= Essay = I've written on AJAX previously, so you are aware of my stance on AJAX, but my fears of balkanization are coming to fruition. Take for example, Gates' recent memos on Web 2.0. Microsoft is finally deciding that they can not only leverage their use of a web browser throughout their own products on a client workstation install, but they can turn Microsoft Office into a web service. Why is this significant? Because before AJAX, there was literally no reason to enable javascript in a browser. Now software providers are finding more and more ways to make it a requirement. Instead of thin services relying upon the basic CGI process, we'll have thick (in terms of bandwidth) services that will be ever so uselessly interactive. Bandwidth requirements go up, computing power needs on the client go up, more Intel Processors ship, the ISPs justify running more and more bandwidth to homes, thick xml-based protocols promulgate, and all the corporations are happy. Why is this bad for us? First, it's bad because the developing world is not going to catch up to the system requirements any time soon. The developing world is thus not able to afford the thick clients and pipes needed for an AJAX-based Web 2.0. Second, As more and more services are required to be download-based, asymmetry continues to develop in the ISP's motivation to deliver bandwidth to end users. Upload prices will remain high as they reap large rewards from corporations that desire to be in on the Web 2.0 world that they develop. Download bandwidth prices drop, but the only parties able to deliver the content will be corporations. Thus the Web 2.0, far from being the democratic medium it was supposed to be, becomes fractured by ever-growing bandwidth and clock cycle differentiation. Peer-to-peer systems and other democratic forms of communication become more and more unfeasable as the asymmetry between download and upload speed continues to grow. Web 2.0 becomes more broadcast-based and less peer-to-peer. Democracy flounders. So, how do we combat this? Well, I have a few ides. Content providers who care can pursue the following policies: First, they should continue to provide standards-compliant and accessibility-driven websites that follow core protocols and ignore implementing core utility features in javascript. The semantic web is not a javascript world. The W3C, the IETF, nor any working group of the Internet Society did not standardize ECMAscript/javascript. ECMA did. ECMA is essentially clueless on policy decisions. Second, instead of making crazy, proprietary thick-clients that only use part of the standards stack, we should continue to stick to W3C-based protocols such as XHTML, CSS, and XSTL. I believe the free software community has done a pretty good job with this, especially in regard to open documentation and document formats. Third, instead of promulgating arbitrary XML schemas, we need to ensure that the XML schemas are well-documented. Microsoft, for example releases completely undocumented XML schemas. Some free software projects also have a problem with this, in that they are all documented in implementation, but they have no formal documentation that third party developers can use to develop against while expecting stability. Fourth, if you have to use AJAX, make sure that it's got a very good reason, that it is accessibility-friendly, and that it's well-documented such that people can easily write accessible interfaces to the XML to present it using ones own scripts. By following some simple guidelines (more are welcome), we can make sure Web 2.0 doesn't Balkanize the Internet into the content-providers and content-consumers, undemocratically. Why will we prevail? Because the very existince of these alternatives allows the continued interconnectedness of under-interconnected populaces. With an alternative system provided, people will have a choice of participating in either system freely, and once the infrastructure is in place, created by free software and the cooperation of developing countries' governments to fund cheap, mass hardware deployments (hopefully enabled by the rejection of foreign patents) the content classes can create their private networks without risking the destruction of the public commons already created. That still leaves only one major trick we have to perform: the democratization of ideas through the rejection of software and trivial hardware patents. Software patents and absurd hardware patent enforcement will eliminate all the rosiness of my outlook and allow the content producers to monopolize, through a government-enforced monopoly, all the technologies that would have allowed the democracy of ideas to flourish.