Google Wave: You need to pay attention to this

Wave, in a lot of ways, is just a XMPP client/aggregator. I don't mean that in a condescending way, it's very cool, but it's hardly "the first truly revolutionary leap since HTTP itself".

The article says that this revolution hinges on XMPP letting you "talk directly to ANYTHING, and ANYONE", but that's just false (it's decentralised but not peer-to-peer) and hardly "HUGE": SMTP is such a protocol where "Social networks and proprietary transports no longer have an exclusive license to deliver content", it's just too old and crusty for us to slap some buzzwords on it and claim that "it can easily turn into a bona-fide platform for next-generation applications".

(woke up grumpy)

A. XMPP is the major leap since HTTP. Wave is just some bloated junk that abuses it.

B. XMPP is not the intermediary network that allows anyone in the world to talk to each other, the Internet is. XMPP is just the model that makes it practical.

Just as the main job of the HTTP protocol is to define client-server communication. The main job of the XMPP protocol is to define client-gateway-gateway-client communication.

Wave is just a bloated, locked-in protocol for asynchronous user group communication. Since it targets users directly, and not communication in general, it will ultimately lose to customer-driven social networks. Meanwhile, you can build stuff on XMPP that's a lot more awesome, because like all the successful protocols before it (IP, DNS, FTP, HTTP, TCP, E-mail, Web Services), it doesn't really do that much. You decide what it does.

This some good analysis. I had no idea of the bigger vision of the XMPP protocol.

In the near term I've been working through how we could use Wave within some of our (relatively) simple content-creation processes at work, particularly around those items that require review by a number of stakeholders before being made public.

The key thing for me around Google's implementation is how one Wave, full of modifications and history, can spawn a new Wave in the form of a published version. We continually struggle with content management, and the balance between easy-to-use and "actually-managed." Version history, content re-use, blah.

If Wave can deliver that alone, it'll be worth our effort to use it. If it can deliver all of what Jason talks about this article, it'll be one of the most leveraged chunks of technology in the whole business . . . and insulate us from the whims of Big Software Companies.

This is exciting.

I have no idea what wave is. I keep hearing about it and follow up by purposefully avoiding learning anything about it.

I figure if its really really good, it will simply work it's way into my life somehow without any effort on my part.

Wave is an immature and new technology. And the Wave client does not speak XMPP; it's not an XMPP client.

Federated Wave servers, potentially owned by different parties, need a standardized way to open a persistent socket and send messages back and forth, with security, authentication, etc. XMPP is used as a transport layer for that.

XMPP is defined in two RFCs. RFC 3920 specifies the low level transport between servers and clients. Wave uses the server to sever transport defined in RFC 3920. RFC 3921 specifies messaging and presence. Wave ignores RFC 3921.

I keep saying this with every Wave article: Wave is HARD! It's not the XMPP server-to-server stuff, it's the server-to-client Javascripting that is a giant pain in the ass. You literally have to have a HUGE textarea or designMode parser or just flatout write your own input methods.

I think it might be impossible for Google Wave to live up to the hype at this point. The problem with declaring something is revolutionary or will change the Internet as we know it (blah blah) is those choices are made by massive numbers of users at a grassroots level. You can't just say hey, this is great, you should stop using all the tools you're content with because it's technically better. It doesn't work that way. I'm positive it will be a useful tool to a certain group of people but it doesn't pass my "will my Mom be using this in 3 years?" test. It doesn't even come close. Google should have probably kept this one under wraps until they were ready for a pretty good public beta. Announcing it so far ahead of time creates an environment where lots of people want to project lots of different ideas on top of it without any understanding yet of how people may actually want to use the service.

I have a wave account. I logged in, and I could not even figure out if the thing could send an email to a normal account.

Google Wave is very hard to understand at the moment. It's something that has to be addressed.

I found this to be one of the more coherent and simple introductions to wave.

Meh.