anonymous morphed into parisse

anonymous morphed into Monica Palmirani

anonymous morphed into Michael Neuren

anonymous morphed into Catherine Tabone

anonymous morphed into Grant Vergottini

Fabio Vitali: hello

Fabio Vitali: joining in right now

Thomas Cucchietti: Hello everybody

Monica Palmirani: hello online

Fabio Vitali: Can you hear me?

anonymous morphed into Thomas Francart

Grant Vergottini: Sorry, I did not have a chance

anonymous morphed into Jonathan Germann

Fabio Vitali: Looking at line 103 vs line 138 of the Google sheet document Fabio Vitali: Do we need many different designations or just one (subdivision), whose values are the subdivision names of the citation?

Fabio Vitali: Fabio is afraid of the added complexity of the first solution

Fabio Vitali: John agrees and see it would lead to unwieldy long lists Fabio Vitali: but no one has definite opinions

Fabio Vitali: Thomas F: We need to keep track of the subdivision types, not just arbitrary names, such as subdivsion A, subdivision B, etc.

John Joergensen: My solution is to take the specific and special elements that are in citations and abstract them. A more general designation that applies to many kinds of documents.

John Heywood: I came in late...where may I find the spreadsheet?

Fabio Vitali: leave names in the actual feature tagging, but outside of the conceptual modeling we are creating Fabio Vitali: John: https://docs.google.com/spreadsheets/d/1e-Uzk4b0ms60oCIupAHrkjufFY8f4KMIygOdXLuG9gE/edit?usp=sharing

John Heywood: Thanks!

John Joergensen: Point taken, but in that case, a possible solution is to keep a more general list of designations, but work with the ORDER of presentation, the syntax that we use, so that the order of the subdivisions are interpretable for that specific document for any resolver that resolves that set of documents.

Monica Palmirani: The hierarchical order is not always known. The order of presentation is important as well as the interpretation ordered by hierarchical principle. We need two annotation for both.

Fabio Vitali: Monica: I do not understand John Joergensen: Remember, we are writing a standard that will make it possible for resolvers to resolve. That means we don't need to completely disambiguate for all documents, but provide enough resources (tags and syntax rules) to allow a resolver programmer to write rules for any given set of documents. We need to balance between covering all possibilities (where the resolver programmer does no work), and a standard that is rich enough to make a programmer's work *possible*. We only need possible.

Fabio Vitali: John, I agree

John Heywood: I also agree.

Fabio Vitali: Open discussion item: Whether to leave one term for subdivision and put the local names in the values vs. have a feature name for every subdivision. Examples and counterexamples are welcome, the matter will be decided next conf call Catherine Tabone: For example of subdivision citation mark-up see annotations in http://www.legislation.gov.uk/ukpga/1998/42/introduction

Fabio Vitali: Fabio's homework: fix function so that co-existence of feature name in source and interpretation frames do not cause troubles Catherine Tabone: You can see underlying mark-up at http://www.legislation.gov.uk/ukpga/1998/42/introduction/data.xml