Re: Calculating entropy for UTF-8
- From: Kless <jonas.esp@xxxxxxxxxxxxxx>
- Date: Thu, 31 Jul 2008 08:59:40 -0700 (PDT)
On 31 jul, 16:40, Ilmari Karonen <usen...@xxxxxxxxxxxxxx> wrote:
On 31.07.2008, Kless <jonas....@xxxxxxxxxxxxxx> wrote:>Yes, it's true, you have many reason. I was confused.
In either case, the way you encode the characters (UTF-8 or anything
else) doesn't matter. All that matters is the number of possible
characters.
I'm not as sure about this since that if i.e. you need 16 bytes for a
IV, and is created with ASCII and latin-1 characters, then its length
will be more than 16 bytes althought it shows 16 characters.
Obviously. However, if you need 16 random bytes for an IV, then 16
random bytes (as in, each byte randomly chosen from the range 0-255)
is what you should generate. If you limit your bytes to, say, valid
printable UTF-8 characters, then some of the bits in them will not be
random, and therefore the IV will contain less than 16 bytes of
entropy.
Thanks!
.
- References:
- Calculating entropy for UTF-8
- From: Kless
- Re: Calculating entropy for UTF-8
- From: Ilmari Karonen
- Re: Calculating entropy for UTF-8
- From: Kless
- Re: Calculating entropy for UTF-8
- From: Ilmari Karonen
- Calculating entropy for UTF-8
- Prev by Date: Re: Calculating entropy for UTF-8
- Next by Date: Re: Calculating entropy for UTF-8
- Previous by thread: Re: Calculating entropy for UTF-8
- Next by thread: Re: Calculating entropy for UTF-8
- Index(es):
Relevant Pages
|