Path: utzoo!attcan!uunet!zephyr!tektronix!psueea!parsely!bucket!leonard
From: leonard@bucket.UUCP (Leonard Erickson)
Newsgroups: comp.binaries.ibm.pc.d
Subject: Re: An oddity with looz and turbo pascal's random
Keywords: zoo, looz, turbo pascal, random
Message-ID: <1511@bucket.UUCP>
Date: 29 Jun 89 05:40:43 GMT
References: <3008@sol.warwick.ac.uk>
Reply-To: leonard@bucket.UUCP (Leonard Erickson)
Organization: Rick's Home Grown Unix; Portland, OR
Lines: 35

In article <3008@sol.warwick.ac.uk> bsrdp@warwick.ac.uk (Hylton Boothroyd) writes:
<A couple of years ago I wrote a short turbo pascal 3 programme to generate
<8-character random alpha strings.
<
<I've just been surprised by the result of putting a copy in a .ZOO and
<running it under MSDOS 3.20 on an XT clone. Using
<	   looz xx zoohive rndname
<I get the same string each time:
<	   kkiioriv .
<
<Giving the programme a different name and putting it in a different .ZOO
<yields the same string!
<
<The command works OK at DOS command level - or more strictly at the DOS/CED
<command level that one of Rahul's remarks nudged me towards.
<
<The odd behaviour of this one programme is not important to me in itself. I
<just happened to use it as an example in testing out a set of command
<structures.  What disturbs me is that I can't see why it might happen, and I
<therefore can't guess the class of circumstances in which there might be
<changed behaviour.

As I recall the random routines in TP3 use some chunk of memory for a seed. 
When you run it from inside the Zoo file, this chunk of the DOS environment
will *always* be the same. Try putting the program in your autoexec and
you'll get the same kind of result.

Remember, programs *can't* generate truly random numbers, only psuedo-
random ones. If the machine state is the same, the output will be the same.

-- 
Leonard Erickson		...!tektronix!reed!percival!bucket!leonard
CIS: [70465,203]
"I'm all in favor of keeping dangerous weapons out of the hands of fools.
Let's start with typewriters." -- Solomon Short