Re: SHA Question

"Neo" <neoscandal@xxxxxxxxx> wrote in message
news:Xns973CCBAFBE8FENeo@xxxxxxxxxxxxxxxxxx
> Paul Rubin <http://phr.cx@xxxxxxxxxxxxxx> wrote in
> news:7xslsbjugc.fsf@xxxxxxxxxxxxxxxxxxx:
>
>> Neo <neoscandal@xxxxxxxxx> writes:
>>> My question is thus.. given 160 bits... is it possible to predict if
>>> it is a SHA hash or not.. or in other word... if i were to SHA all
>>> 2^160 possibilities of data, would i get a pretty varied hashes
>>> (ideally) .
>>
>> You will get about 63% of the 2^160 possible outputs and about 37%
>> will be missing. There's no known feasible way to find out whether
>> any specific output occurs or not.
>>
>> This is pretty basic stuff in cryptography. There's nothing wrong
>> with not knowing it (everyone has to start somewhere) but if you don't
>> know it, this isn't the right time to be designing your own encryption
>> schemes.
>>
> I'm not designing my own encryption scheme. Its just password management
> i'm worried about.
>> If you're somewhat mathematically adept and determined and you want to
>> know this stuff thoroughly, try to slog through
>>
>> http://www-cse.ucsd.edu/users/mihir/cse207/classnotes.html
>>
>> After that, you'll be ready, or at least a lot closer.
>
>
> I'm going to implemenmt SHA-256. AES 256 would seem sufficient but i'll
> provide the option to the user. Also is there any documentation online
> that
> brings forth issues during hard disk file encryption.. like the memory
> being copied to the hard disk swap file.. I guess you can increase the
> priority of ur thread to maximum if you want to prevent this.
>
> Thanks to all.

No. That is not the way to fix that.

Are you sure you have sufficient background to make crypotographic software
that other people will be using?

--
LTP

:)


.

Loading