On Mon, 1 Oct 2001, Wayne Wilson wrote: ... > Very few of the efforts being discussed on this forum are attempting to > define health care web services. There was a discussion of a means of doing > this for drug databases which came the closest. OIO can export in XML.
Wayne, The OIO project does aim to provide a framework for health care web services. However, we only have the more basic "services" implemented :-). Currently (OIO-0.9.9), the system can import/export forms-metadata, patient data, and multi-lingual translations via XML. Using the OIO Library, very primitive discovery of these "services" are possible. Direct transport of the XML documents via HTTP between OIO servers is trivial. The challenge, IMHO, is to provide adequate tools for the authoring and inter-operations of these "services". Again, we only have the simplest kinds of tools in these areas - but it seems to be adequate for most of our purposes. Within the OIO framework, a "OIO web-form" is the simplest type of "web-service". I believe .net has a similar construct. One of the major next steps for the OIO project is to support more complex "services" - especially those that require a full workflow model. There have been some discussions about this on the open-outcomes-general mailing list: customizable protocol/workflow support, was Re: OHSCA meeting http://www.geocrawler.com/archives/3/5445/2001/9/0/6654759/ >From flexible workflow to ?, was Re: data entry by whom? http://www.geocrawler.com/archives/3/5445/2001/9/0/6668680/ OIO-1.0.0 will have the ability to import/export arbitrary comma-delimited and XML documents (e.g. TkFp patient records, drug database, GnuMed, GEHR systems, OpenEMed, FreePM, FreeMed, etc). There is much much more work ahead, but I am hoping that this will lead to inter-operability between most of the open-source health care "services" - web-based or not. Getting back to the ".net cost" topic, a co-worker's son who works as a programmer for Microsoft commented that this is exactly what Microsoft/he is working on after reviewing the OIO project description. This was back in 2000 when most of the OIO architecture were still just on paper. He did not think it realistic/possible to implement an OIO system without significant resources because this was/is THE major project within Microsoft. This is true but I don't really think we need to/should be writing code de novo and defining a new programming language to build useful web-services. I believe a great deal of the $2 billion or so are spent for "marketing" reasons. Of course, the "funding" model for free/GPL projects is quite different and we may have spent more than $2 billion without even knowing it :-). Of course, it remains to be seen whether the "community" can deliver anything comparable/superior to .net. I believe it is possible and the OIO project can make some contribution in that regard - and Yes - the OIO implementation may have to move to Java eventually :-). Best regards, Andrew --- Andrew P. Ho, M.D. OIO: Open Infrastructure for Outcomes TxOutcome.Org (hosting OIO Library #1) Assistant Clinical Professor Department of Psychiatry, Harbor-UCLA Medical Center University of California, Los Angeles
