Skip to content

Latest commit

 

History

History
352 lines (214 loc) · 10.3 KB

2018-05-01-meeting-and-SAAM-paper-planning.md

File metadata and controls

352 lines (214 loc) · 10.3 KB

Project-meeting 01 May 2018

This document: 2018-05-01

Present: KP, DW, DDeR, JP, GK

Actually, two meetings:

  • first without DDeR, which focused on reviewing progress to date
  • the second was focused on preparing a SAAM paper

Agenda

  • Confirm date for next meeting
    • 2018-05-15 (?)
  • Review actions; note any issues or blockers
  • Review software progress
  • Planning for SAAM paper
  • Review and revise timeline

Resolutions

Previous actions carried forward

  • 20180322.4 JP Add notes about proposal for NiN export triggers to GitHub - this isn't so urgent, may be allowed to drift
  • 20180322.8 JP Multiple/additional working set import and basic display
  • 20180328.6 JP Fragment assembly / internal score generation (UI and logic)
  • 20180328.7 JP MEI document generation from internal score
  • 20180328.8 JP Audio generation from MEI score (0.5d)
    • looked at previous working code; con luded that it should port OK.
    • blocked on MEI document generation (20180328.7)
  • 20180413.3 JP authenticated access to LDP
    • not an immediate requirement, but should be addressed before allowing external access to remixer. Will probably be provided via HTTP server proxy.

New actions

Most actions were not formally recorded. Those that were include:

  • 20180501.1 GK propose interface for auto-generated containers

    • Rough sketch scanned and emailed 2018-05-01

  • 20180501.2 GK Circulate terminology proposal

    • Emailed 2018-05-02

  • 20180501.? ?? ???

Next meeting

Next meeting: 2018-05-15 (time TBD)

Draft agenda next meeting:

@@TBD@@

  • Set date for next meeting
  • Review agenda
  • Review actions; note any issues or blockers
  • Review development activities
  • Review paper preparation activities
  • ???
  • ???
  • Review and revise timeline

Review actions

Previous meeting: 2018-04-23

Previous actions carried forward

  • 20180322.4 JP: Add notes about proposal for NiN export triggers to GitHub - this isn't so urgent, may be allowed to drift
  • 20180322.8 JP: Multiple/additional working set import and basic display - No activity, ongoing, not a vast amount of work to complete, but not urgent - deferred until after demo rehearsal
  • 20180328.6 JP Fragment assembly / internal score generation (UI and logic) (1.5d)
    • Basic logic is done, but UI still needs work
  • 20180328.7 JP MEI document generation from internal score
    • Not complete
  • 20180328.8 JP Audio generation from MEI score (0.5d)
    • Not started
  • 20180413.3 JP authenticated access to LDP
    • not an immediate requirement, but should be addressed before allowing external access to remixer. Will probably be provided via HTTP server proxy.

Actions from last meeting

  • 20180423.1 GK post meeting notes
  • 20180423.2 JP complete UI for remixer and fregment assembly (0.5d)
    • UI still needs completing
  • 20180423.3 JP complete check MEI output and functionality (0.5d)
    • Incomplete
  • 20180423.4 JP complete and test audio generation from MEI (0.5d)
    • No progress
  • 20180423.7 JP (arising from SAAM meeting) add simple new match services to demo system.
    • Not started

Progress on these items was impeded by VM storage problems, and some tasks (esp. audio playback interactions) proving trickier than expected.

Review software progress and blockers

Integration still coming together; also time needed to sort server after VM storage failure.

UI for remixer and fregment assembly:

  • Outstanding: UI layout improvements
  • score cursor (insert point) problems
  • no audio playback yet.

MEI output and functionality:

  • Testing and checking still to do.

Audio generation from MEI:

  • Outstanding: playback of snippets (problems with audio playback interactions)
  • Outstanding: playback of whole thing (from MEI)

Not started on new match services.

Review requirements for SAAM paper

Front-end:

  • Don't currently have UI to select working set container
    • Canned working set isn't enough
    • e.g. provide URI entry box, and more
  • Needs tio be clear from display what we're looking at
    • working set?
    • candidate fragments for selection?
    • etc.
  • Need to get CSS working for columnar display of selection/options
  • Show fragment-related metadata (to underscore semantics)

Back end:

  • LDP containers holding anotations
  • Separate service to generate compatible keys container
    • code to create annotations works
    • is still manually triggered, which is OK for now

Terminology (to clarify and be consistent with narrative):

Need to generate real data for cmpeling narrative

Selection of additional match criteria

what are we trying to get over as contribution of paper?

  • use of end-to-end semantics
  • need to play up this aspect; e.g.
    • semantics from NiN
    • key, mode pulled out from thios and used in assembly process

Two issues

  • multi-part (multi-voice)
  • quantified constraints (NOT NEEDED FOR NOW)

NEED (for multi-part)

  • length match (i.e. same-length)

WANT

  • key selection (lockable across time and/or parts)

EASY

  • scale mapping (trivally similar to key)
  • algorithm mapping (trivally similar to key)

Also to consider:

  • external description of ensemble; map to instruments (parts)
  • Instrument selection (fixed for part over duration of composition)

ACTION: GK think about discovery of interface for auto-generated containers within match service.

ACTION: GK send email with proposal for naming components

Additional notes for paper

Separation of concerns; process structure and data specifics (e.g. FP combinators). Uniform API / REST. HO fns for quantified matching?

Using linked data instead of new services (uniform interface + representation)

Planning for SAAM paper

Excerpts from Kevin's email:

This is project SOFA (SOFA Ontological Fragment Assembler) (?)

  • Graham is away next week. It is best to assume his contributions will be writing up of the architecture/figures rather than implementation. If there are any questions to him on the architecture or implementation, the sooner the better this week (seeAlso body vs. target discussion).
  • John will continue with implementation. We've cleared some of David's other project commitments until after the SAAM deadline so he can help with implementation.
  • By default David's implementation assistance would seem best applied to UI elements manipulating the MEI/SVG, and those parts of the services which are likely to be adopted into MELD-web-services in the future (and identifying which those parts are).
  • An immediate action for John and David is, therefore, to work out the division of implementation subtasks.
  • We should target the end of next week (11th) for wrapping up the additional implementation. This clears the final week for writing.
  • KP will write up an outline paper structure based on our discussions and circulate.
  • Graham and David are meeting Thursday (2018-05-03) to discuss arch issues.

Next steps and timeline

Finish off essential parts of remixer demo.

Impement new features for SAAM paper.

SAAM paper writing.

JP time over next 3 weeks - no breaks planned, so 12 days to next meeting.

(See also notes above from KP email)

Next week timeline

????

Overall rough timeline

  • 2018-05-01 to 2018-05-11

    Basic functionality for SAAM paper

  • 2018-05-01 to 2018-05-18

    SAAM paper writing

  • 2018-05-01 to ???

    Complete core UI for remixer

    • finish fragment assembly / score generation Complete MEI document generation ("score") Complete audio playback

  • Deferred to later

    DMO Export logic and UI

    • may drop this unless there a clear requirement arises

    Multiple working set import and basic display (0.5d)