Dashboard Brainstorm - Part II

For context, see notes for Dashboard Design Session Part I: http://wiki.osafoundation.org/bin/view/Journal/DashboardCollabDesign20051215

Last time, we identified the following stages for the lifecycle of information items and tasks

• Blackbox
• Research
• Waiting fors
• Actionable
• Done

We also talked about Timeframes associated with these items, essentially periods of time when we want to FOCUS on or be aware of these items in some way:

• Due dates
• Milestones
• Start thinking about dates

We also talked about distinguishing between PUSH and PULL items. Information items that we want the system to PUSH to us with reminders versus information that we just want to store so we can go back and retrieve it at some later date.

Design challenge How does Chandler help you decide and keep track of what items to FOCUS on at any given time?

• Brainstorm What questions do I ask myself? What questions do I need to answer in order to make progress on my information items and tasks?
• What can I do next?
• How long will it take?
• What data do I need to collect?
• What are my goals?
• What happens if I don't do this?
• Have priorties/dependencies changed?
• Is this relevant?
• Am I waiting on something?
• How long can I let this slide?
• Can I delegate?

We then revisited the list of "tasks on people's minds" from last time and asked the same questions again:

• What size truck does Jed need?
• Which branch of UHAUL?
• Alec found a company for dumpster.
• Who gets what Xmas gifts?
• Katie assigned a timeframe for her "20% dev time" task (need to talk about it at next staff meeting) and defined a next action: talk to someone at Google
• Bryan was waiting for Lori to get back to him about health benefits
• Jeffrey emailed somebody to get a friend's address
• Alec had to find an architect for his kitchen remodeling project
• Jeffrey needed to research what his tax bracket was and which tax shelters were appropriate?
• Jeffrey needed to define what his financial goals were.
• Bryan needed to find the appropriate time and context to read his healthcare plan material.

So what are the axes of metadata that people need to define in order to make progress on their information?

• Who does this relate to? in what capacity?
• Size of the task
• What is the next action?
• What supporting information do I need?
• What is the urgency of this item?
• What is the appropriate context for completing this task? (Interestingly, nobody brought this up.)

Then there is this entire category of information that will be hard for a PIM to "capture" because they are inherently "mushy" concepts that many people may not be able to or want to articulate explicitly. The design challenge here is to figure out implicit or indirect ways the UI can encourage people to address and capture some of these questions without making it so literal that people get turned off by it.

• What are my goals?
• What happens if I don't do this?
• Have priorties/dependencies changed?
• Is this relevant?

• Possible solutions:
• Helping people keep track of tasks that may no longer be relevant...see Andi's Gold Box idea.

Brainstorm: Different ways to attach metadata to items in 0.7 and possibly beyond

Jeffrey suggested a palette of "attribute value" stickers that you can drag and drop onto various target zones:

• item detail views, as a way of adding metadata to items
• items in the summary pane, as a way of adding metadata to items
• search box as a way of searching on that attribute
• sidebar as a way of creating a persistent of view of data (aka collection) based on that attribute

Jeffrey suggested in particular, a People palette, perhaps a "related" Contacts pane for each collection in the sidebar.

First apply the sticker, worry about which specific attribute it applies to later. For example, you could drop a "person sticker: Lou-Ellen" on an item and it would automatically render itself as a Who: Lou-Ellen attribute. You could come back later to specify "in what capacity" Lou-Ellen is related to this item (ie. Author, Invitee, For, Owner, etc).

Other GUI ideas for manipulating Chandler's query-based views of data:

• Drag and Drop search terms from the search box into the sidebar
• Drag and Drop attributes from the detail view direction into the search box
• Drag and Drop attributes from the detail view directly into the sidebar to define new collections
• Drag and Drop attributes from the sidebar to the detail view to label items

Of course all of these actions would also be duplicable via context menus and menu items.

Solutions we might implement in 0.7

Labeling and tagging through the CLI-for-dummies UI proposed a couple of weeks ago.

Generic "label" column in the summary table view. That way we can accommodate lots of different attributes and lots of user-defined attributes, without having to force the user to spawn a whole new column for each new attribute.

Viewing "supporting material" in groups

Alec suggested having the ability to view clusters of items in the summary pane (a la thread view in mail clients) ie:

• next actions
• related resources and other related information items

Keeping track of old and forgotten items...Cleaning out old and forgotten items

Andi suggested an Amazon Gold Box feature that queues up "old and forgotten" information items that you entered a long time ago, but haven't touched in a long time.

Katie followed up with a "If you liked this item, you might like these items" feature.