[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <a1a061cac945280cff1877cba6082f17@smtp.hushmail.com>
Date: Mon, 29 Apr 2013 20:20:12 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: Node support
On 29 Apr, 2013, at 19:20 , magnum <john.magnum@...hmail.com> wrote:
> On 29 Apr, 2013, at 17:20 , magnum <john.magnum@...hmail.com> wrote:
>> I think I see a bug with -stdin that exist in core too.
>
> Trying to establish the problem in core. Actually the same problem is seen with --wordlist too, which makes sense because core uses mostly the same code for wordlist/stdin. I have only seen problem with -stdin in bleeding.
OK I now understand what happens. It's because of my uneven splitting (sometimes I play clever). The code is wrong for the case I used, ie:
Session A running 1-3/7
Session B running 4/7
Session C running 5-6/7
Session D running 7/7
What happens is,
A reads line 1, 6, 11 and so on.
B reads line 4, 11, 18...
C reads line 5, 11, 17...
D reads line 7, 14, 21...
And no-one reads line 2, 3, 8, well you get it.
I think A and C have to to read 3 and 2 consecutive words (even if that means slightly less optimal spread). And skip 4 / 5 words after that. I can't think of any better solution but I wouldn't be surprised if you can.
magnum
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
