How to fix CKM biggest issue
pazospablo at hotmail.com
Sat Mar 14 10:00:26 EDT 2015
IMHO ADL is very readable. More than XML. But of course depends on how much knowlegde the reader has about the model below. Without knowing the ADL syntax and the AOM/AOP models, reading ADL is almost imposible.
I would love to use GitHub for versioning, but I need the mindmaps and the tabular views of the header and content of the archetype that the CKM provides.
I think that the main proprietary portion of the CKM is the one that handles the versioning, if that part could use GitHub, I think most of yhe problem can be solved.
Sent from my LG Mobile
------ Original message------From: Marcus BawDate: Sat, Mar 14, 2015 10:21 AMTo: For openEHR clinical discussions;Subject:Re: How to fix CKM biggest issue
On 14 March 2015 at 04:53, pablo pazos <pazospablo at hotmail.com> wrote:
For me the biggest concern, besides the limited publishing capabilities or non editors, is that the CKM is made over proprietary software, that doesn't allow us to create our own instances of the CKM for free, and share archetypes in a distributed / versioned way, like GitHub does.
Pablo, you've nailed the problem here. The CKM is proprietary.
"All contributions to CKM is on a voluntary basis, and all CKM content is open source and freely available under a Creative Commons licence" From openEHR Foundation website: http://www.openehr.org/programs/clinicalmodels/documentation
There's a disconnect there. I have in the past been in the middle of trying to explain openEHR to open source 'purists' and been left with some uncomfortable questions to answer about the tooling used not being freely available. (no, despite what may appear to be my OSS zealotry I am actually not even close to being a Richard Stallman-esque OSS purist)
'community' computing is very definitely moving away from anything that is dependent on proprietary platforms, towards cross-platform, open source, generic systems. Open source languages, and Git for version control.
If we could find some way to wrap ADL in a more readable language then perhaps we really could just use GitHub for archetype sharing one day! One of the primary reasons for reliance on a GUI is that ADL in its raw form is so unreadable. If it could be read and understood in a text editor then there would be less need for a GUI. I accept that clinician led review would still benefit from a GUI.
Another benefit of using a mature version control system such as Git is that some of the metadata about archetype authoring and details of who did a certain translation could reside in the version control commit history and would therefore not need to reside inside the archetype itself. This would reduce the size of archetypes, and would also obviate some of the problems such as the one Silje mentioned on another thread - in which there isn't room to record more than one translator.
BTW this post is very definitely not intended as a criticism of any individuals, and I recognise the massive amount of hard work that has gone before to even get where we are now.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the openEHR-clinical