Sun XML architect Jon Bosak talks about how the Universal Business Language can play a part in enterprises right now.
Since the advent of electronic commerce, standards bodies have been in search of a lingua franca for business communication. Unfortunately, early efforts such as EDI (electronic data interchange) had a high barrier to entry in terms of cost and complexity. However, a new specificationUniversal Business Languagemay actually deliver on the common-language promise.
More recently, XML and the public Internet were thought to be likely candidates to succeed EDI. XML, however, was so open that it fragmented into various schemas. Although these schemas are technically compatible with each other, they are next to useless for organizations trying to conduct e-commercejust because the alphabets of various languages share the same characters does not mean that everyone is speaking the same language.
Still, four of the major XML schemas are widely used: cXMLor Commerce XMLis used for automated order receipt and fulfillment, and was spearheaded by Ariba and Sterling Commerce, among others. xCBL, which was developed by Veo Systems (purchased by Commerce One) and was funded in part by the Department of Commerce, is widely used by Commerce One customers and their suppliers. RosettaNet and OAGIS (Open Application Group Integration specification), meanwhile, are two of the more mature standards for business-to-business interoperability.
UBL, built on xCBL and governed by OASIS, is meant to be exactly what its name implies: a universal standard for business-to-business communication. The UBL technical committee released its first draft of the specification last week. eWEEK Labs Director John Taschek recently spoke with Sun Microsystems Inc. XML Architect Jon BosakUBLs leading proponent, a founding member of OASIS and the former chair of the W3C XML Coordination Groupabout how UBL can play a part in enterprises right now.
Can UBL be summarized in a pithy statement as an early attempt to standardize EDI as XML?
That would be a good characterization of ebXML, which basically is an XML or Web services realization of EDI. The UBL effort actually started as a program to complete part of the ebXML stack by providing specifications for the standard documents corresponding to the EDI "transaction sets" from X12 and UN/EDIFACT [Ed. note: The Accredited Standards Committees X12 was the first cross-industry standard for e-commerce. The governing body of the UN/EDIFACT (United Nations Electronic Data Interchange for Administration, Commerce and Trade) and ASC jointly developed core components for electronic business with eBXML.]
So from that angle, yes, the UBL schemas are the XML versions of EDI message specifications. But its equally correct to view UBL messages as the electronic versions of the traditional paper and fax documents upon which most trade is actually based. So one view of UBL is as the way well get EDI onto Web services. But its also the way well get traditional business into electronic commerce.
Is the UBL announcement too preliminary to matter to most organizations?
No, I dont think so. The schemas weve just released for review are sufficient to implement the basic buy/sell relationship that accounts for most actual trade. They will need further customization for the specialized versions used in certain industries, but I believe that those specialized versions will be based on the generic documents and data components instantiated in this release. The UBL component library has been developed in close coordination with [UN/EDIFACT] guidelines and is slated for contribution to the UN business semantic registry. So what youre seeing now is quite likely very close to what the world will use for XML-based B2B for the next few years. If I were in a business organization considering a move to XML, Id want to take a good hard look at this review package to make sure that it meets my business requirements before it moves on to standardization.
Siebel has developed a bet-the-company strategy called UANor Universal Application Network. Does this announcement affect Siebels vision of what will happen with UAN? What does it mean for other companies trying to build universal application networks and how business processes fit into them?
I dont believe that any of the players in this space have really thought through what theyre going to do with business messages outside the enterprise. Microsoft had a try at this with its .Net registry of XML schemas, but that died quietly a while ago, and I havent seen them propose anything to replace it.