BP - use of RFC2119

Hi,

A comment about the use of SHOULD, MUST, etc.
It comes from my reading of the section on vocabularies
http://w3c.github.io/dwbp/bp.html#dataVocabularies
but it may be in many more places in the vocabularies...

I am wondering whether we over-use RFC2119, or don't use it with the correct sentences. Our injunctions read sometimes weird, at least to me.
For example:

"Humans SHOULD be able to understand the vocabulary." and "Any person who wants to use or reuse a vocabulary SHOULD be able to do so."
-> I cannot but smile at the idea our document specifies RFC2119 requirements on humans ;-)
To me requirements are rather on the artefacts than on the persons who interact with them.

"The data SHOULD not be more complex to produce and re-use than what is necessary;"
-> This hints that we already foresee cases in which it's fully understandable that in spite of our recommendations, it is ok to be more complex than mecessary. What would be these cases?

"a first preliminary step SHOULD be to analyze whether"
-> Can we stay that a step SHOULD be something?

Compare with the following sentence, which read much less debatable:
"the vocabulary SHOULD be actively maintained"
"the vocabularies MUST have an associated sufficient documentation"

I'm ready to accept I'm nitpicking on this, but I thought I would still ask, in case I'm not the only one a bit puzzled.

Best,

Antoine

Received on Wednesday, 21 January 2015 23:12:58 UTC