Re: [Cbor] Private Use Tags

On 2019-02-20 01:27, Carsten Bormann wrote:
> Hi Jörg,
> 
> The IETF has a long and checkered history with private-use allocations.
> 
> Generally, the observation is that private-use allocations start to get used for something that does get wide deployment, and suddenly, those allocations take on specific semantics and no longer truly are available for private use.
> This is somewhat related to the “X-Dash problem”, which is discussed in RFC 6648.

In my not unbiased opinion, it's also related to the limited domains problem [https://tools.ietf.org/html/draft-carpenter-limited-domains] and the whole topic of protocol extensions [https://tools.ietf.org/html/rfc6709]. Perhaps not a topic for this list, but I agree with Carsten that registering a tag intended for private use is much safer than using a potentially ambiguous value.

    Brian

> 
> Instead, you can just go ahead and register the tags “Jörg Richter Private 1” to “Jörg Richter Private 10”.  (Preferably in the 1+2 or, if the allocation is huge, 1+4 byte space.)  The Tag number space is huge, so it is unlikely we will run out because of such private use allocations.  And if one of these suddenly takes off, you can keep the number and change the name (and supply a specification, I hope).  (You might want to use a middle person for the allocation if you don’t even want to disclose that you are using CBOR Tags.)
> 
> Grüße, Carsten
> 
>> On Feb 19, 2019, at 11:11, Richter, Jörg <Joerg.Richter@pdv-FS.de> wrote:
>>
>> I had the idea to tag some internal types. These types are 
>> never transferred to other communication partners.
>> Would it be possible to provide a tag range that is explicitly
>> marked as "Private Use Area"? This would exclude the
>> possibility that these tags will be assigned in the future and
>> that there will be a collision.
>>  
>>  
>> _______________________________________________
>> CBOR mailing list
>> CBOR@ietf.org
>> https://www.ietf.org/mailman/listinfo/cbor
> 
> _______________________________________________
> CBOR mailing list
> CBOR@ietf.org
> https://www.ietf.org/mailman/listinfo/cbor
> 