[
Lists Home |
Date Index |
Thread Index
]
Indeed. The accronym that I find most useful when thinking about
security matters is CAIN :-
(C)onfidentiality (message encryption at the transport or message
(preferred) level)
(A)uthentication/Authorisation (various types of secure token e.g. certs)
(I)ntegrity (tamper-proofing - usually dig sig)
(N)on repudiation (usually dig sig)
So depending on what features you need (often more than one), select
your poison.
Fraser.
On 14/07/06, Mitch Amiano <mitch.amiano@agilemarkup.com> wrote:
> An encrypted file need not be signed at all, and a signed file need not
> be encrypted.
> The two things - signing and encrypting - are distinct operations.
> One you do to ensure no one can read the data that shouldn't be reading it.
> The other you do to ensure that no one has tampered with data that
> shouldn't be tampered with, while not necessarily encumbering the
> ability to read it.
>
> Now, I'm not a security expert. Someone with more experience in this
> area may correct me on this, and could speak to the issue a bit more
> practically.
>
> But encryption alone is insufficient. One reason is that someone might
> well encrypt another file and substitute it for your original encrypted
> package. With a signature, both you and the receiver can perform a
> subsequent test that the signature and file still match up. Of course,
> if the signature is also with the original data, and that's your only
> copy, then someone could replace the signature too. Even if not, you or
> the receiver could conceivably maliciously replace both the file and
> the signature, thus creating an uncertainty about whose copy is authentic.
>
> Dave Pawson wrote:
> > On Fri, 2006-07-14 at 13:21 -0400, Mitch Amiano wrote:
> >
> >> https lets you send the data within a stream of packets of encrypted data.
> >> The signature gives you confidence that an unencrypted packet of data
> >> hasn't been altered.
> >> To take a document and encrypt it, so it is unreadable without
> >> decrypting, you could use encryption software such as GNU Privacy Guard
> >> or an API's crypt function.
> >>
> >
> > Would this be prior to or after 'signing' it?
> > If xml-sig is an analogy of an sha1sum, surely after?
> >
> > I guess a 'standard' crypt library is good enough for data protection
> > act, due care etc;
> > though I do need one for a Java client and Microsoft server (which
> > doesn't make
> > for an easy life :-)
> >
> > regards
> >
> >
> >
>
>
> -----------------------------------------------------------------
> The xml-dev list is sponsored by XML.org <http://www.xml.org>, an
> initiative of OASIS <http://www.oasis-open.org>
>
> The list archives are at http://lists.xml.org/archives/xml-dev/
>
> To subscribe or unsubscribe from this list use the subscription
> manager: <http://www.oasis-open.org/mlmanage/index.php>
>
>
|
