Google Wave is Google's attempt to blast every other communication framework out there into the dust. Will it be that successful? I doubt it, but it's still gonna be making waves (ha ha pun) for several years.
If you don't feel like watching the intensely long video that Google recorded when it debuted the software at its I/O conference back in May of 2009, you're in luck. I'm gonna describe Wave to you, though it will probably be a little strange.
Google Wave is the crossbreed of email, instant messaging, wiki, and forums. And that definition itself is probably glossing over half a dozen features that don't come from any of those four technologies. It aimed to combine the best elements of each, and in my opinion, it has. As it is (I'm looking just at the invite-only preview now), there are a lot of rough patches on the software, but it has delivered admirably on it core functionality.
Google Wave is, unsurprisingly, built around a core construct called a "wave". Each wave is a conversation stored on central servers (either Google's or some other cloud-based provider--more on that later) and displayed to all of its participants. It functions somewhat like a forum thread in this respect, giving the participants one centralized copy of the conversation. Where it differs from a forum thread is that in Google Wave, replies can be inserted at any point in the conversation. Not only can replies be made to posts at any stage, but they can be made to particular parts of posts, instantly splitting the post to allow the replies to be made at the designated point.
It is also possible for participants to edit not only their own postings, but messages posted by other users in the wave. This allows for collaborative editing to be done in waves, right alongside comments and discussions about the project being worked on. In order to prevent confusion from erupting should participants join late or should someone go through and change participants' writings, Google Wave includes a "playback" feature, which allows users to step through all of the activity in a wave and see how the present wave was reached.
Currently, Google Wave is confined to the wave client program, accessible either through an AJAX web page or via ssh commands in some cases. Google hopes to be able to integrate waves directly into standard HTML web pages, allowing users to make waves available on their own personal websites or blogs. As of the time of writing, this feature has not been made public, probably because Google is still trying to control the number of users that can access its beta. Similarly, Google is working on ways to pull information from other web applications into waves. The video showcases syntheses of Wave and Twitter and of Wave and Orkut.
Waves are given added functionality by adding robots to them. These robots edit the waves as regular users are editing them to do certain things. Some of the more interesting robots that have already been produced are: Tweety, which interfaces with Twitter; a dynamic translation bot which I cannot seem to recall the name of; a dynamic spelling-correction bot which I believe is automatically included in every wave; and a bot which interfaces with the Blogger/Blogspot network.
Perhaps Google Wave's most exciting feature is that it allows multiple participants to edit a post at the same time. By keeping a single copy of the conversation stored on the server, Google Wave makes it possible for lots of users to constantly ping in changes to the document and keep their own copies updated relative to the original. I am not sure how well this works on real-world internet connections (eventually a speed issue must present), but it seemed very robust in the Google presentation.
For all of its shiny goodness, Google Wave has several problems. Some of them are high-level problems with the fundamental structure of the application and unlikely to be changed, but some of them are pretty basic and will hopefully be ironed out by the public release.
The interface for Google Wave is presently rather cumbersome: it is split into many windows, some of which are mis-sized, and the minimization functionality offered as a solution to this problem is not as effective as one would like. It is also currently rather difficult to find contacts on Google Wave. Currently one must take a hit-or-miss approach of finding people in public waves or guessing at contact email addresses. The software is also plagued by speed problems, though it is unclear whether this issue can be resolved by Google, or if the problem is just with insufficiently speedy client connections (recall that the demo was conducted over a LAN to the servers which were hanging out right next door).
On the broader scale, Google Wave has drawn fire from the same groups that have been attacking the cloud model entirely, on many of the same charges. Google Wave's excessive reliance on centralized servers has both privacy and security groups a little on edge, because it gives Google so much control over user content and because it makes Google a single point of failure in the system. The production of Google Wave and of Google Chrome OS also have antitrust concerns worried about Google's increasing dominance over large sectors of online activity, prompting fears that "World Domination" has finally gone on the roadmap.
barrythefigment noted, "I believe Google released all the specs for the Wave protocol, so it should be possible for interoperable open-source implementations. If you're going to talk about worries about centralized servers and excessive Google control, you might want to double check that and mention it as a counterpoint."
Right you are barry, and though this does ease fears of a Googlopoly on Wave, the fact remains that the system is still centralized to servers, whoever is running them, and this poses the same security risk.