[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Tue, 12 Dec 2023 09:17:29 -0800
From: Stephen Hemminger <stephen@...workplumber.org>
To: "Fernando F. Mancera" <ffmancera@...eup.net>
Cc: Donald Hunter <donald.hunter@...il.com>, netdev@...r.kernel.org
Subject: Re: Preferred term for netdev master / slave
On Tue, 12 Dec 2023 10:57:05 +0100
"Fernando F. Mancera" <ffmancera@...eup.net> wrote:
> On 11/12/2023 15:05, Donald Hunter wrote:
> > I'm working on updates to the YNL spec for RTLINK and per
> > Documentation/process/coding-style I want to avoid any new use
> > of master / slave.
> >
> > Recommended replacements include:
> >
> > '{primary,main} / {secondary,replica,subordinate}'
> > '{initiator,requester} / {target,responder}'
> > '{controller,host} / {device,worker,proxy}'
> >
> > Is there an existing preference for what to use in the context
> > of e.g. bridge master / slave?
> >
>
> Hi Donald,
>
> In other projects like NetworkManager, rust-netlink libraries.. the
> terms that are being used in the context of linux bridging are
> controller / port.
>
> > If not, then how about one of:
> >
> > primary / secondary
> > main / subordinate
> >
> > Thanks!
> >
>
> Thanks!
>
For iproute2 my intention is to replace slave with member
and not use the term master and instead refer to the interface as bridge.
Ditto for bonding.
This aligns with some other OS and implementations.
Powered by blists - more mailing lists