Too much coffee, and too many huge unknowns. It’s amazing how fast this stuff piles up.

Mail backlog. I need a (much) better approach to managing conversations – something that puts every communication with every person on a timeline, so I can see what’s going on with that person both in terms of my communications with them, but also grabbing their tweets and RSS feeds so I can see how my stream of comms with a person is intersecting with their lifestream and possibly calendar. I want to be able to visualize/contextualize somebody’s life right in my “email” client please. Somebody want to write that with me?

OmniMail? God knows OmniGraffle is worth about 10 points of IQ. Imagine what a mail client could do if we started from scratch assuming a lifestreaming environment!

I think it’s the kind of thing which might be an excellent use of the new SVG rendering capabilities of modern browsers. LifeMail / VisualMail / FocusMail type things.

Also mail that indicates in some kind of coherent way when and if a reply is needed. That might be as simple as a social convention in titles: [r10d] (please reply within 10 days) which would be parsed by your mail program. Or god, even another email header – how’s that for old school? Remember when you could add arbitrary headers in your mail client?

We really just haven’t even started on integrating email into the social web.

The other thing we haven’t got a good handle on is how close or far things are apart. A map doesn’t really tell you, intuitively, if you are close to something because it’s defined in distance, and what actually mattes is not distance, it’s travel time. Finsbury Park is close to Brixton. Sydenham is close to Hoxton. Now consider two people in London, or three, or four.

Where’s the piece of software where I can punch in all of our locations, and it will generate a list of cafes which have roughly the same travel time for all parties? Now can we do the same thing, but not in time terms but in cost terms for international meetings?

See what I’m saying?

We haven’t even begun to write social software yet. There’s 20 or 30 years of tech development just from pushing the existing social software tools we have to their logical integrated-with-the-real-world conclusions. I don’t talk about software much these days, I got distracted, but I used to be quite good at it. In fact, I used to really enjoy it.

Anyway, final thought for the night – Processing is the right language for data visualization. Now where’s the mail client that joins latent semantic analysis to Processing-style visualization to show me my emails clustered by emergent topic (“Meeting on Tuesday”, “What do we do about buying a new server?”) with accurate maps of the people who’s emails are typically more important, and some keyword-based “contextual over-rides” to allow trusted people to cut through my filters directly to me?

We haven’t finished email yet. We just need next-generation email interfaces.