Developing the "OCL for OpenMRS" Application

[darius (Darius Jazayeri)]

@paulk

Confirmed. Additionally, please confirm if you are comfortable with us using zoom as opposed to hangouts for the next call

I have a slight preference to use Hangouts, but I can do zoom if we need to switch for @akanter.

@paulk

Hello @mohitd This is the right place to report any issues. We have noted your issue and we shall get back to you

This thread is fine for reporting feedback, though it’s high-traffic and it gets in the weeds of dev process. You can also use this thread, which is lower-traffic, and specifically about feedback: OCL for OpenMRS User Feedback discussion

@mohitd

One more query i had is the traditional OCL does not support the export of all the classes and the data types, does the new OCL support them?

@mohitd can you give examples of some of the classes and datatypes that you’re not seeing? E.g. Bahmni adds some classes that aren’t generally used in other OpenMRS installations. Are you looking for those, or something else?

(The “OCL for OpenMRS” application uses the same underlying OCL model, so if we’re going to have a discussion around specific classes we might start a new thread with @paynejd.)

@karuhanga

I’ve got a query in regards to the ‘create another kind of concept’. I’d like to understand why the classes of concepts which are already accounted for with other options on the menu, e.g questions, sets, drugs, procedures, etc are still included as possible options when creating another kind of concept. The only rationale I can currently think up for this is we might have situation where say, a question has a set in addition to its answers. Is this a likely scenario?

@karuhanga, the example that you thought of is spot-on, and there some others.

If someone is creating a dictionary from scratch, we hope they’ll follow the modern conventions. But sometimes they don’t, either by choice, or because they have older concepts, or there’s some use case we haven’t thought of. And we don’t want to prevent that.

Examples:

• A concept was created as Q-and-A, but then someone wants to use it with a new UI widget that expects it to be a set => user decides to duplicate all the answers as set members
• Today we prefer for a diagnosis to have datatype=N/A, i.e. it’s not a question, just an answer. But before many implementations created diagnoses with datatype=boolean, and captured malaria=yes or malaria=no. Going forward we don’t prefer this, but it exists in many dictionaries (including CIEL).