MS IE5 XML not entirely pure, and what's this patent?

Just when the company's making noises about openness, this sort of stuff happens...

Great emphasis is being given by Microsoft to XML, prompting the suspicion that it won't be long before we read of impure XML. Parsing Ballmer's recent comments and looking at the sub-text, when he said that software would "metamorphose", and that it could not do so "in today's environment of monolithic server code", it was clear that Ballmer was working from a carefully prepared text, behaving uncharacteristically, and was not at all playing his usual cheerleader role. What had focussed Ballmer was the role that XML was to play in Microsoft's future plans. If there's one thing we can be certain about, it's that Microsoft's intentions towards XML are less that honourable. It is true that Microsoft was active with W3C in bringing about XML, and that it has made a number of moves towards incorporating XML in its products. If this were Java, it would just be a matter of time before Microsoft's route from so-called monolithic Web sites to integration was only possible with a Microsoft version of XML. Is, as some analysts are saying, XML to big for Microsoft to be able to do this? We'll see. But the nature of the XML support in IE5, and some related patenting shenanigans are causes for concern. XML is already in IE5, Ballmer said (or Windows 2000, as the press release put it, politically correctly), and in extensions for Office 2000, with "a very fast native XML parser". MS SQL also supports XML, as does/will SOAP (simple object access protocol), the Babylon integration server (which is an XML transaction integrator), and the BizTalk framework. He also said: "IE is the first browser to support XML." But a different version of this was included in a press release on 31 March this year, entitled: "Microsoft delivers industry's first XML-compliant browser. Comprehensive XML support enables developers to build new generation of data-driven applications". It was said that "Internet Explorer 5 is the first commercially available browser software" to support XML. There are three things wrong with this statement. First, IE is not the first browser to incorporate XML: it wasn't, since the first was Jumbo, from the Virtual School of Molecular Science and the Virtual Information Technology College. This is a Java application to demonstrate XML and appears to be the world's first XML browser. Second, Microsoft qualifies "first" with "commercially", which is a bad mistake for a defendant in an antitrust case that has been saying that IE was "free". Third, IE is not XML compliant in many respects, nor in intention, it would seem. The Web Standards Project is a ginger group of leading developers that attempts to get software developers to follow Internet standards. It has promoted an Open Letter to Microsoft urging that the standards be followed exactly, and castigated both Netscape and Microsoft for proprietary extensions. WSP said in a press release that IE5 should have been delayed as it did not fully support key standards. So far as XML is concerned, WSP identified bugs in its interpretation of XML data despite there being freeware XML parsers available since 1997. There is also deliberate violation in IE's handling of XML namespaces, preventing a developer using more than one XML-based language in the same Web page. It is not as though Microsoft does not understand where it is at fault, since it played a role in the development of the XML specification. IE also has hard-wired support for the "html" prefix, so that users cannot override this, as should be possible. IE also defaults to XSL (extensible stylesheet language, still not stable and experimental) if XSL and CSS (cascading style sheets) are present in the same page. It's no surprise that Microsoft's implementation of XSL has proprietary keywords and syntax not in the W3C drafts. Consequently, IE5 is not compatible with the XML standard. Another important concern is Microsoft's use of key concepts from W3C's cascading style sheets in US patent 5860073, which is assigned to Microsoft and was granted on 12 January 1999. The WSP feels that Microsoft should assign the patent to W3C. It was unethical for Microsoft to have participated in the W3C standardisation process and not inform W3C (as did Intermind about its patent for P3P - Platform for Privacy Preferences) or assign the patent to W3C. The Microsoft patent also appears to incorporate part of a 1989 Open Text Corporation patent. The industry needs to watch very closely any further XML deviations by Microsoft. How just it would be if it were illegal for those who did not achieve compatibility with a de jure standard to be legally denied the use of the names like XML. Perhaps this is one for Judge Jackson. ® Next part: The march away from COM Analysis - Inside AppCenter Windows becomes the Internet platform How pure is IE5's XML? The march away from COM

Sponsored: Today’s most dangerous security threats