Re: [VCARDDAV] Signed vCards

[DataPacRat <[e-mail]>]

On Mon, Jul 15, 2013 at 1:12 PM, Simon Perreault
<[e-mail]> wrote:
> Le 2013-07-15 19:00, DataPacRat a écrit :
>> A vCard which is compatible with the signed vCard extensions MAY add
>> an 's' to the version number.
>> A vCard which uses any of the signed vCard extensions SHOULD add an
>> 's' to the version number.
>> A vCard which uses a negative number in the CONFIDENCE property (ie,
>> has a confidence of under 50%) MUST add an 's' to the version number.
>
> What's the purpose of the 's'?
>
> Why isn't the mere presence of a property a sufficient indicator?

My reasoning is that a piece of software can claim to be vCard-4.0
compatible by interpreting the standard specification, and simply
dropping and ignoring any properties that it doesn't understand (such
as the BIRTHPLACE property introduced in RFC 6474). For the properties
I'm proposing for signed vCards, that behaviour will still lead to
acceptable results... with the exception of signed vCards used for
revocation, those with CONFIDENCEs of <50%. Adjusting the version
number for that particular case will allow such programs to still
claim full compatibility with the 4.0 spec, even if they misinterpret
revocation vCards.


Another option that comes to mind could be to roll up the new tags
from RFCs 6473 and 6474 along with the signed vCard tags into the base
spec, and call the whole thing vCard 4.1.


Thank you for your time,
--
DataPacRat
"Then again, I could be wrong."