Crucial decisions we have to make soon (with current answers/thinking):

• How much support for Phones/PDAs?
  • Current answer: We are embarking a study of the Phone and PDA market as the key piece of the ChandlerEcosystem20030808 project. We will have of recommendations from this study.

• What is our web presentation/client strategy?
  • Current answer: Our document architecture should provide a web presentation/interaction solution.

• Is well integrated IM in or out?

• How much do we leverage the web for PIM data (from URL management to active polling of web data)?

• External file formats inline support?

• Do we support parcels/components NOT written in python?
  • Current answer: No, because of scope of project. Need to investigate SOAP and other more loosely coupled solutions for integration non-python components

• Do we provide automatic secure email?
  • Decision made: No automatic secure email in Canoga. Privacy is not a key push in Canoga. The payoff for this effort is uncertain and it doesn't reinforce other cool features we can provide.

Crucial areas of Chandler design we need to nail down

• What are recognizers? We need to design and spec this out so that it's do-able within Canoga timeframe (at least as a first version)
• Document architecture is still not well defined enough (but increasingly better defined).
  • What widgets are in the Document Architecture?
• We need to nail down UI design much more concretely

Architecture

• Selection of important external libraries/standards we will rely on may further constrain or change how we build Chandler. We need to nail down the important libraries soon:
  • HTML renderer/editor
  • email libraries
  • Spam filter
  • Spell checker
  • Sync with PDAs/Phones (should we use SyncML?)
  • How much of RDF do we support?
• Data framework is coming along well but requires more application validation
• Agent framework requires application validation

Current answer: we just hired a UI designer.

-- ChaoLam - 21 Aug 2003