Skip to content

Latest commit

 

History

History
177 lines (134 loc) · 5.26 KB

2024_09_16.md

File metadata and controls

177 lines (134 loc) · 5.26 KB

tinyButt 2024-09-16

(or 15th depending on your location)

Agenda

  • review last times actions
  • features needed : align on what we think current utc they are --> see below
  • back-casting : what do we need to do to deliver something useable

Actions from last time (Aug 1/2, 2024)

All

  • write up what your wishlist / requirement list for a scheduling app - aiming for short paragraph about what we are doing for a grant

Mix

  • send cal invites
  • talk with Sukie about Grantwriting

Christian

  • start git org
    • move the code repo into org
    • start Notes repo
    • start designs repo

Nanomonkey

  • plan grant funding with Mix

TinySched requirements

Mix's

  • gathering format
    • title
    • description
    • space-time coords [start, end]
    • host(s)
    • attendees
  • "follows"
    • we want to be able to organise-by-source
      • we want to support "curated programs for conference"
    • ideally there could be multiple "authors" for e.g. "dweb camp"
    • may be off-chain/ on-chain state
  • org format
    • who are valid authors for an organisation

Questions

  • is the follows a user visable feature?
    • Mix: not sure!

CFT

https://hackmd.io/sNJAFzf-TdKNO4IuN8wHKw?both

Minimal Data Structures needed for a tinySched MVP

  • structured text-only with color codes, transported over tinySSB or similar
    • non-text media only accessible and transported via URL and Internet (pictures, PDF, recordings)
    • color-codes ~= categories
  • usable for the main conference (dWeb camp) as well as its unconference (called "emergent day")
  • usable both for the organizers as well as the attendees
  • role support: conf owner designates curators who accept attendees and assign talks (and their speakers) to sessions with rooms
  • supports many views: from
    • "all talks" to
    • "this session's schedule" to
    • "my conf program"
  • bridge to the Web
    • some way to read tinySSB data on the web

Questions:

  1. does this need to be able to work in 2 directions? Web <--> TinySSB
    • implications about keys/ trust
  2. how do we gather the fields we need?
    • JSON?
    • implementation detail

Seven logical data structures/tables

a) conference (description, ref to owner, ref to curators)

  • organisation/ group?
  • could just be a network_id ??
    • might need meta data as well - roles, permissions etc. b) list of talks (description, ref to person)
  • rename to event?
  • note is could be a workshop c) list of places (description)
  • has resources (e.g. 14 seats, a whiteboard) d) list of sessions (title, time-span, ref to place, ref to steward)
  • this is a thematic parallel set of talks... a.k.a "stream"
  • Sched doesn't seem to have this: https://dwebcamp2024.sched.com/
  • is this the same as tag/ category? e) list of bookings (ref to session, ref to talk, time, color) - naming:
    • assignment?
    • event - joins: talks with places at a time/duration
    • and session
    • and color - A talk can have only one booking - Warn user if a place has two bookings at same time - question:
    • why is this different than a talk? => yes f) list of roles - these are assigned to a "person" (ref to userID, role={attendee/presenter/steward/curator}) - joins ?:
    • [conferenceId, userId, "curator"] - question:
    1. what level of granularity are we doing access control/ authorization
      • e.g. event wide? location level? g) list of attendance intents (ref to person, ref to booking) - for participants, way to track/ plan - for organisers, this might be useful for planning place

Views to be generated out of above data

a) all lists above, with selectable order (A-Z/Z-A, time+reversed, submission time, room size?) b) ordered talks of a given session c) ordered talks of a given room, per day d) full program roster, per day e) inconsistencies (room double bookings, sessions with holes) f) "my conference program" g) the dWeb camp's unconference is just another tinySched conference

Dimensioning -- not so tiny, after all

  • 1,000 persons (attendee/steward/speaker etc is defined in "roles")
  • 3,000 role entries (including versioning)
  • 500 talks (including versioning)
  • 100 places (incl...)
  • 100 sessions (incl...)
  • 1,000 bookings (incl..)
  • 10,000 attendance intents which translates to approx ... MBytes

Questions:

  1. what is the consequences of moving 15k messages?
    • a) what is the update latency?
    • b) does the system get blocked?
    • c) any trials are gonna require large data + node numbers

Discussion

  • text-only mode deliberately chosen as a demo for "Internet is an extension and if it works you have a rich experience and otherwise you still have basic conference orientation"
  • app can still cache pictures, but they could be out of date?
  • do we need "talk categories" for navigating/filtering a huge number of talks?
  • I tried avoiding thinking in terms of append-only logs - this is for later

Actions

Michael

  • book October meetup?

CFT

  • request more input from others

Nanomonkey

  • post up proposal work
  • set up grant conversation with Mix

Next Agenda

  • complete the walk-through of requirements, synthesis