If you think about it, the complexities of building out modern data center infrastructure (locally and inside the Google Cloud) necessitates really watching and learning from the data traffic patterns across a field of Wave servers. There is a great deal of complex data being exchanged in real time as it interconnects with so many parts of the Internet to interact with your Wave(s). This isn’t just simple text like the old bulletin board days, it’s rich media. It’s multimedia in every thread of conversation. I imagine Google needs time to model this this opaque wad of machine to machine, machine to person and person to machine exchange and scale it – again in real time. We’re talking hundreds of thousands of users scaling to tens of millions. How does one do something like this gracefully and quickly? You don’t.
Let’s step back to set the stage.
Some differences between Gmail and Google Wave to consider is that once you’re done with an email conversation you don’t go back for updates. In fact you can’t really use email for much back and forth discussion on issues. It’s cumbersome.
In email, conversation threads remain pretty static as a result after a few back and forth cycles. On occasion, someone responds to something said but that results in adding entire copy of the conversation thread plus his remarks. There is no interaction of the data ‘in’ the email. It’s static, fixed. You can’t reach into an email string and cause the text or numbers to be updated or commented on. You can’t have a program reach into your mail file to update a calculation so the next time you view it the numbers are updated. You cannot for example update an object inside email that has already been received.
In Google Wave you could have dozens ofgoing on different topics. When you come back they are all updated and ready for your quick review. This includes updates to objects like graphs, polls, widgets and so forth embedded in the thread of conversions in a Wave. They are constantly updating or at least there is the potential for that. In reality too much updating can feel like you’re overloaded but this issue is temporary. Most users of Wave tend to jump in and treat it like a CB Radio or as I said like the old bulletin-board days. As that settles and people get more familiar with the art of engagement in Waves so will the perception of it’s use.
Setting up Waves of conversation threads reminds me of the old modem dial in bulletin board days. For free you could download a bulletin board software package and get your friends to dial in to it and talk. We didn’t really have open source back then, free was free. Go get it and help us all learn how to use it better. For the Internet elders this was like The Well which was an amazing collection of conversations managed in San Francisco which was only accessible (originally) via slow telephone modem so text was the only viable thing to exchange in conversation. Of course once a group of people interacting online gets over a certain size it becomes unmanageable. We all experienced this in the early days of mailing lists too. With too many people it becomes disastrous. You bail out. What gets lost is the more nuanced interaction with people on conversations that really means something to you.
Google Wave not only lets you set up your own discussions on your own server and hand select who you want to participate, it also includes the use of multi-media elements inside the pages of conversations. Ok well it will over time. You can drag in whatever type of media or widget you wish and use it as part of the conversation. This includes the ability to drag in little robots and widgets that ‘do’ things for you automatically online. You have the ability to build a tiny world of connected conversations that have value based on who you’re sharing it with. Very much like you do with personal email only now you have personal Waves where a select group of friends exchange ideas, work on problems, debate issues and so forth. If you think of the new economy as being based on the exchange of knowledge and intellectual property then Google Wave is an tool for making this happen. The Waves platform deftly blends person to person with person to computer (and visa versa). More on this in a later post as I familiarize myself more with what’s possible.
Of course for the moment, the Google Wave servers seems to bog down and the availability of robots and widgets to use inside a Google Wave is limited. But this . The IT guys have to learn more about the unique aspects of Google Wave servers to implement predictable capacity planning across a cloud of data centers.course is the purpose of a beta with some regulation over the scaling process
Darned exciting times if you ask me. Though oddly, this whole Google Wave thing sure reminds me of the early Lotus Notes server days. Different but somehow similar. We’ll be learning more soon I’m sure.