M as a language is not quite as dead as the patients whose lives are being lost each day because the functionality of VistA is restricted to the VA. The urgency in the IOM reports (among others) merits working with what we have now. Otherwise, I don't think we have done our patients a service. What is happening now is that medical open source is spinning its wheels (maybe a little traction here and there) while Cerner et al attempt to become the "next Microsoft".
I can easily see this list in the future saying to its collective self, "Oh, well, we tried, but Cerner really has the best functionality and is the only show in town, so why bother?" After all, Brian Bray, the founder of the list, has already said that open source must follow the taillights of proprietary code.
That would be (remotely) excusable, if VistA didn't exist. But it does, and to condemn it to death because it's written in a "dead" language takes a lot of patients down with it in the death spiral.
Please note that the VA is not spinning its wheels with VistA. They are the first to implement barcode scanning of medications at the point of care. Etc. etc. etc. It is very interesting to see Kizer on the board of directors of Medsphere. He is definitely a mover and shaker.
I admit that if you remove the moral argument then VistA is just another program and we can wait. If we were talking about predicting the price of grain futures, we certainly wouldn't be talking about M. I recognize that.
As a programmer for some 30 years and with a lot of languages under my belt, I'm not interested in learning another language that doesn't move me forward in my career .... I have enough dead languages already in my brain.where I work I push for building (large) components that interconnect through standard APIs (soap/corba/whatever) goal being to allow graceful transitions to future technologies/code/apis I have no illusion that today's fav language won't be eclipsed by the next-great-thing. And for that matter the external API will change as well ... CORBA->SOAP->whatever. But if I've built clean (large) components well it will be easier to evolve, lower long term costs. I'm not against vista/m if it buys you something now. But (imho) you're not going to pull in a lot of good-will-oss coders if they're going to work in an older language. That may not be important to you ... if your target coders are good-will doctors having fun hacking on the side then m/mumps may be great. But for a professional programmer the time spent learning mumps is counter to their development as a professional. java or c# center of action python next tier down possible ruby/eiffel/your-favorite-scripting-language down another tier? btw, I haven't played w/ but watch the GNUe effort and it may be possible to implement a med system in it w/ less effort because it provides a framework http://www.gnuenterprise.org/project/what.php which runs counter to what I said above re languages, but in this case you code templates/schemas and you don't really live as much in language ... just a thought to the extent this email list works on implementations (as opposed to just design/theory) suggest working on one of the modular/componentized oss efforts even if it isn't as complete as a vista now. Need a mechanism for shorter code-test-deploy-evolve cycle. Get a working testbed up and functioning to play w/. and I apologize for any out-of-contextedness in the above Heitzso
