Closed agreiner closed 2 years ago
@agreiner I agree.
'Releases' are often given names. So I agree. And even release 'numbers' often have three dot-separated parts, therefore not a number.
Please note that the current definition is complemented by a usage note:
DCAT does not prescribe how a version name / identifier should be specified, and refers for guidance to [DWBP]'s Best Practice 7: Provide a version indicator.
So, we are not prescribing the use of numbers only.
For the records, the definition of dcat:version
re-uses the one of pav:version
, where the notion of "number" is used in a broad sense - quoting:
The version number of a resource. This is a freetext string, typical values are "1.5" or "21". The URI identifying the previous version can be provided using prov:previousVersion.
This property is normally used in a functional way, although PAV does not formally restrict this.
Yes, I understand that we do not want to prescribe the use of numbers, which is why I think we need to define it without describing it as a number. The document should be consistent in this respect. -Annette
version = "The version number of a resource"
version = a number or text that identifies a particular rendition of the resource version = a number or text that identifies a particular variation of the resource version = a number or text that identifies a particular variant of the resource version = a number or text that identifies the content of a particular instance of the resource
("content of" could be used with the others as well)
Ok, none of these feel absolutely right, but the attempt here is to avoid using "version" in the definition of "version".
version = a number or text that identifies a particular revision of the resource
We discussed this issue and the PR #1451 in tonight call.
We agreed that PR #1451 addresses the core of this issue "version should not be limited to numbers" replacing the current definition of dcat:version
The version number of a resource [[?PAV]].
with
The version indicator (name or identifier) of a resource`
Resolution https://www.w3.org/2022/02/08-dxwgdcat-minutes.html#r03 summarizes the result of tonight discussion on the argument that we shouldn't use 'version' to define 'version'.
The proposal is to keep "version" as using the word version in the normative part leaves open to other kinds of versions than revisions. And at this stage, we do not want to restrict the ontological commitment without full evidence that version = revision .
Does this sound reasonable to @agreiner and @kcoyle?
yep, looks fine to me.
In section 6.4.27, we have Property: version = "The version number of a resource", but we don't want to prescribe numbers only.