Is Google Wave just Gmail on Waves?

In a previous article the question whether Google Wave can tackle the ever growing inbox was raised. A few tips were tossed, mainly suggesting user discipline assisted by Google Wave.

I haven’t played with Google Wave yet, my thoughts in this article are solely based on screenshots of the Google Wave client (see for example Figure 1). But when I compare those screenshots of the Google Wave client to Gmail it seems to me there’s not much difference. We still have a central inbox, only now it contains waves. While I love the wave concept, I fear Wave extensions and robots will only increase the size of our inboxes.



Figure 1. Google Wave. Source: The Official Google Blog

There’s no doubt waves have the potential of drastically changing communication and collaboration. However, assuming the screenshots are representative, I do doubt the current Google Wave client will improve the management of my inbox. In this article I will try to figure out a way to collect and present waves to the user without cluttering their inbox.

Inbox vs. Archiving

On of the key points from the previous article is to delete, act on and/or archive mails in your inbox. This in fact got me wondering, how or what do I actually archive? In my case, I have about 1400 messages in my inbox, another 1300 archived. Archived mails mostly include newsletters, iTunes receipts, serial number of programs I bought, et cetera. On the contrary, my inbox contains all kinds of mails, but mostly conversations with contacts. This means my archive is actually cleaner or better structured than my inbox! Or at least the way I am archiving is not sufficient to cover all emails in my inbox.

Before going further, let’s take a look at some other inboxes.

Snail mail Regardless what you do once you’ve opened the envelope and read your mail, you never put it back into the mailbox! In the laziest case, you either throw it away or put it on a pile. Otherwise you classify the mail in categories such as correspondences, bills, magazines, commercial, ... Mostly without even realizing.

Task Managers They often provide an inbox for quickly gathering new tasks from emails, sms, webpages, et cetera. Once you have a spare moment of time, you start labeling the tasks, pick a due date and assign them to projects or more general context (cfr. GTD).

By some mysterious mechanism these kinds of inboxes end up empty. Mysterious? No. Snail mail of course takes up physical volume which is either limited by your mailbox or your hallway. Regarding task managers, I guess we simply prefer shorter task lists over longer ones.

So do we need ways to classify waves in our inbox? We already have labels and filters, you say? True. We all know the combination of labels and search serves as an alternative to folders, that’s what Gmail is about and Google Wave seems to use the same principles. However, labeled e-mail keeps sitting in front of us in the bloated inbox unless we archive it which is just putting it on a second pile. So there’s actually no real difference between the Wave inbox and the Wave archive. Both pile up to a confusing mess of waves which we have to query to find anything (luckily Google is good at that).

Another View

So we need ways to classify waves and move them out of our inbox. We need to bring some of the inbox cleaning mechanisms from snail mail and task managers to the wave inbox.

So let’s start with automatically moving waves out of the inbox once we’ve opened the wave and start classifying the wave right away. The default classification could be archive. I know, now we’re creating one big pile again, but let’s take the Gmail concept of labeling and saved searches to a higher level. Since we can expect an explosion of robots and gadgets, we can use them for a more intelligent classification (even before we opened the wave).
In my eyes the central view on Google Wave should be something along the lines of Figure 2; an (almost) empty inbox and saved searches. I call those saved searches views on my wave collection. Like the current filtering system, we can create views based on labels, based on contacts (Friends, Family, ...), but also based on collaborations and documents (Project X, Project Y, ...), based on robots (Facebook, Bloggie, ...), based on gadgets (Chess, polls, ...), date ranges and so on.
You might even go as far as deleting the notion of an inbox altogether and simply define a view of new waves. Conceptually this is even more beautiful.



Figure 2. A mockup of the concept of Views in Google Wave.

You’ll notice each view has a badge counting the number of new waves for that view. We could even add more badges such as a badge for the number of active waves. Views based on robots and gadgets could define their own badges. For example, Taskie may count the pending tasks, while Facebookie counts the number of updates and wall posts separately, ...

So with one eye drop on this Google Wave mockup I see what’s actually new in my inbox, using the top view icons I see a summary of new, active or otherwise classified waves in my wave collection. Those are positioned central instead of tucked away in a sidebar. Finally, the active view presents me a filtered list of only those waves I am currently interested in.

Start

While Google say they started with the question “What would email be like if it was invented today?”, I can’t help but think they actually started with “So we have our mails in Gmail, what else can we add?”. They’ve build an amazing platform, but it’s not quite there yet. However, I think under the Google Wave hood we have the right engine to go beyond Gmail on Waves and change our views on waves.

Early Google Wave Dissection

(guest blog by Steven Devijver)

Over the last couple of days I wrote a seven-part analysis on Google Wave. I have looked at different aspects of the platform and wrote this overview for your convenience.

Part 1: Architecture

Google Wave is based on an associative memory architecture which is unique in the world of social platforms. As with e-mail anybody is free to host their own Wave provider which Google will release as open-source. Because of its architecture a Wave provider never leaks content to the outside when a wave is between participants on the same server, just as with e-mail.

This is a huge benefit of Google Wave because now any organization can deploy its own Wave provider for its own people and prevent sensitive content from leaving the safety of their private networks.

Part 2: Unified Messaging

Like Twitter, e-mail and instant messaging Google Wave has the free-text message as its content model. Yet Wave is different. First of all, participants can be machines in the form of robots. Secondly, participants can add any kind of content to a wave: text, chatting, video, pictures, gadgets, … . Thirdly, each participant can instantly become a co-author.

The Wave team has understood that unified messaging is really about two things: each content type requires certain kinds of communication (e.g. instant messaging requires P2P communication), and content types are defined by our internal articulations rather than by its technical characteristics. In other words, the Wave team has understood that content is a social object as has build the Wave platform accordingly.

Part 3: Extensibility

Robots and gadgets are first-class citizens in Google Wave. This means a robot can modify the content of a wave just as a human can. In the demo video we see several examples of this: converting www.google.com to http://www.google.com and the fabulous spell checker.

Yet the real power of extensibility is that anybody can build third-party additions for their own purposes. These robots can be hosted on any Wave provider, the only thing that is required is to add a robot as a participant to a wave.

Part 4: Collaboration

Google Wave will dramatically change the way we work together as soon as we realize that the waves we create can be harvested for metadata. With this metadata we’ll be able to build smarter robots that know us better.

Part 5: Version Control

Google Wave is not just a collaborative word processor, it also comes with a distributed version control system built in. This is a significant feature, especially for long-term group collaborations.

There are many open-source version control tools available but they’re used by software developers and don't lend themselves for integration in productivity suites like Office or Google Wave. That is way the Wave team has taken the ideas behind Git and built a new version control system for collaboration.

Part 6: Social Network Platform

Google Wave is not just an awesome social platform, it will dramatically change our expectation of what a social platform has to do for us. In general, social platforms have to be able to evolve with the users and communities they host. Facebook and Ning are notoriously unable to do this while Twitter is notorious lenient towards the conventions its users prefer.

It's probably true that complicated data models for social platforms as sported by Facebook and Ning are not a thing of the future. Instead, free-form content models like Wave and Twitter give much more freedom and allow us to organize unified messaging as we see fit.

Part 7: Marketplace

Many business problems are hard to solve for the same reasons that social platform like Facebook and Ning are not very malleable: it's painful and expensive to let dedicated data models evolve with reality. Look at CRM tools, look at sales tools, look at procurement tools. They all suffer from the same design flaw.

Enter Google Wave. We can integrate any business information we want to keep close in Wave through robots and gadgets. Businesses will demand that these third-party extensions run on their own networks. Et voila, the soon-to-be booming market of third-party Wave extensions.

I've enjoyed writing about Google Wave and the response has been overwhelming. Thank you. I'm looking forward to the actual release of Wave and working with others to make their Wave dreams come true.

Can Google Wave increase email productivity ?

Most people will agree that productivity gains could be reached by optimizing the way we handle email. However, a great deal of optimization is already possible today with tools in our current email clients, such as Gmail. In some cases, lack of knowledge of these tools or discipline lie at the cause of many email annoyances.

It looks like following real problem, that currently can't be tackled with Gmail//email, can be solved with Google Wave.

Tackling management of one-way updates, such as newsletters

The way you can manage a newsletter with GMail is to create a filter, enabling you to make the message skip the inbox. The problem consists of the fact that all the messages are sitting under the filter, as unrelated, atomic items. But because these messages are newsletters from the same organization, it's highly likely that their content will be related.

Let's imagine that you will keep receiving plain email newsletters, once you are using wave. If your wave client allows you to channel all the incoming newsletters in the same wave, you can include a robot or gadget to do certain tasks for you. It could start archiving and processing older newsletters, and present their combined content in a much more accessible way. For example a tag cloud, or a searchable index.

Taking one step further, when newsletters won't get sent out in plain email anymore, but are created as waves, the whole concept of newsletter can be redefined. The wave, sitting in your email client could be a live updated list of top5 newest items on the source site, or 5 most popular items.

In any case, we might be able to get rid of recurring, individual email messages, just because it doesn't require an individual message to update the wave, and bring new content to your attention.

To be continued by how Wave can tackle the Ever Growing Inbox

Watching your contact type ... nothing new ?

From the dark past, where instant messaging was still associated with a flower that shifted between the colours red and green, and messages had an annoying "oh oh" sound, I remember the feature where you could see your contact type every character.

A blogpost from 2008 confirms this as well, although I can't seem to find any screenshots.

Of course, you were typing in your own little window there, and didn't have any concurrent editing options. But they did have the character per character communication.
It would be interesting to see the technologies compared. It's already sure that connection response times won't pose any theoretical issues, if even non-broadband connections could handle this.

Ben Hyde on Conflict resolution, Background, Federation and Access Rights

Ben Hyde wrote in his blog about the use of conflict resolution techniques, familiar from source control, in wave.

He indicates that, instead of the conflict resolution algorithm making an average of 2 values, entered at the same spot, that it picks one. I could imagine a wave gadget that gives certain users a "master" right to a document, where their own edits can never be editted by other users, unless they specifically request access for it. Ok, not really a leap forward for collaboration, but might come in handy in, for example a student-teacher collaboration relation, or another situation where you want a tighter grip on control.

He also goes into detail about the historical background of concurrent editing, how different wave servers can be federated, and how access rights will be managed.

He states that it's totally unclear at this point how access rights are managed in the deeper structures, constructed in a wavelet.

Watching those Waves

Did the Google Wave demo hit you like a brick wall ? Although it might look like this at first sight, it's not just about solving a few web-annoyances, nor is it just the next incremental step of new functionality.

The generic concepts and the open platform will enable levels of interoperability that we haven't seen before. Google seems to have struck a good balance between offering a generic interface, in order to build different applications, and just adding enough semantics to ensure that the same API's are relevant for the intended uses.

To clarify this, let's compare with SOAP: while SOAP webservices are very generic. It's a technical standard where you define your own API's and semantics for every application you think of.

On the other hand, many succesful standards are very specific, and therefor limiting their use. RSS is a fine protocol and way for one-way communication, but it stops there. Err, does it really ? Put a powerful search on top of RSS, and there you have Twitter. But does it "work" perfectly well ? If you start a twitter "conversation" the information is not captured into a unit, and will be very hard to retrieve afterwards. So yes, two way communication is possible, but this illustrate that water comes through the cracks when protocols get "abused" for stuff they aren't meant for.

As long as we're talking information that can be represented in text, and a process of communication or collaboration ... it looks like we'll be able to handle that with one single protocol.

Maybe it's a dream and reality will soon kick in, but really, the examples shown in the demonstration are just the tip of the iceberg.

Just like email has become an important chain in many processes, waves might become a cornerstone in every day workflows as well. The amount of communication that we are involved with has been rising without stop. As far as I know, there are no tools available yet that seamlessly blend this all together. This is exactly what might get accomplished with waves.

It's going to be an interesting journey ... and wavewatch is in for the ride.