Re: [apps-discuss] RFC 5785: Registration of .well-known services under HTTP to First Come

[Phillip Hallam-Baker <[e-mail]>]

On Fri, Jan 15, 2016 at 10:18 AM, Darrel Miller <[e-mail]> wrote:
> Philip,
>
>> From: apps-discuss [mailto:[e-mail]] On Behalf Of
> Phillip Hallam-Baker
>>
>> Most Web services make almost no use of HTTP features beyond the protocol
> identification described above > and the message framing. If one was
> developing a protocol that was only for Web Services, it would not have >
> features such as caching which are almost never desirable.
>
> I believe the assertion you make here is evidence that a designated expert
> is required.  You and I obviously live in very different parts of the web.
> My role on a daily basis is to interact with developers who are building
> services on the web.  My experience over the last two years is that
> developers are attempting to make more use of HTTPs capabilities as an
> application protocol and are shunning the use of HTTP simply as tunnel for
> bytes.

People keep making such assertions. I have yet to see anyone make a
specific example.

When I make a proposal, I give the technical basis for it. I do not
say that people should do it my way because I am the expert which is
essentially what you are doing here.

Now the inexperienced designer might take your claim to expertise at
face value and accept it. However the fact that you are trying to
one-up one of the original designers of HTTP by pulling rank suggests
that you are rather less expert than you might imagine.

This is yet another example of the insider clique behavior that I want
to see ended.


> I fully support the use of designated experts to control what becomes part
> of standard registries even though I have seen registrations refused that I
> believed should have been accepted.  I would much rather the bar be set high
> than see registries filled with poor practices.

Again, no specifics.

The bar to participation in this forum is an opinion and a keyboard.
It is deliberately set low because that is how the Internet emerged in
the first place.

There were plenty of people telling Clark, Cerf et al that their
design was 'mindnumbingly stupid' and the people telling them that
considered themselves to be the experts.


What you are proposing is actually contrary to one of the canons of
layered architecture. Dependencies between layers should be kept to a
minimum. I do not want a presentation layer that interacts with my
application layer. I want to decouple the two to the maximum extent
possible.

I suggest that you consider your arguments more carefully before this
discussion moves to the IETF list.