Re: [IPP] [Technical Errata Reported] RFC8011 (6085)

Follow-up from today's IPP WG conference call - the consensus is to change section 5.1.4 to read:

The 'keyword' attribute syntax is a sequence of characters, of length 1 to 255, containing only the US-ASCII [RFC20] encoded values for uppercase letters ("A"-"Z"), lowercase letters ("a"-"z"), digits ("0"-"9"), hyphen ("-"), dot ("."), and underscore ("_"). The first character SHOULD be a lower letter, and all letters SHOULD be lowercase.

Status should change to "confirmed", and we (at some point) will want to do an update (sigh).

> On Apr 10, 2020, at 8:31 PM, Michael Sweet via ipp <ipp@pwg.org> wrote:
> 
> Ira,
> 
> Looks like my attempts to get my contact email changed haven't completely been successful, either... :/  Comments below...
> 
> 
>> On Apr 10, 2020, at 2:07 PM, Ira McDonald <blueroofmusic@gmail.com> wrote:
>> 
>> Hi Mike,
>> 
>> We need to look into this errata on RFC 8011.
>> ...
>> --------------------------------------
>> Type: Technical
>> Reported by: Thomas Urban <thomas.urban@cepharum.de>
>> 
>> Section: 5.1.4
>> 
>> Original Text
>> -------------
>> The 'keyword' attribute syntax is a sequence of characters, of length
>> 1 to 255, containing only the US-ASCII [RFC20] encoded values for
>> lowercase letters ("a"-"z"), digits ("0"-"9"), hyphen ("-"), dot
>> ("."), and underscore ("_").  The first character MUST be a lowercase
>> letter.
>> 
>> Corrected Text
>> --------------
>> The 'keyword' attribute syntax is a sequence of characters, of length
>> 1 to 255, containing only the US-ASCII [RFC20] encoded values for
>> uppercase letters ("A"-"Z"), lowercase letters ("a"-"z"), digits 
>> ("0"-"9"), hyphen ("-"), dot ("."), and underscore ("_").
> 
> I think we need to be careful about changing this, but AFAIK this change shouldn't cause a problem (CUPS will accept this, at least...)
> 
>> Notes
>> -----
>> First, the "keyword" syntax is applicable to values of enumerations according to section 5.1.5 stating
>> 
>>   Each value has an associated 'keyword' name.
>> 
>> However, section 5.4.15 is declaring some enum-type attribute with names per integer value using uppercase letters in violation of 5.1.4. Those names are commonly used all over the specification and thus it is rather common to assume those values are meant to be keyword-compliant names of given enumeration.
> 
> First, these are operation names, and second the keyword associated with an enum is informative and not transmitted over the wire...
> 
>> Second, section 5.1.4 is stating
>> 
>>    The first character MUST be a lowercase letter.
>> 
>> referring to "a"-"z" according to enumeration of accepted characters given right before that. In opposition to that statement 5.4.14 is declaring
>> 
>>    The following standard 'keyword' values are defined in this document:
>> 
>>    * '1.0' [..]
>>    * '1.1' [..]
>> 
>> Neither of the two "keywords" start with a lowercase letter.
> 
> True.
> 
> ________________________
> Michael Sweet
> 
> 
> 
> _______________________________________________
> ipp mailing list
> ipp@pwg.org
> https://www.pwg.org/mailman/listinfo/ipp
> 

________________________
Michael Sweet