Express old hat? (December 1999)

Jenny Meader (POSC Consultants Ltd.) speaking at PIM Ď99 reviewed the standard organization scene, concluding that progress was limited by a natural predilection for the complex. One way forward, according to Meader is to re-tool from the esoteric Express modeling language to the more widely understood and used Unified Modeling Language (UML).

Meader has spent six years working with standards organizations and 2 years with POSC and POSC/CAESAR Association (PCA) 'trying to make standards happen and to make them useful to industry.' Meader is therefore well placed to investigate what makes a successful standard, whether it is de facto or de jure. Unchallenged examples of successful standards are Microsoft Windows, HTML, and the emerging XML protocol. Meader notes the accelerating process of standardization - beginning with the formation of the International Telegraph Union (ITU) in the 1860's, then the IEEE in 1884, ANSI (1918) and ISO (1947). Considerable acceleration has taken place since then with IETF (1986) and closer to home Epistle Core model in 1999. Of particular note is the recent success of 'brain dead' HTML.

why?

What makes standards successful? - For Meader they must be easy to understand, easy to use and freely available. They should not be encumbered by ípoliticalí issues such as who is in charge (vendors or users), who organizes things and who pays. Also important is the degree of verticality (market sector) that is right for a successful standard. Meader concludes that one standard per sector is desirable.

implementable

This should be vendor and user driven. It should be directly and unambiguously implementable and easy to use. Intellectual property rights need to be clear and the standard should be freely available. In the PIM context Meader further concludes that the merger of the PCA/AP221 and Epistle data models is a good sign. Vendor and user cooperation is also a plus. One the downside, PCA suffers from the 'Ivory Tower' syndrome - the underlying technology is neither widely used, nor directly implementable and is not easy to understand. Future directions for PIM related standards should embrace vendor technology and emerging standards such as UML and the XML Schema and graphics standard. A migration from EXPRESS to a language such as UML/XML, for which tools exist and which is directly implementatable is needed soon - 'or we will be bypassed by the rest of the world.'ź

Click here to comment on this article

If your browser does not work with the MailTo button, send mail to pdm@oilit.com with PDM_V_3.3_9912_15 as the subject. Copyright © 2000 The Data Room - all rights reserved.

© Oil IT Journal - all rights reserved.