Archive of UserLand's first discussion group, started October 5, 1998.

Next Generation Email

Author:Eric Soroos
Posted:9/21/1999; 9:01:02 PM
Topic:Next Generation Email
Msg #:11331
Prev/Next:11330 / 11332

The most recent AlertBox, from Jakob Nielson mentions that seemingly the only advance in the last 30 years of email has been the elminiation of the hop!hop!hop addressing requirements.

I've got an interest in email (hardly surprising to some of you out there.) It's my opinion that email reached it's peak of utility with respect to the client interfaces some time before netscape communicator and outlook express arrived on the scene. Email certainly was more useful when every piece of email in your inbox was from someone that you might want to write back. (pre spam, circa 1993. Yes Virginia, there was such a time.) For writing and editing email, (or any text doc for that matter), It's hard to beat the purpose built tools such as Eudora, Emacs, Alpha or BBedit. (Word is a different animal, useful for when you are working where wysiwymg is actually a possibility)

We've taken a pretty big step backward now, accepting web based email and the swiss army knife browser/news/email clients. The tradeoff in these less advanced UIs is that in some cases, we have powerful servers storing a lot of information for us that we can access from anywhere there's a net connection. They can do indexing, threading, and anything else that you might desire, while being accessible from anywhere. (generally the same things that a unix guru was doing 10 years ago with telnet and emacs, but now it's accessable to the general computer user.)

There are three or four ways to access email that seem to be prevalent or relevant today:

  1. POP3 - the old standard. It's not sexy, but it works
  2. IMAP - Complicated as all hell to do a good server, Very good over local net connections though. Solves the "Is the message on the client or the server" problem
  3. Webmail - Webmail is now nearly ubiquitous. Good access, but in general the user interface and responsiveness is miles behind what you can get on your local machine
  4. Mail, the next generation - Rich client, network storage.

Now we're getting down to the point of all this. finally

Where do we go on the next generation of email?

We've got some fantastically powerful tools. We have, or will be getting soon, bandwidth. We have an uprising saying that things are too difficult to use. So for the era of 24/7 connections, multiple machines (laptop, desktop, palm, cellphone), what should we do to email?

I'm thinking that an xml-rpc/SOAP type interface could be a good middle ground between POP3 simplicity and IMAP functionality.

I'm thinking store it on the server, cache it on the client if necessary. Don't ever delete it, if possible. (the deleting thing cuts out 1/2 of IMAP's complexity). (Keep in mind when reading that last statement that you're listening to someone who just this last week realized that he lost a business card with a webiste on it. So, when altavista.com failed, I grepped my web cache (not browser) logs for a small piece of text I knew was in the URL. Email is my memory hole.)

I'm thinking that important email deserves more attention than non-important email. And that the user shouldn't be burdened with dealing with email below some utility threshold.

I'm thinking that without backwards compatibility, it's useless.

So what should the next generation of email bring? How can we make this a bigger jump than just eliminating the hop routing of the early internet? What would make email a better tool for the user?

eric

ps. I'm trying out the Jakob Nielson way of writing for the web with this piece.

There are responses to this message:

This page was archived on 6/13/2001; 4:52:45 PM.

© Copyright 1998-2001 UserLand Software, Inc.