Something hilarious…checking the Wikipedia definition of “semantic interoperabilty”–there’s a box at the top that warns “All or part of this article may be CONFUSING or UNCLEAR.” Well–that’s it in a nutshell, isn’t it!!!

http://www.outoftheslipstream.com/node/125

I’m hoping that this will go some way to providing a rational and sensible analysis of the situation.

The #1 reason I think MUMPS is not a viable platform today is because of the weak MUMPS ecosystem (or the smaller ecosystems within everything we call “MUMPS”). When compared to something like, say, Java, MUMPS has almost no tooling, enterprise vendor support, commercial or open source libraries upon which to lean. I believe that if we had a vibrant ecosystem, any deficiencies we have would be quickly patched up or worked around.

Witness Ruby on Rails. Sorry to bring in the “hype” or “fad” of the day, but it is an excellent example. Rails originated in Ruby. It was not the first MVC web framework of the day, but it was the first framework to popularize this concept (among other innovations). Today, just a few years later we have Rails-like frameworks for Java (Groovy on Grails), .NET (Monorail), Perl (Catalyst), PHP (Cake), Python (Django), Smalltalk (Seaside), and so on. A few years ago you could claim “but PHP doesn’t have an MVC framework!” Now you have CakePHP.

PHP had enough community support that, when they collectively realized what a great thing Ruby on Rails was, they went ahead, adapted the concepts and built their own framework in PHP. Thus, they improved their own ecosystem, borrowing from outside.

This is just one example of what a vibrant ecosystem brings you. Don’t forget tools like code coverage, automated testing, source control systems or adapters to existing source control systems, compilers that compile your code into another language or platform, third-party backup vendors, multiple IDEs, debuggers…

So I ask you: do you think MUMPS (or your specific variant) is a vibrant ecosystem? If not, what do you think you’re missing? I’ve explained a bit of what you’re missing, but do you agree with my list? Don’t you think this kind of stuff matters?

Do you think that you, the MUMPS community, is more vibrant and more innovative than the rest of the world, combined? Don’t you think you could adapt concepts originating from the outside the community?

–
And I’m not saying we need a MUMPS on Rails framework, please don’t

You can program in A, B, C or D for all I care – but please don’t tell me that you need a specialized language or data model for health care when the rest of the world is using relational databases for a reason. The only reason Mumps is around is because people dislike to move from something they know to something more efficient.

I compare to sticking with a VAX and using a CISC processor instead of moving to UNIX and RISC. It makes no sense to stick with the old just because it does a few operations better – you need to look at overall performance.

Same with MUMPS. Using a database that excels at hierarchical storage just because you can find lots of hierarchies in health care is hardly a reason to give up on a relational database and object orientated programming. Just doesn’t make sense, unless you are only a MUMPS programmer and looking for job security (kidding).

The next generation that will end up calling the shots will never roll back the language and design improvements that have advanced software development over the last decade. And CIO’s will want easy access to their database sans any viruses