Re: [Sbcl-devel] problem seeding randomness

[Joshua P. <jpa...@gm...>]

Hi Juho,

Thanks. How does one get a hold of the 0.9.12.3 version? I was able  
to find and build 0.9.12 from SourceForge, but the test I sent still  
doesn't work, so I imagine it isn't the ".3"

Thanks again!

Best,

Josh

******************************************
Joshua Parmenter
joshp@u.washington.edu
Post-Doctoral Research Associate - Center for Digital Arts and  
Experimental Media
Raitt Hall - University of Washington
Seattle, Washington 98195

http://www.dxarts.washington.edu
http://www.realizedsound.net/josh/


On Apr 27, 2006, at 9:05 AM, Juho Snellman wrote:

>
> Joshua Parmenter <jpa...@gm...> writes:
>> Hi,  Sorry if this has come up before, but I am new to the list and
>> couldn't find much that related to this in the documentation.
>>
>> I seem to be having a problem seeding the state of *random-state* on
>> sbcl 0.9.11 on an Intel Mac.  Here is an example:
> [...]
>> * (defvar *seed* (make-random-state))
>> SEED
>> * (progn
>>    (setq *random-state* (make-random-state *seed*))
>>    (loop for i from 0 below 10 collect (random 1000)))
> [...]
>
> The REPL calls the compiler to evaluate the LOOP form, and the
> compiler needs to call RANDOM every now and then. So *RANDOM-STATE*
> might or might not get modified between the SETQ and the first call
> your code makes to RANDOM.
>
> As it happens, the same issue was last mentioned on this list a couple
> of days ago ;-) This has been fixed in 0.9.12.3.
>
> -- 
> Juho Snellman
>