# DC TAP meeting Thursday, May 12, 2022

**Zoom link:** https://us02web.zoom.us/j/85164307523?pwd=QTMybkFlSTJoUHA3cHp0NkZkU1ZZdz09

**HackMD link:** https://hackmd.io/kGU-JVISQQGQWHb7fR__SA

**Time:** 15:00 UTC ([check time](https://www.timeanddate.com/worldclock/fixedtime.html?msg=DC+TAP&iso=20220512T15&p1=%3A&ah=1))

## Participants
* Karen, Phil, Tom, Nishad, John

## Announcements
* Tom: continued work on python program

## Agenda

* [DC Tabular Application Profiles (DC TAP) - Primer](/DErWH403RaWiBippMFosaw) questions
* [Re-written intro](/H8rRdV4LSNeMAcdvCEJskA) (see [comparison](https://docs.google.com/document/d/1NuVaaB6dfVE9b_n1jFKpa6pT_XZlAXaCTJaDhCF3INY/edit))
* Should we move cardinality to the end of the statement section?
* "case insensitive" - maybe make this just a suggestion
* Multiple values in cells and close [#19](https://github.com/dcmi/dctap/issues/19)?
* defining statement constraint and constraints on statements

## Minutes

KC will begin editing Primer again. Will provide diffs.

### Introductory wording

Comments from Tom.
* data function -> drop. end with application
* other wording accepted
* understandable by people who aren't coders - "don't know schema languages" - but if don't know schema languages don't know what a schema language is. Phil: not understandable by all members of metadata community
* table-based APs - folks don't use "tabular" - "APs in the form of tables"

Other edits
* describing more than one thing / defining - Phil: can do both, can define something that exists, but in other cases you are defining something

Phil to do extension paragraph for primer.

phil: value and link is ambiguous. "Value of a statement may be a link or may be text"

kc: alphanumeric string? characters that represent strings or numbers.

john: problem of what is the statement? value *in* the statement.

phil: just say "the value"

phil comment: profile defines rules and provides further information...

"property must have been previously defined in a metadata vocabulary"

phil: metadata can have strings or dates or numbers and the profile says which you use.

phil: a profile might specify that values of some property may have to be a specific type of date.

kc: two questions:
1. Which columns in the TAP make up the statement. What does cardinality refer to? Not just the property, but is on the whole statement. Tom: put cardinality down below.

Tom: "Statement constraints" is ambiguous because we talk about datatypes, etc. as constraints. prefer: individual columns are constraints. Set of statement constraints? awkward. Use *template*.

phil: style guide : constraints on statements are the rows, individual constraints are the columns. But not every column is a constraint. "constraints on statements are *in* the rows"

phil: is constraint the total, or are the elements the constraints? I like that constraints are the totality of the row

tom: call the set a *statement constraint* that aligns with ShEx

kc: statements are in the metadata; constraints are in the TAP

tom: call them just statement constraints

kc: use constraints on statements in the document

john: confusing to call set of columns a statement constraint. we need a word for an entire row. not all of them are constraints, e.g. labels

tom: *template* doesn't appy that everything in it is a constraint

phil: and I like patterns

kc: individual things are rules and the overall is a pattern, like Tom's template

phil: pattern is in the primer, but one of the value constraints can be a pattern

john: one confusing area is "set of statement constraints" - is a set a row or the whole thing? We really need a word for a row

tom: set of statement patterns?

nishad: ap is not just patterns but is also definitions. template is better. pattern doesn't distinguish AP from a constraint language. AP explains the application of the patterns; explains the local decisions

kc: so if we call it a template and we define template as being all of the local choices about the metadata? A TAP might not include much about local semantics. Is it still a template?

john: statement configuration? "local choices" - implies an action, making a choice

kc: looking at constraint heading in primer: call it "Statement templates". Cardinality can move down, note is on its own. Not bound to statement template

phil: I argue that we shouldn't have statements about the shapes on the same rows as the statement templates. Note is just about statement template. I've done separate tables for shapes

kc: this needs a github issue

kc: where to put cardinality?

phil: fine where it is

kc: need to be sure that cardinality refers to entire statement unit - clarify in the text

### Tom's program changes
See email: https://lists.dublincore.org/pipermail/application-profiles-ig/2022-May/000659.html