Re: 301/302

On Wed, 3 Sep 1997, Larry Masinter wrote:

> >   I strongly agree with the 307 proposal cited above.  It is most
> >   unfair to those who have made the effort to read and understand the
> >   specifications to change them now by swapping the codes, even if
> >   there is time for them to recover (which we cannot know in any
> >   case).
> I don't think it is a good policy to make technical decisions on the
> grounds of whether or not the decision is 'unfair' to one or another
> group of implementors. We need to justify our decisions based on the

The swapping of 301/302/303 is an protocol change which is incompatible
with previous versions of the protocol published as RFCs. There is no
potential of additional incompatibility adding 307. It is impossible to
reach a 100% accurate conclusion about the impact of swapping 301/302/303.
Even as we debate HTTP applications are being built, referencing existing
RFCs and books based on those RFCs, by people who can't spell 'working
group'. Given the insignificant level of 'cruft' added to the protocol
by defining a new 307 status and protecting existing implementations,
why are we continuing to debate the issue and pondering the level of
damage we might do?

Add 307, deprecate 302 and let's get on with bigger issues.

Dave Morris

Received on Wednesday, 3 September 1997 12:29:48 UTC