Ripe for Remixing
Nat Torkington
Jan. 26, 2005 11:54 AM
Permalink
The iTunes music store shakes up "the album". Print-on-demand and Safari U shake up "the book". I've got a hit-list for the next things that need to be remixed. Top of that list is email.
Email
When I look at my mail client, it hasn't changed substantially from 1988. There have been efforts in academia, business, and open source to improve email clients, but most "improvements" have been incremental. While this was going on, though, the volume of both wanted and unwanted mail we recieve has dramatically increased. The demand for better mail facilities far outstrips the supply of improvements.
The only people I've seen looking hard at throwing out the old type of email clients are IBM. Esther Dyson wrote a report on the state of email that makes great reading. I've been watching ReMail for a while, and loving what I see. It's the sort of research that's begging to making it into a product.
Nat Friedman, at Foo Camp, talked about how people relate to mail. You only view your mail in terms of people (who sent it, who's on the CC: line), time, or project. The project is the bit that folders are supposed to solve, but they're horrible at it. Am I really supposed to use a Victorian filing cabinet model for my mail, assigning each message to its position in a hierarchy of folders?
Prompted by Rael, I've been tipping all my mail into only two folders: "saved messages" and "paperwork" (receipts, which are critical to fidn quickly). When I need to find something, I search my inbox. I'm struggling under Mail.app's limited searching (sometimes I want to refine a search by using multiple criteria, which I won't be able to do until Tiger). However, it's a definitely improvement over the two step search I had to do before: "where would I have filed this message, what criteria will work best to find it?"
It's interesting that I never uniquely find the message with a search. That is, a search simply narrows the range of options down to where I can recognize the message from a list. So if I'm trying to find the message where I explain OSCON committee duties, I'll search my saved messages for mail to the committee list and then order by date to find the earliest message. There's no search I could make to uniquely find that message, and that's fine. Yahoo Search doesn't find the page I want, it clears the irrelevant pages away so I can find the page I want.
The logical alternative to folders is tagging. This will solve the problem of one message talking about two different things: which folder to file it in, or do I copy it to file it in both places? Tagging and "folksonomies" (I do loathe that term) are all the rage these days, so hopefully it won't be long before some bright spark adds this to a mainstream mail client (preferably mine :-).
But filing isn't the biggest gap. The problem is that changing focus in your client is intentional, a multi-stage process: hit search, type your terms, get your list, sort by another criteria. The type of focus I want is project-based: I want to be able to define people and tags and terms that make up a project, and only view my mailbox in terms of those tags. Then, as I do my reading and writing of mail, the client adapts. If I'm sending mail to Cory about patents, behind my composition window I should only see messages about patents and messages to Cory. When I send the mail, it can revert back to my inbox view.
There's no sense at all of the mail client actively helping me. I can initiate searches, I can supply the keywords, and I can file my messages. About the only intelligence my mail program has is spam filtering, and it's the feature of my mail client that I love the most.
The mail client has ossified the way the music business and the web browser did before iTunes and Firefox shook things up. I say it's first in line when the revolution comes.
Buddies
Another area that needs a shakeup, though, is the whole concept of my friends. I'm not talking all this forced social network nonsense where I have to try and force my friendships and acquaintanceships into a mediocre programmer's data model and a didine capitalist's business model. I'm talking about the address book and the buddy list, where my data exists on my machine. How can I organize this, or better yet have the network effect kick in and let groups determine groups?
Once I have a cloud (and it can be a fuzzy cloud, the way tagging provides you with fuzzy folders) of related people that I can point to, I can use it with mail, web, appointments, IM, etc. Think of it as an internalized Outlook. Inlook, perhaps? I don't want to have to commit my life to a Microsoft server to reap those benefits--there's no reason we can't distribute the knowledge of groups and associations to get similar effects (simple scheduling, notification) and extend the benefits to other filters and views of data in other applications.
The buddy list is slowly accreting information and slowly interoperating with the address book. I'm even seeing printed business cards with AIM handles. The buddy list is just a specialized address book, and I argue that it's time to stop thinking of them as separate. If "but I can't do ___ with the address book/buddy list" springs to your lips, then think about what it would mean to do that. Bring the social categorization to the business categories of the address book. Bring the rich address book data and integration to the buddy list.
The network effects here are particularly interesting. Why should everyone working on the "Advanced Smeg Hacking" book have to create their own informal list of people related to the book, so that their mail client can do smart filtering? If I've added you to a list, I should be able to make it public and now it shows up as a possible list for your programs to use. How to do that? LOAF? Ping? Sxip? Lid? I don't know, Digital Identity is still beyond me, even after two conferences.
Where's My Singularity Already?
All this, of course, falls under Nat Friedman's Dashboard and Beagle visions, though they are both add-on applications. In other words, they still think of filtering and contextual displays as a separate app rather than something built into every app. Of course, some people think the very concept of an app is outdated--eventually one program will do everything. Mozilla's heading there, but it still has a very long way to go. I see a lot of benefit to the focus of an app--let's fix mail handling, then fix web browsing, then fix IM, then .... Even the Mozilla folks partition their efforts--they're not releasing Frankenbrowsers that do everything poorly. Their lead product is focused on being a great web browser, and they've got separate development on the other application areas.
Chandler has this kind of all-encompassing vision. They're building a Personal Information Manager, which shares, slices, and dices. It's a dessert wax and a platform too! They're trying to get a 2005 dogfood release ("Kibble"), which might shake things up. They've also been bitten by their initial attempt to solve all problems in one app. They're instead identifying a core to work on and will build out from that. I'm still not convinced that Borg.app is the right approach, but I'll withhold judgement until I see Kibble.
What I can't understand is why we're not seeing this kind of innovation from Microsoft or Apple. Sure, Mail.app in Tiger adds better search and smart folders. But it's an incremental improvement. Who at the major vendors is remixing your inbox? Who's remixing your rolodex?
Perhaps this is an opportunity for open source and small software companies, much like IE's dominance and then stagnation was for Firefox and Mozilla. The only innovation in the browser market has come from smaller players, while IE rotted on the vine. I've seen some steps but I'd love to see more.
--Nat
Nat Torkington
is conference planner for the Open Source Convention, OSCON Europe, and other O'Reilly conferences. He was project manager for Perl 6, is on the board of The Perl Foundation, and is a frequent speaker on open source topics. He cowrote the bestselling Perl Cookbook.
Showing messages 1 through 2 of 2.
-
Nathan Torkington is still puzzled
2005-01-31 10:58:35
Johannes-Ernst
[View]
-
worth looking at.
2005-01-26 12:26:43
monkeyt
[View]
|
Showing messages 1 through 2 of 2.
|
Return to weblogs.oreilly.com.
Weblog authors are solely responsible for the content
and accuracy of their weblogs, including opinions they
express, and O'Reilly Media, Inc., disclaims any and
all liabililty for that content, its accuracy, and
opinions it may contain.
This work is licensed under a
Creative Commons License.
|
I don't know, Digital Identity is still beyond me, even after two conferences.I sympathize, but I think the answer simpler than he thinks: don't look at Digital Identity as a product, but look at it as a feature... more...