27;t write an implementation
that handles them in different threads without having any thread
locking. This is what PureTLS does.
The major exception is handshake, which does have to be done in
lockstep. Also, you have to make sure that the threads interact
properly during closure.
pdate()
function already.
As for why it worked with SSLeay? That's puzzling, I admit. Perhaps
the function name changed or was only recently exposed to dynamic
linkage or something.
Try #defining SHA1_Update() to something else in the OpenSSL build
and see if that fixes the
which is to prevent active attack on the connection.
The vast majority of the complexity of SSL is there to prevent
active attack. By choosing to use unauthenticated certificates,
you are opening the door to a broad class of attacks.
-Ekr
__
oes
have certain technical properties which make it better than DH
in some circumstances.
Since then, the Merkle-Hellman patent has expired and DH has become
free. The RSA patent lasted a little longer, that's all.
-E
approved as an Informational RFC.
HTTP Upgrade (draft-ietf-tls-http-upgrade-05.txt) has been
approved as a Proposed Standard. Both documents are waiting
to pop out of the RFC Editor Queue.
They should be taken to be relatively stable, since only editorial
changes are supposed to occur
etc.) However, most of these end up being configured as RSA
only. In short, you're likely SOL.
-Ekr
__
OpenSSL Project http://www.openssl.org
User Support Mailing List
y the server does private key operations. Thus, if you want
to optimize your server, you might choose to use multiprime but
you would still interoperate with old clients.
In the face of hardware acceleration, this doesn
ipher.
Thus, since you've removed all the export ciphers from the
cipher list, the first handshake fails and the whole
process doesn't work.
-Ekr
--
[Eric Rescorla [EMAIL PROTECTED]]
PureTLS - free SSLv3/TLS software for Java
ht
s a function that takes no parameters (in C, f() is
the same as f(...)).
--snip--
-Ekr
--
[Eric Rescorla [EMAIL PROTECTED]]
PureTLS - free SSLv3/TLS software for Java
http://www.rtfm.com/puretls/
_
Dr Stephen Henson <[EMAIL PROTECTED]> writes:
> Are you sure its a SEQUENCE of and not just indefinite length encoding?
>
> If the latter then OpenSSL should handle it just fine.
It should be indefinite length encoding. That's what TIPEM has
emitted every time I'
enerate the primitive definite length
encoding, you have to read the entire file, encrypt it, and then
spit it out in one shot (So you can write the length octets
at the beginning.) This approach allows you to encrypt the
file in a single pass without buffering, since you write it
out one bite-si
an
early HMAC. TLS uses HMAC.
3. The Finished messages are different.
4. RSA key exchange in SSLv3 implementations does not comply with the
spec. It's supposed to be an opaque <2^16-1> but actually they
just put bytes directly in the record.
5. TLS has more alerts.
6. TLS requires DS
the web I find no reference to v4, so I
> conclude that either it doesn't exist, or is in a very preliminary
> development stage.
As far as I know there is currently no such thing as SSLv4 and never
will be.
TLS 1.0 is the successor to SSLv3.
-Ekr
--
[Eric Rescorla
occur to us to worry about the DH patent expiring. This
would not have been a convincing reason not to use RSA.
-Ekr
P.S. SSLv1 and v2 were not designed by Kocher et al. They were
designed by Kipp Hickman (also a Netscape employee) in the
fall of 1994.
--
[Eric Res
(afaik) - is the first
> (2) above?
Yes, this is a serious worry. OTOH, if you managed to securely
generate a private key, you must have had plenty of entropy
around at the time. You can store this entropy (using OpenSSL's
random seed file) and use it at signing time. Moroever, the random
hat difficult with
TLSGold. Though, again, you'd have to program it yourself
rather than just get some open source software that uses
OpenSSL, since the interface (though analagous of course)
is incompatible.
-Ekr
--
[Eri
Andrew Cooke <[EMAIL PROTECTED]> writes:
> EKR wrote:
> > Andrew Cooke <[EMAIL PROTECTED]> writes:
> > > Nicolas Roumiantzeff wrote:
> > > > Does anybody know why both IE and Netscape browser implement exclusively RSA
> > > > certificates?
icant amount of integration
work to plug it into a web server. If all you want is a single
secure web server, you'd be better buying one.
The only reason to buy the non-RSA version of TLSGold would be if
you wanted support, since it does essentially the same job
as OpenSSL.
Cheers,
-Ekr
Discla
initially a 'test site'. We need to rebuild the
> server in the near future, .. and I would be very interested in pros
> and cons.
You've missed at least one interesting option: use IIS on Windows. You
get SS
Vin McLellan <[EMAIL PROTECTED]> writes:
> Ekr> I don't believe this was the case. The original SSLv3 drafts
> Ekr> did not have DH/DSS/RC4 support. TLSv1 continued this.
> Ekr> The evidence that this was simply a glitch is that
> Ekr> DH_anon _was_ defined w
; non-Rivest ARC-4 source, but still labelling the new TLS ciphersuites as
> including "RC4."
I don't believe this was the case. The original SSLv3 drafts
did not have DH/DSS/RC4 support. TLSv1 continued this.
The evidence that this was simply a glitch is that
DH_anon _was_ de
NG, because the PRNG distills the randomness.
For instance, a badly biased RNG can still be used
to seed a PRNG. You just need more data.
-Ekr
--
[Eric Rescorla [EMAIL PROTECTED]]
PureTLS - free SSLv3/TLS software for Java
http://www.r
oken. It generate the DSA signature as concatenated
r,s rather than as a SEQUENCE OF as it's supposed to.
This may have been fixed in 4.7. I haven't checked.
-Ekr
--
[Eric Rescorla [EMAIL PROTECTED]]
PureTLS - free SSLv3/TLS sof
ooking for reference implementations so we can push this through to
> RFC status. If anyone has an implementation and is interested in collaborating
> - please contact me directly.
You do not need a reference implementation to go to Proposed
Standard -- the first step on Standards Track. To
rations is two digests, and MD5 digest
and a SHA-1 digest. When you're using RSA, you concatenate them
into one 36 byte string and RSA sign the entire thing. When you're
using DSA, you sign the SHA-1 digest only.
-Ekr
--
[Eric
25 matches
Mail list logo
