Re: How to do the sunrpc revert?

["Carlos O'Donell" <carlos at systemhalted dot org>]

On Wed, May 9, 2012 at 2:13 PM, Roland McGrath <roland@hack.frob.com> wrote:
> If it's done (which I've said I'm against), then it certainly should
> not be done as an attempt to undo the work done in 2.14. ?If anything,
> it should be the minimum necessary to restore the API availability,
> while not being too desperately unclean--so perhaps more than the
> minimal patch that Fedora and OpenSUSE are using, but closer to that
> than to the pre-2.14 state. ?We still want to remove sunrpc from the
> libc API as soon as it's convenient to do so, not prepare to support
> it forever.

I think what we're going to checkin is minimal, and well tested.

I think it's a hack and should be fixed for 2.17 e.g. full reversion.

I would love to see TI-RPC replace the libc API, but that's going to
take work from more volunteers and the facts on the ground haven't
changed.

I am strongly against simply tearing out sunrpc just because we don't
want it, that's abdicating our responsibility to an API we've had in
our sources for a long time.

As glibc developers we are going to need to do the work to bootstrap
the TI-RPC project to the point where it works, where we show it
working, and then hand off to the distros in a responsible fashion.

It would make a good project for someone on the distro teams :-)

Cheers,
Carlos.