Machine processable specs

Hi Thomas,

Is it possible to have a machine processable spec for BMM?

I could not find it, and I would like to be productive on this.

Thanks in advance for your answer

Bert

Well we could express BMM in BMM, and you could consume it with Archie. Otherwise, there is the XMI from the MagicDraw project, which is a better known standard obviously, but pretty painful to process.

What do you want to do?

Please correct me if I misunderstand. I was thinking that BMM is a model, not a notation language.

So if you can publish it by a parsable notation language, then I generate the model myself in a programmming language I choose later. It seems to mee that the BMM model can be defined in XSD, or maybe in an OpenEHR notation language for which are grammars in Archie.

I think you are right that XMI is not very suitable, it is complex and it is not really a standard.
So, if we can avoid using that, we better do.

Best regards
Bert.

Hi Thomas,

Maybe, after thinking about it, I should explain how I plan to use it.

BMM for me is a container-engine to serve in-memory-models. So what I can do with it is create a BMM for any purpose. (and have an AOM, and AQL definition according to that)

(By the way:The thing I am missing in BMM is a predefined archetype-base-model, if that would exist, BMM could be used to define an archetypable reference model out of the box. But maybe that is not what you want, so I can create that later, myself.)

My problem with BMM is a startup problem that it is harder to write the engine itself then necessary. This because, as I learn now, you have (an XMI schema) in which you define the BMM specs.

I don’t like XMI very much, like you. So maybe it is possible to export it in f.e. XSD? Or in a BMM-structure, but then we have the versions, which version is Archie using? Because it is all in a fluid stage, maybe it is changing, and I don’t want to solve incompatibility-issues. That is even harder than writing the BMM engine by hand. :wink:
So maybe Magic-Draw supports exporting to XSD? Or exporting to Java-code (I can use that as a schema too) or whatever, Eiffel will do too.

If it is not too much effort to publish or send me privately an easy hassle-free machine-processable BMM structure

It is not very likely that I keep on using Archie, because Archie is very much OpenEhr oriented. My goal is much broader.
It is an archetypable database structure and validation-egine and query engine, just like OpenEhr, but then for any purpose and for any programming language and for any framework, like Node.js, Spring, Golang, POJO.

When it is ready, one must just get the JSON BMM-definition, and one is running, documented, semantic, compatible, queryable
I want to use it in combination with document databases.

Of course I will honour the inventors of OpenEhr very much and be forever grateful for this wonderful idea.

Just for fun. I am sixty-three, so I want to do more and more things just for fun.

Best regards
Bert

1 Like

Better is, never mind the trouble, I do it myself, running xslt on the html-definitions is easy, must give a good result because the structure of html is always the same.

Thanks anyway for reading to here

Best regards
Bert

I am happy to build a BMM file of the BMM model - I’ve been thinking about it for some time. There are some recent updates BTW in the expressions, also computed constants, have a look at the latest version.

I will see what else MagicDraw can export as well.

Would be nice, thank you, I discovered that the html-files have many validation-errors, not closed img-tags, etc. Browsers handle this well but XML-processors want me to repair that before executing xslt. :wink:
So thank you very much if you can help me with this.

Bert