----- Original Message ----- 
From: "Ryan Thompson" <[EMAIL PROTECTED]>


> Bill Landry wrote to Ryan Thompson:
>
> >    No documentation found for "check_whitelist".
> >
> > I'm running SA 2.64.
>
> Ahh. That explains it. check_whitelist is quite a bit better in 3.0. :-)
> 2.64's doesn't have perldoc, for one thing. For another thing, I just
> tried running it, and I was able to reproduce your results from a copy
> of a known-good database that shows proper scores with 3.0's
> check_whitelist.
>
> > Anyway, I figured that this was the column
> > discriptions, however, I was confused as to why my first and second
columns
> > showed the same score, instead of an average in the first column.  From
my
> > previous post:
> >
> >    35.2        (35.2/4)  --  [EMAIL PROTECTED]|ip=216.93
> >    16.9        (16.9/2)  --  [EMAIL PROTECTED]|ip=209.206
> >    32.0        (32.0/2)  --  [EMAIL PROTECTED]|ip=69.6
> >    30.3        (30.3/9)  --  [EMAIL PROTECTED]|ip=64.28
> >    22.3        (22.3/4)  --  [EMAIL PROTECTED]|ip=216.21
>
> It's the second column (TOTSCORE) that's out. See this:
>
>          $t/$v, (sprintf "(%.1f/%d)",$t/$v,$v),
>
> This is quite an obvious bug in check_whitelist; $t/$v is used for both
> the average score and the total score. The second $t/$v should be just
> $t. That line should read:
>
>          $t/$v, (sprintf "(%.1f/%d)",$t,$v),
>
> To verify, I ran 2.64's check_whitelist against 3.00's check_whitelist,
> on the same auto-whitelist file, and:
>
> # diff 2.64.txt 3.00.txt
> #
>
> So, with the fix, above, the results are identical.
>
> That also means you could just grab check_whitelist from 3.0's tools/
> directory. It's better documented, more robust, and it actually works.
>
> Still, I smell a bugzilla report coming on... :-/

Thanks Ryan, I grabbed a copy of check_whitelist from my 3.0.0.rc1 test
server and now have perldoc info and the output from the dump that I had
expected.

Thanks again for you help!

Bill

Reply via email to