[Python Wpg] Tested seed()

[Mike Pfaiffer]

Richard Fillion wrote:
> You shouldn't pass a static value to a random seed generator.  Pass  
> something that will change, such as the current time, or date, or ...  
> anything not-constant.
> 
> Richard Fillion
> e: rick at centrix.ca
> p: 204.291.5800

	We had a discussion about this at the meeting. If you leave out the "2" 
the seed is based on the system time. I was curious if putting a 
constant in the seed() function would return the same values each time 
it's run. Nobody knew for sure. Nobody caught anything in the 
documentation. Now we know.

	Historically there are only two random generators I've seen which "self 
seed". The first is the Apple// (which I discussed at the meeting) and 
the second is an *old* PC card which got its seed from the amount of 
static electricity in the air.

				Later
				Mike



> On Jun 28, 2007, at 12:05 AM, Mike Pfaiffer wrote:
> 
>> 	Here is some code.
>>
>> #!/usr/bin/python
>>
>> from random import seed, randint
>>
>> seed(2)
>> for c in range(10):
>>      print c,randint(1,10)
>>
>>
>> 	Ran it twice under idle and the same results came out three times.
>> Looks like it is best to run it without a value in seed().
>>
>> 	Here is what I got...
>>
>> IDLE 1.2.1
>>>>> ================================ RESTART
>> ================================
>> 0 10
>> 1 10
>> 2 1
>> 3 1
>> 4 9
>> 5 8
>> 6 7
>> 7 4
>> 8 7
>> 9 7
>> 				Later
>> 				Mike
>>
>> _______________________________________________
>> Winnipeg mailing list
>> Winnipeg at python.org
>> http://mail.python.org/mailman/listinfo/winnipeg
>>
> 
> _______________________________________________
> Winnipeg mailing list
> Winnipeg at python.org
> http://mail.python.org/mailman/listinfo/winnipeg
>