My heartfelt thanks go out to the great number of IETF participants who have endorsed the device upload proposal. Sadly, my W3C sources tell me that there is still insufficient support for the device upload proposal within the W3C HTML Working Group. The most substantial objection to device upload is currently that it overlaps with future work planned for "XForms." I participated in the XForms effort for most of last year, and it has left me very discouraged about the W3C process. In November, the Chair of the XForms group told me that it would take two years to completion (although the published "HTML Roadmap" claims that it will be ready around the end of the year.) The complete redesign of forms that they are undertaking will in my honest opinion serve no purposes that aren't addressed by server-side software in widespread use today. The enormity of the effort, and my experiences with its leadership, leave me with the impression that it could take several years to reach implementation. In the mean time, non-wintel web browsers will still not be able to record audio. I have devoted my last three years to rectifying this situation, and I ask that you continue your polite requests that the W3C HTML Working Group formally Recommend the device upload proposal. Now, along the subject of distance education, forms, and XML, I would like to announce that the Instructional Management System (IMS) Project has released their Question and Test Interoperability (Q&TI) specification, and has requested public review and comment: http://www.imsproject.org/question/index.html Given that web-based distance education is growing quickly, these Q&TI specifications will prove to be very useful and important for the publication and exchange of, and thereby the development of a market in, assessment content. A number of companies have stated that they intend to support assessment import and export based on these IMS standards. IMS managed to get research scientists from Educational Testing Service to contribute to the requirements and the specification, so few should question the pedigree of the spec. As a practical matter, the spec covers almost every kind of assessment in modern web-based instructional systems: multiple choice, fill-in-the-blank, etc. I hope as many companies as possible will participate in the interoperability trials being conducted by the IMS Project over the next three months. I have no doubt that those companies adopting the Q&TI spec will benefit significantly from doing so. My only concerns with the v.1 specifications is that the results processing directives do not yet provide a way to score the correctness of the results of a "Standard Order Objects" question, although such questions can be represented as items -- see: http://www.imsproject.org/question/qtbest01.html#4.1.9 http://www.imsproject.org/question/qtinfo01.html#A1.10 Also, the "other" condition method, "explicit undefined condition processing," is very unclear, being used as a catch-all without any given logical semantics. There are numerous uses of "other" in the examples that do provide some hints as to when it should return true, but it really needs to be defined. I am sure that these problems will be addressed. I hope that the "Standard Order Objects" questions allow partial-order answer keys so that such questions can easily include more than one correct answer without enumerating all of them. Cheers, James Salsman
