>>>>> "Jason" == Jason Gunthorpe <[EMAIL PROTECTED]> writes:

    Jason> On 4 Mar 2001, Brian May wrote:

    >> So you think we should support IPv6 without the USAGI stuff,
    >> even if it means hacking servers to ignore the error code
    >> generated by bind?

    Jason> It would be nice if glibc did not return both v6 and v4
    Jason> sockets when you request a bind list on kernels that did
    Jason> not need that. IMHO that is the best solution.

Agreed. Ideally glibc should be able to insulate applications from
changes to the kernel.

Anyone willing to fix this and/or file a bug report?

I assume (hope!) that it is possible?
-- 
Brian May <[EMAIL PROTECTED]>


Reply via email to