On Wed, Oct 26, 2011 at 12:36 AM, <[email protected]> wrote:

> "Milan Mimica"<[email protected]> wrote:
> > Yes, having the salt randomly generated and storing it with a hash is a
> > better idea. Note taken. Combining it with a fixed salt (and trying to
> keep
> > it secret) is even better. Keeping a hardcoded salt in the image running
> on
> > the remote machine serving WEB pages makes it quite secret IMO.
>
> I was referring to Mariano's intent (at least how I understand it) to
> hardcode it "in code". If he's confident he'll be able to keep the code
> secret then hey may as well have the password in it in plain text, hashing
> it with or without salt, doesn't make much difference IMO.
>
>
Yes, indeed. I didn't give details. It was something very very stupid and
simple. All I wanted to do is to commit a class I use to build my images and
such class sets my username/password for squeaksource repotistories. But I
didn't want to put such password in the code... at the end what I did
(because my scenario is really stupid and only for me), is to read the
password from a file in my machine :)  hahahahha

Anyway, I learn from the thread :)



> Generating random salt and keeping a hashed password on a deployed system
> is a different scenario. In this case it's different and unpredictable with
> every deployment. When it's hardcoded it's the same everywhere.
>
>


-- 
Mariano
http://marianopeck.wordpress.com

Reply via email to