Re: [Cbor] CBOR tag range IANA allocation policy

[Thiago Macieira <[e-mail]>]

On Thursday, 25 July 2019 15:25:23 PDT Carsten Bormann wrote:
> On Jul 25, 2019, at 18:19, Sean Leonard <[e-mail]> wrote:
> > 24-255 = RFC Required
> 
> I like that.
> (That includes the independent stream.  But there is a level of quality
> checking in publishing an RFC, so only high quality specifications will get
> these.  Of course, the designated expert *could* do the same quality
> checking, but there may be a slightly higher tendency to just give in among
> us weaker DEs.)

The problem is less the quality checking and more the time required to write 
an RFC, with the overhead required of it.

For example, I am sitting on a bunch of simple types I wanted to recommend, 
but never had the time to review, much less write RFCs for.

https://gitlab.com/thiagomacieira/qtbase/blob/master/src/corelib/
serialization/cbor-tags-geometry.rst
https://gitlab.com/thiagomacieira/qtbase/blob/master/src/corelib/
serialization/cbor-tag-bitarray.rst

As you can see from the tag numbers, I wrote them a while ago but never told 
anyone about them. I find that those are generically useful types that 
probably should exist in the 32-255 range, but I'd like the community to agree 
with me.

And if we conclude that the range is correct, then I probably won't find the 
time. Nor do I think that an RFC is required to define a couple of tags for 2D 
and 3D geometry types.

-- 
Thiago Macieira - thiago.macieira (AT) intel.com
  Software Architect - Intel System Software Products