[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <571D56B1.4070209@openwall.net>
Date: Sun, 24 Apr 2016 18:28:49 -0500
From: jfoug <jfoug@...nwall.net>
To: john-users@...ts.openwall.com
Subject: Re: Rules 'C' command, wtf???
On 4/24/2016 4:35 PM, Solar Designer wrote:
> Jim,
>
> On Sun, Apr 24, 2016 at 12:24:58PM -0500, jfoug wrote:
>> c capitalize
>> C lowercase the first character, and uppercase the rest
>>
>> I do understand the why of the McXyz handling within C but it is not
>> documented that way, and there are other cases which would also fall
>> into this situation. I really think this is the wrong place to put this
> I agree. This is something I did in very early versions of JtR, and it
> just stayed in there. Some years back, I finally regretted doing it.
> Maybe we should drop this undocumented special handling now, given that
> it's a minor detail and is probably not replicated by other tools
> implementing similar rule commands. If you drop this from jumbo now, I
> will also drop it from core.
I will make the change to 'C' and 'c' handling in jumbo. It took me for
a loop, since I had rules that 'should' have been identical all of a
sudden find 'differing' outputs. The same rules were very trivial. The
rule u and CT0 should have been identical. But all of a sudden, I had
a run with some random data input where they were not?? Ended up being
a line starting with mCr It took me for a loop that the u was MCR but
for some reason, the CT0 ended up with MCr
Yes, I DO think that having a way to do casing and reverse casing of the
McXxxx type name is a good thing, but having that be the 'default' for
'c' or 'C' likely is not what the end user wants.
Btw, I did look at other tools, and they work like how things are
documented. c upcases first letter, and lowers the rest. C is the
inverse. So removing this strange behavior from john will get the final
result set closer.
https://github.com/magnumripper/JohnTheRipper/commit/a08c37672949941a471dc8531210c98ae7445744
The above commit removes the McXyz logic from jumbo. Now the logic for
c and C match the documentation. Was this something that Cracker Jack
did ? Or was this just a 'wow neat idea' you had early on (which btw,
'is' a neat idea and probably does help crack names, but seems ugly for
default behavior).
Jim.
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.
