Hi,

In message "[ruby-talk:01073] Re: rand() / drand48()"
    on 00/01/08, Yukihiro Matsumoto <matz / netlab.co.jp> writes:

>I don't get it, sorry.  So do you think we should preserve this
>undocumented behavior or not?  For me, current behavior (returns
>random value for each run) is less confusing.

Sorry for my unclear words.  I agree your patch now, but you should
document the difference from the previous one.  For rand() family
user, the current behavior is strange a little.

-- gotoken