egghelp/eggheads community Discussion of eggdrop bots, shell accounts and tcl scripts. 2002-06-18T18:23:00-04:00 https://forum.eggheads.org/app.php/feed/topic/1982 2002-06-18T18:23:00-04:00 2002-06-18T18:23:00-04:00 https://forum.eggheads.org/viewtopic.php?p=8143#p8143 <![CDATA[Random numbers (srand)]]>
Second, the eggdrop rand generator is already seeded. It's probably just a crappy algorithm.

You could try using tcl's instead, or you could use the md5 function to create pseudo-random numbers.

Code: 

set _last_rand [clock seconds]proc md5rand {len} {  global _last_rand  set r [md5 "[clock seconds] $_last_rand"]  set _last_rand $r  set x1 [string range $r 0 7]  set x2 [string range $r 8 15]  set x3 [string range $r 16 23]  set x4 [string range $r 24 31]  scan "$x1 $x2 $x3 $x4" "%x %x %x %x" i1 i2 i3 i4  return [expr ($i1 + $i2 + $i3 + $i4) % $len]}
I did a little test to see the randomness of this vs. rand:

.tcl for {set i 0} {$i < 100} {incr i} {lappend md5rands [md5rand 10]}

.tcl for {set i 0} {$i < 100} {incr i} {lappend rands [rand 10]}

.tcl for {set i 0} {$i < 100} {incr i} {lappend tclrands [expr round(rand() * 1000000) % 10]}

.tcl putlog [lsort $tclrands]


.tcl putlog [lsort $md5rands]
[13:59] 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 8 8 8 8 8 8 8 9 9 9 9 9 9 9 9 9

.tcl putlog [lsort $rands]
[13:59] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 2 2 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 8 8 8 8 8 8 9 9 9 9 9 9 9 9 9 9 9

.tcl putlog [lsort $tclrands]
[13:59] 0 0 0 0 0 0 0 0 0 0 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 8 8 8 8 8 9 9 9 9 9 9 9 9 9 9 9

To be perfectly distributed, each number should appear 10 times. In md5rand, the most deviant waa 8 (appeared 14 times). For rand, the most deviant was 2 (appeared 2 times). For tclrand, the most deviant was 1 (appeared 4 times). You'd have to compute more statistics to really tell, but it looks like md5rand is superior.

Statistics: Posted by stdragon — Tue Jun 18, 2002 6:23 pm

]]> 2002-06-18T08:23:00-04:00 2002-06-18T08:23:00-04:00 https://forum.eggheads.org/viewtopic.php?p=8132#p8132 <![CDATA[Random numbers (srand)]]>
putserv "PRIVMSG $chan :...[lindex $response_array [rand [llength $response_array]]]"

Which works, and gives a fairly random set of responses. But its seems to pick some answers a lot more often than others, and some it misses out entirely.

In other languages Ive used you have to seed the random number generator before using it. But I cant seem to do that in my scripts.

srand [clock clicks]

gives this error

Tcl error [do_response_chan]: invalid command name "srand"

even tho srand seems to be in the tcl manual. Any ideas what Im doing wrong? Do I need to the seed the generator, or am i just being paranoid?

Statistics: Posted by Guest — Tue Jun 18, 2002 8:23 am

]]>