ZedAI telcon 20091102
From zedwiki
Note: this meeting will be held at 2100h UTC. Find the meeting time in your geographical location.
Contents |
Present
Markus (scribe), Sam, James, Dennis, Josh, Boris, Matt
Regrets
Previous Telcon
Action Items
New
- Matt: Find out: could we use DC for the always required set? Propose a map.
- Markus spell out how to do external meta linking
- Markus: set up Boris's object documentation
- Markus: book Matt and Sam this week for lists
- Sam - write a sample of content selection feature - multilingual document - declaration of human rights
- Boris/Markus - ask ViewPlus et al about the SVG namespace-inject issue [ONGOING]
- Kenny - own the page continuator issue
- Markus Matt, James, to create a wiki formulating an out-of-line metadata proposal. Then circulate to interested parties (IFLA group, WIPO, board, etc) [ONGOING]
- Markus implement role module [ONGOING]
- Boris: help deal with svg integration, help with object documentation and samples, address issue 9 [ONGOING]
- Markus: describe our use of xlink:simpletype as an implicit default a la docbook5
- Dennis: solve Issue 60
- Dennis: solve Issue 61
- Dennis: solve Issue 46
- Josh: solve Issue 50
- Markus to add spec prose on cascading level for inline styles: "For the purposes of the CSS cascade, these values are considered author-level"
- James: make sure the spec no longer claims to reserve the z3986-profile and feature values; instead define above rdfa solution [DONE, reservation for PLS remains]
Agenda/Minutes
Reminder: ZedAI_Iteration4_Report#Deliverables_for_the_next_iteration deliverables for the next iteration
On Math
Dennis: Final draft is out. Now is the time to look at issue 60.
Dennis: on simplification (Issue 46): to do simple inline math without using mathml
Metadata
Status update - topic page: ZedAI_Meta_Data_-_OutOfLine_Proposal
James: inline metadata only focused on the zedai document itself (creator (name, location), unique identifier)
Propose that required stuff be as minimal as possible. No need for catalogue searching etc here.
Discussion on keeping the simple use cases simple; adding some items (title and author) to retain backwards compatibility with prior versions of the standard.
Matt: could we not do this by saying that a dublin core meta set should be attached? James, yes we could! If required, kind of inclined to put in the document itself.
Matt: for the spec, maybe not make req to include DC; but make recommendation to also attach dublin core.
Summary
- always required:
- creators name
- creators location
- unique identifier
- version number
- release date
- recommended for backwards compatibility, some items from Dublin Core
- dc:title ()
- dc:creator (referring to the author of the original document)
- TODO some other DC:s
Remaining issues: what do properties do we use for the top required set?
James: with these five items; are there clear unambiguous mapping in Dublin Core? Markus: creator is ambiguous. James: if we use DC its consistent.
@Matt: Find out: could we use DC for the always required set? Propose a map.
James: in 2.02 and z2005, dc:publisher meant RFB&D, TPB etc. We could use that for the creator field in the always required set. Same for dc:identifier.
@Markus spell out how to do external meta linking
Lists
Status update - topic page: ZedAI_Lists_Proposal
- Thoughts/input? Next steps
Markus, Matt, Sam book new date. Markus send proposal.
SVG and Object
(See overview in previous minutes: ZedAI_telcon_20091019#SVG_and_Object)
- Update, budding proposal from Boris and Markus:
On intertwinings re Object/fallback text/annotation
Solution: we need to point out that fallback text and annos have different behaviors and slightly overlapping semantics. If the author provides an annotion (with a specific role?), then fallback text is not requried. At least one of them is required, but both can be provided
James: NIMAC books with a heading, decorative graphic layout, heavily styled text. Here, object is a good way of implementing it as compared to current solutions, to express equivalence.
Annotations: we have no role to state "this is really the same thing"; so heres where object fallback comes in.
@boris: in documenting role, make this distinction.
Object Mimetypes
- base profile defines a limited list of allowed object mimetypes. (New: a Profile must do this)
- a new feature called "arbitrary object inclusion" enables documents to contain types beyond those listed in the above list. These "arbitrary objects" should have object children (entries from the approved list) where applicable, but else fall back to the general rule of each object having at least fallback text or an annotation
James: alternative: define a minimal list that all PA's must support - then anyone can reference any object as long as fallbacks from the minimal list are provided. Boris: through innermost text being required, we've already got the fallback.
SVG Integration
(discussions still ongoing with viewplus)
- make a SVG a feature, just like MathML
- ?allow only one or both of 1.1 and 1.2 Tiny?
- ?allow both CDI and CDR? (compare MathML, which atm in the draft is assumed to be used CDI)
- Host language inline text injection needed - this more straightforward in a CDI context
- in case of CDR - feature would on activation "contribute" its mimetype to object
Periodicals profile
Reminder: periodicals profile will be removed at the end of this iteration.
(Suggestion on considering a magazine/journals profile; may do a VF2F on this topic in the coming iteration)
Next call
16 November
