Closed raffazizzi closed 1 year ago
@ebeshero unfortunately having both old and new classes at the same time would define the attribute twice, breaking the build/schema
Sure, an attribute can't be a member of multiple classes unless they are inherited... I guess I was wondering if the new version could be on dev branch safely, but probably not given the length of the deprecation period.
Really, I wish we did not have to handle this with deprecation, since it's a change involving class memberships and the improvements explain the usage of @calendar
so much more clearly! Just reviewing the decision for a moment, I suppose we are deprecating an overuse of @calendar
that has always been problematic but wasn't explained sufficiently. Of course the deprecation is simply to warn people this is going away on some elements.
If we want the improvements to be available as soon as possible, I don't like the persistence of @calendar
in the old class much at all.
Not sure if this is viable: What if we revised / updated classes now (no commenting out of what we intend to do), and for the deprecation period, we temporarily supply @calendar
on the elements where it is to be removed--just without its class and with the deprecation warning notice?
Moved calendar to its own class and deprecated on all att.datable members except
<date>
,<time>
,<origDate>
, and<docDate>
. Fixes #2045.I was not able to use
@validUntil
on@calendar
because it's not to exclude@calendar
from membership toatt.datable
on the four elements where it needs to remain. So there is a schematron check targeting all the elements it will be removed from.After the deprecation period, we will need to make
<date>
,<time>
,<origDate>
, and<docDate>
members of the newatt.calendarSystem
.If anyone has suggestions on how to do this better, please let me know!
Also made
<docDate>
member of att.datable. Fixes #2432.