Forum: Ruby create a md5 / md5 passwd with a salt

Announcement (2017-05-07): www.ruby-forum.com is now read-only since I unfortunately do not have the time to support and maintain the forum any more. Please see rubyonrails.org/community and ruby-lang.org/en/community for other Rails- und Ruby-related community platforms.
19df66bc8d455d20a253bb1446f48349?d=identicon&s=25 Peter Woodsky (webspider)
on 2008-11-20 19:25
Hi list

This problem is so above me my head is swimming.

I have a forum software that recently changed the way passwords were
stored. I use a ruby script to update the user table when the user
changes the password but my script no longer works so I hope someone
here can help or point me in the right direction.

The forum now uses this format for storing passwords md5(md5(PASSWORD) .
salt) where previously it was md5(PASSWORD) which was pretty easy to
deal with.

This is what I used before:

require 'cgi'
cgi = CGI.new
require 'digest/md5'
user_pwd = hash_class.hexdigest("#{password}")
end

Like I said I hope someone has an idea or solution.

Thanks in advance,
Peter
3b1756d05466b4a78afd9aea7bb845c2?d=identicon&s=25 Aaron Turner (Guest)
on 2008-11-20 19:50
(Received via mailing list)
On Thu, Nov 20, 2008 at 10:21 AM, Peter Woodsky <peter@iwebsl.com>
wrote:
> salt) where previously it was md5(PASSWORD) which was pretty easy to
> deal with.
>
> This is what I used before:
>
> require 'cgi'
> cgi = CGI.new
> require 'digest/md5'
> user_pwd = hash_class.hexdigest("#{password}")
> end


user_pwd = Digest::MD5.hexdigest(Digest::MD5.hexdigest(password) + salt)

I will point out this though: for this to work, somewhere you need to
store the PLAIN TEXT salt in the DB.  Usually this is done by
pre-pending the salt to the hashed pasword like this:

user_pwd = salt + Digest::MD5.hexdigest(Digest::MD5.hexdigest(password)
+ salt)

Or it might be stored in another field in the DB.  That way when the
user logs in, you can use the salt again to test to see if the hashes
match.
753dcb78b3a3651127665da4bed3c782?d=identicon&s=25 Brian Candler (candlerb)
on 2008-11-20 21:23
Peter Woodsky wrote:
> The forum now uses this format for storing passwords md5(md5(PASSWORD) .
> salt) where previously it was md5(PASSWORD) which was pretty easy to
> deal with.

You should give an example of a password, a salt, and the expected
output. That's not any "standard" way of hashing an output that I've
seen.

Is the inner md5() function producing a binary output (16 bytes) or hex
(32 characters)? Or something else, e.g. base64 (24 characters)?

Anyway, it'll be easy enough to code if you know what you're looking
for. You can test if you're doing it right in irb.

Example: I've arbitarily chosen inner md5 is binary, outer MD5 is hex:

irb(main):001:0> require 'digest/md5'
=> true
irb(main):002:0> Digest::MD5.digest("hello")
=> "]A@*\274K*v\271q\235\221\020\027\305\222"
irb(main):003:0> Digest::MD5.digest("hello") + "mysalt"
=> "]A@*\274K*v\271q\235\221\020\027\305\222mysalt"
irb(main):004:0> Digest::MD5.hexdigest(Digest::MD5.digest("hello") +
"mysalt")
=> "263058590f9387d91f98266a832fabaf"
19df66bc8d455d20a253bb1446f48349?d=identicon&s=25 Peter Woodsky (webspider)
on 2008-11-20 21:40
Brian Candler wrote:
> Peter Woodsky wrote:
>> The forum now uses this format for storing passwords md5(md5(PASSWORD) .
>> salt) where previously it was md5(PASSWORD) which was pretty easy to
>> deal with.
>
> You should give an example of a password, a salt, and the expected
> output. That's not any "standard" way of hashing an output that I've
> seen.
>
> Is the inner md5() function producing a binary output (16 bytes) or hex
> (32 characters)? Or something else, e.g. base64 (24 characters)?
>
> Anyway, it'll be easy enough to code if you know what you're looking
> for. You can test if you're doing it right in irb.
>
> Example: I've arbitarily chosen inner md5 is binary, outer MD5 is hex:
>
> irb(main):001:0> require 'digest/md5'
> => true
> irb(main):002:0> Digest::MD5.digest("hello")
> => "]A@*\274K*v\271q\235\221\020\027\305\222"
> irb(main):003:0> Digest::MD5.digest("hello") + "mysalt"
> => "]A@*\274K*v\271q\235\221\020\027\305\222mysalt"
> irb(main):004:0> Digest::MD5.hexdigest(Digest::MD5.digest("hello") +
> "mysalt")
> => "263058590f9387d91f98266a832fabaf"

From the first solution given it works as it should. All that was
required was "" for the vars password and salt.

md5_w_salt = Digest::MD5.hexdigest(Digest::MD5.hexdigest("#{password}")
+ "#{row[0]}")

Thanks again!
753dcb78b3a3651127665da4bed3c782?d=identicon&s=25 Brian Candler (candlerb)
on 2008-11-20 22:18
Peter Woodsky wrote:
> From the first solution given it works as it should. All that was
> required was "" for the vars password and salt.
>
> md5_w_salt = Digest::MD5.hexdigest(Digest::MD5.hexdigest("#{password}")
> + "#{row[0]}")

Glad it works. Note that if password and row[0] are Strings, you should
not need to wrap them in "#{..}" since you are just interpolating a
String into another String.
Aad37b5f7116c8d1f547d23b37566032?d=identicon&s=25 Greg Donald (destiney)
on 2008-11-20 23:10
(Received via mailing list)
On Thu, Nov 20, 2008 at 12:21 PM, Peter Woodsky <peter@iwebsl.com>
wrote:
> require 'digest/md5'

MD5 is no longer secure:

http://www.google.com/search?q=md5+broken


There are a number of MD5 lookup dictionaries already online for a
couple years now, for example:

http://gdataonline.com/seekhash.php
http://passcracking.com/
753dcb78b3a3651127665da4bed3c782?d=identicon&s=25 Brian Candler (candlerb)
on 2008-11-21 10:12
Greg Donald wrote:
> On Thu, Nov 20, 2008 at 12:21 PM, Peter Woodsky <peter@iwebsl.com>
> wrote:
>> require 'digest/md5'
>
> MD5 is no longer secure:
>
> http://www.google.com/search?q=md5+broken
>
>
> There are a number of MD5 lookup dictionaries already online for a
> couple years now, for example:
>
> http://gdataonline.com/seekhash.php
> http://passcracking.com/

I think you're confusing two things.

1. MD5 has been broken, in the sense that you can make two versions of a
document with the same MD5 hash

2. You can have lookup dictionaries for hashes. However the same could
be done for SHA1 or any other hash, and does not depend on (1).

But in any case:

* the OP's algorithm has a salt. If well chosen (random and long
enough), this should eliminate the dictionary attack. (It would be
better if it were a HMAC construction though)

* the OP seems to have no choice in the algorithm to use anyway

* the OP wasn't asking for security advice :-)
This topic is locked and can not be replied to.