On Tue, 28 Feb 2006 01:26:12 +0100
"Dr. Stephen Henson" <[EMAIL PROTECTED]> wrote:

> On Mon, Feb 27, 2006, Michael B Allen wrote:
> 
> > Hi,
> > 
> > I have an idea for a new interface. At least from looking at the SSL man
> > pages it looks like everything depends heavily on reading and writing
> > directly to/from sockets. If the herein suggested interface already
> > exists then I'd really like to know more about it.
> > 
> 
> Actually the SSL library doesn't depend on sockets at all. Everything is done
> thorugh BIOs which are an I/O abstraction. 
> 
> These BIOs often communicate with sockets which is why several examples
> refer to them but they aren't compulsory.
> 
> If you want to use a different form of I/O you have two options.
> 
> One is to use a BIO pair which is analagous to a socket pair and then all data
> transfer is handled by the applications. There are various manual pages and 
> examples of the relevant discipline so I wont repeat that here.

Interesting. This seems like a very reasonable way to go. I had seen
the BIO stuff but the IO part led me to believe it was some kind of
I/O multiplexer. Looks like mostly an ops struct type thing now. That's
good. I was worried for a minute.

> 
> The other option is to write a BIO. That is less well documented but there are
> plenty of examples in the existing implementations. The socket BIO would be an
> interesting starting point.

This sounds like it would be more direct though. Might save a copy over
the BIO pair. And bss_sock.c doesn't look to complicated. Sounds like
I can write a BIO that does exactly what I want. I'll check it out.

Thanks,
Mike
______________________________________________________________________
OpenSSL Project                                 http://www.openssl.org
Development Mailing List                       [email protected]
Automated List Manager                           [EMAIL PROTECTED]

Reply via email to