search

w3c / rdf-tests

Repository for the RDF Tests Community Group
w3c.github.io/rdf-tests
Other
42 stars 23 forks source link

Missing/wrong(?) status of tests in manifest file #137

Open kasei opened 2 months ago

kasei commented 2 months ago

The N-Triples 1.1 test manifest file uses rdft:approval for a number of tests, but many have no approval data. For the ones that do, there are 2 rdft:Approved and 20 rdft:Proposed. It seem based on the tests in https://www.w3.org/2013/N-TriplesTests/manifest.ttl that this situation has existed since the 1.1 standard (and the related implementation report includes data for the proposed tests).

Is there a known reason for the Proposed tests? I assume that anything marked approved or with no approval data was implicitly approved by the 1.1 WG and should be marked approved and given some dawgt:approvedBy data referencing the spec's moving to REC…?

I would guess that the "Proposed" tests are also mean to be approved? If there are concerns about simply marking them as such, maybe we can get the WG to take a quick look and bless the bulk approval of them.

kasei commented 2 months ago

(Wasn't sure if this issue belonged in the spec repo or the tests repo. Can move as appropriate.)

gkellogg commented 2 months ago

Probably should go in the rdf-tests repo. All the tests in the 1.1 directory should be considered approved, although there may be no corresponding WG approval. In the 1.2 directory, we haven’t taken any approval action. I would think that at some point we can make a resolution that all such tests are approved and some policy for approving tests that come in after that. Generally, we need some approval to merge, and that is probably adequate going forward.