Shoehorns and fossils

[Jesper]

Good day, list.

I think I have a good idea of what's wrong with today's email clients,
and it's not about buttons or slowness or database formats, although
they play a part as well.

The two traits that sink M.S. Generic Mail Client are:

* Shoehorning. I was actually going to post something very similar to
Brent's topic earlier. It's not just a question of spam vs ham. You
have dumb links from family members; you have files sent across
computers because it's 2010 and this is the best we can do (sob); you
have deep insightful existentialist discussions with friends; you have
a link to that YouTube video; you have discussion lists; you have
support mail; you have love letters; you have automated mails from
various services (moderate comments, friend requests); you have
business mailings. And *then* you have spam. This is the scale of
human communication. If I'd classify the entirety of spoken
communication as "annoying pushy mobile phone and carrier contract
salesman in the mall" and "other", you'd want to have me committed,
but this is exactly what we're doing to email.

   Shoehorning is a bit harsh if you look at it from the wrong angle.
Mail clients didn't epically fail from the outset by only recognizing
one type of mail. There are many axes along which to measure the type
of mail. But they have failed in-so-far that they haven't adapted.
It's not that mail clients shoehorn. It's that their feature set and
the ways of organization that they make available to you force *you*
to shoehorn.

* Fossils. Mail clients largely haven't evolved that much. Even when
Apple does it, it's been a question of improving the presentation of
what you write. (Mail's Stationery makes me think of a web site with a
link to a PDF product brochure in that I'm not sure this is the wisest
way of them to spend their time.)

The only big thing that has happened to email in the past decade was
Gmail. Gmail made some good choices, like focusing on threads instead
of individual messages, including your part in the mix and trying to
hide quoted parts to cut away redundancy.

But Gmail doesn't fundamentally fix anything. It's a nice
implementation of these ideas and it's a good case study for how far a
web app can go, but the rest of its functionality is *mostly* the
essentials from a good mail client.

I hope that the aim of this list, although we're mostly developers,
will be high: to solve the problem. Not to make a faster horse, but to
figure out how we really want to deal with email, and then implement
something that will work in that way. Hopefully in a technically
advantageous way, naturally, with background work in GCD, a flexible
storage model and suitably customizable.

/Jesper
(the "waffle" guy)