as we're looking at die-rollers... check mine out too!  :)
>>
>>There isn't any information there on how you generate the rolls.
>>Where do you get your entropy?
>>How do you avoid bias?
>
>Are we going to hear an argument on why D&D requires such perfect
>randomness so that rand() just won't cut it?

If you want :)

rand() is usually very very poor for simulating dice, particularly if you use
any sort of modulo operation to produce your rolls. For example if you
generate 100 rolls with rand() modulo 10, you can see patterns in the rolls
without much effort.

Bruno's roller is interesting (thanks Bruno!). The disadvantage with entropy
based rollers is you can't reproduce the pattern of rolls from a seed value, a
feature which is nice in some game situations.

>More to the point would be to wonder why anyone would use a Web-based
>die-roller in the first place.  Writing one in order to teach oneself
>CGI, maybe.

No idea, but Bruno's code can be easily adapted to other uses.

Mark
-----
Mark Thomas thomasSPAMGUARD@clark.net http://www.clark.net/pub/thomas
PBEM Eldritch ----------------------> http://www.pbegames.com
Remove SPAMGUARD from address when replying.

References

Re: my dice roller thingy...
mmeiss@borg.uits.indiana.edu
Thu, 28 Sep 2000 00:00:00 +0000
my dice roller thingy...
"Chris Williams" <williac6@rjrt.com>
Thu, 28 Sep 2000 00:00:00 +0000
Re: my dice roller thingy...
bruno@cerberus.csd.uwm.edu (Bruno Wolff III)
Thu, 28 Sep 2000 00:00:00 +0000