RPC and IPv6

[Thorsten Kukuk]

On Fri, May 31, Ulrich Drepper wrote:

> On Fri, 2002-05-31 at 00:20, Thorsten Kukuk wrote:
> 
> > because with ti-rpc we can communicate 
> > with all other IPv6 ready RPC implementations. With everything else we
> > are not able to do so.
> 
> According to you we cannot even talk to earlier glibc versions which is
> completely unacceptable.

No, I never said this. I wrote that we would break the protocol if we
would add IPv6 support to the current sunrpc code as we tried in the 
past. ti-rpc speaks the old and the new protocol.

> > I don't see your problems: We have symbol versioning with which we
> > should be able to solve the problems. But more after I know exact
> > what will break.
> 
> You haven't read what I wrote, have you?  

I did. But it seems you mix ti-rpc with our tries to add IPv6 to
sunrpc in the past.

> Symbol versioning cannot cover
> up the complet breakage of a large number of interfaces reliably

We don't break a large number of interfaces.

> and it cannot at all cover up protocol incompatibilities.

That's the reason for using ti-rpc instead of hacking sunrpc: With
ti-rpc we don't have the protocol incompatibilities we would introduce
with a hacked sunrpc version.

  Thorsten

-- 
Thorsten Kukuk       http://www.suse.de/~kukuk/        kukuk@suse.de
SuSE Linux AG        Deutschherrnstr. 15-19        D-90429 Nuernberg
--------------------------------------------------------------------    
Key fingerprint = A368 676B 5E1B 3E46 CFCE  2D97 F8FD 4E23 56C6 FB4B
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://sourceware.org/pipermail/libc-alpha/attachments/20020531/848aa21a/attachment.sig>