by J.D. Falk
Director of Product Strategy, Receiver Services
What’s that we see, waving through the raindrops? Isn’t email supposed to be dead? You already know I’m going to say no; as usual, once you see past the refraction and the rainbows, reality is somewhat more complicated.
The recent, ongoing launch of Google Wave has almost everything we’ve come to expect. It begins with a slow roll-out, with people begging for invitations. Then comes the headlines proclaiming the death of email, often based on nothing more than a short preview video and someone else’s interview with Wave’s creators. This all leads to gigantic, Google-sized expectations. But with Wave those expectations have yet to be met; It is either such a gigantic paradigm shift that most of us can’t yet comprehend the enormity of their genius, or it’s an incomplete product that shouldn’t have launched until there was something more to show off than a Google-y user interface. Either way, Wave appears to have crested quickly, and we’re left waiting and wondering.
Then the Mozilla Thunderbird team filled that void by introducing a new concept they call Raindrop. This, like Wave, is an attempt to reconceptualize the way we interact with messages. The difference is that instead of creating an entirely new kind of message, they’re starting with the messages we already interact with: email, social status updates, et cetera; surely they’ll be adding IM, SMS, voicemail transcripts, and so forth.
Both Raindrop and Wave have the potential to change (and hopefully improve) the way we communicate online. Everyone who works in email, and in messaging in general, will be watching and blogging and exchanging lots of email about it.
Thing is, the competition happening here isn’t just between Google and Mozilla. It’s between two entirely different ways of thinking about innovation: starting anew, or building on established successes. It’s also between two entirely different ways of thinking about communication: closed, where only a few can participate — or open for all to interact.
When the internet email protocols were invented, the network — the whole idea of different computers talking to each other — was still a new concept. Trading data between computers didn’t just require finding a cable, or importing from one version of Word into another; often the basic method of constructing and storing data was based on entirely distinct mathematical principles. This stuff was really hard, and didn’t become easy for another twenty years.
In order to get past these differences, internet email had to be both simple and flexible. Obviously it couldn’t assume that everyone is using the same operating system; it couldn’t even assume that everyone’s byte is made up of exactly eight bits. It couldn’t (and still doesn’t) assume that every computer is permanently connected to a network, or that every network is connected directly to every other network. And most importantly for this discussion, the internet email protocols could not assume that mail software would always display messages in the same way.
The rendering challenges of today are nothing compared to what you would’ve faced back then — so much so that nobody really worried about it most of the time. Lines would wrap where the screen ended, whether it was 80 columns or 120 or 40. Bold and italics (if they existed) were reserved for the application, rather than the email message content. People had to pay attention to their words, because the final layout was out of their control. Email was sent uphill both ways in the snow every morning after collecting the eggs from underneath the nastiest hen for three counties, and gosh darn it we liked it.
Today, with few exceptions, the email reading and sending interfaces that most people use have converged on the same basic concepts. There are folders, and there’s an index of messages in each. There’s a way to view a message, a way to reply to a message, a way to forward a message, a way to delete a message. A select subset of headers are shown above the message, and an even smaller subset of headers are editable by the user. There’s a new feature here and there, but most people still interact with their messages the same way they always have; except for the addition of the mouse or trackpad and graphical windows, it hasn’t changed much in decades. Mobile devices offer limited subsets of the same functionality, rather than trying anything new.
Wave, more than anything we’ve seen in years, plays around with these concepts: messages are no longer complete when sent; conversations may never end. Attached documents can (sometimes) be edited in place, and you can watch other people make mistakes as they type — even in another language. To accomplish all that, Google also created an entirely different set of protocols underneath.
For Wave to kill email dead, everyone has to switch to using Wave. Sure, they’re talking about making the protocol open, but early reports are that it’s hideously complex. Existing email software built through years of experience will have to be scrapped.
Conversely, Raindrop kept the existing protocols. They’re changing the interface, finding new ways to display and sort messages. They aren’t changing the way those messages get to you.
Raindrop doesn’t have to win out over email, because it is email. If it lives up to its’ promise, it’ll also be Twitter and Facebook and IM and SMS and everything else. You won’t be required to use Raindrop to communicate with someone who uses Raindrop, just as you don’t have to use Thunderbird to communicate with someone else who uses Thunderbird, or Outlook to communicate with someone else who uses Outlook (though Microsoft keeps trying.)
By building on existing, open, widely understood protocols, the people who designed Raindrop have simply created the next in a long line of standards-based messaging clients. I’d go further and say that Raindrop was inevitable: many others have taken steps towards the single inbox concept before, but they usually keep the same old interface we’re all used to.
Facebook and Twitter are perfect examples that a centrally-managed protocol can succeed, at least for a while. So are AIM, and Yahoo! Messenger, both still far more popular than the Jabber protocol that Google Talk uses. Wave could do the same thing, if they can figure out how to make it appear exciting and interesting besides the scarcity of invitations. But will any of them still hold the same position in twenty years that they do today?
Raindrop may not be around in twenty years, either — but a message sent using Raindrop’s email mode today could be received by an email client twenty years from now, or (with a few caveats) twenty years ago. The interface may change, the security models must change (that’s another article) — but email, both as a concept and as a set of standard protocols, will continue.