Re: I-D Action: draft-carpenter-6man-rfc6874bis-00.txt

[Brian E Carpenter <[e-mail]>]

Nico,
On 07-Jul-21 06:03, Nico Schottelius wrote:
> 
> Andrew Cady <[e-mail]> writes:
>> Please remember two facts and a corollary that I posted here last week:
>>
>>   1.  RFC3986 says percent-decoding is done AFTER parsing into components
>>       and subcomponents
>>
>>   2.  The IPv6 literal subcomponent is NOT percent-encoded.
>>
>>   2a. No percent-decoding is done (or allowed) on this parsed subcomponent.
>>
>> (Ref:<https://mailarchive.ietf.org/arch/msg/ipv6/FPLeDZXqJ1zwE1yF_Qkh7Ldq120/>)
> 
> I've skimmed RFC3986 and have a bit of trouble reading it - and maybe
> that's also the reason why browser developers had trouble properly
> implementing it.
> 
> For me the three statements from Andrew are well readable and
> understandable.
> 
> Would it by any chance be easier to create a specific RFC
> "Parsing of IPv6 literal addresses in URIs"
> that basically re-iterates the three sentences above and maybe even
> includes a parser description?

I would think that *when* we have agreed on a definitive solution,
we should add that text as an appendix to rfc6874bis.

> They are well reasoned, easy to understand and easy to implement. Given
> that we add a "reference parser", which could even be a regexp,
> developers might be able to "copy and paste" the parser into their code
> (+/- language bindings).

I think the browser implementers know their business better than 6MAN does,
but having a sample implementation is never a bad thing.

Regards
   Brian