Re: Another E-Mail VOTE (again with Monday deadline)

Jos:
>As developer:
>still haven't seen independent evidence
>that miscellaneous-010 is OK as test case
>but also no evidence that it is wrong
>(and that since many weeks)

Each of the tests with one pass represents a difficult call ...

but it seems that the "let's wait for developer feedback" period is over,
and we have to make the best choice we can with the evidence we have.

There are 12 tests with one pass. I suggested obsoleting 3, approving 4, and
approving 5 as extra credit.

The choice between approving and obsoleting often depended on the interest
shown in the test by WG members and others during CR. The more interesting
tests have a value, which leads to suggesting approving these tests, on the
evidence that we have. As indicated in the proposal misc-010 misc-011 and
dl-209 are the highest risk items - even if we end up with an erratum for
each (which would be the worst case scenario) that is not too bad ... [I do
not believe these tests to be wrong, but Jos is right to highlight the lack
of independent evidence, for the misc ones - we have fixed earlier problems
reported with the syntax of misc-010 and misc-011]. The value of the
misc-010 and misc-011 tests is that they show other uses of the wine and
food ontology, not merely consistency.
(The fourth of these that I suggest approving, I5.3-014, can be checked by
hand, I am happy to produce a proof if anyone wants)

[I can give further justification for the five extra credit proposals if
anyone wants]

Jeremy

Received on Friday, 28 November 2003 09:38:40 UTC