Re: Warning: old MERGED_VERSION class still in UML... + versioning question

[Thomas Beale <thomas.beale@oceaninformatics.com>]

Hi Erik,

On 22/03/2010 20:45, Erik Sundvall wrote:

The particular diagram you refer to has the MERGED_VERSION removed
from it, but it has not been removed from the underlying XMI-files and
is thus available as
http://www.openehr.org/uml/release-1.0.1/Browsable/_9_5_1_76d0249_1140019952491_116477_6153Report.html
  

interesting....well, I am just waiting to move to a nice new clean UML _expression_, so this problem will evaporate ;-)

(And the 1.0.1 in the URL and some html page headings is a bit
confusing if this is 1.0.2 :-)
  

well this bit was intentional (maybe a bad decision) the idea was to indicate that the UML was stuck on 1.0.1 (it is....)

Erik asked:
  

P.s. A side question:
It seems impossble to add attestations to IMPORTED_VERSIONs, this
sounds reasonable (e.g. due to problems syncing back those
attestations to other systems without creating a new original
version).
    

Tom answered:
  

on the contrary, you can - see
http://www.openehr.org/uml/release-1.0.1/Browsable/_9_0_76d0249_1109326589721_134411_997Report.html
    

I trusted the specification to be more correct than the UML link you
refer to. Look at figure 7 (page 38) in
http://www.openehr.org/releases/1.0.2/architecture/rm/common_im.pdf
and the text/tables in chapter 6.5
The XML specification also follows the1.0.2 common im spec document
and puts attestations only under ORIGINAL_VERSION, not at the VERSION
level as the UML you refer to. See
http://www.openehr.org/releases/1.0.2/its/XML-schema/documentation/Version.xsd.html

It was via the mismatch between the XML spec and my generated UML
diagram I found the differences and then looked in the spec PDF
document...
  

You have got me there. I think the decision was that 'attestation' is only a meaningful concept on original content, i.e. because it is to do with some medical professional signing off on the content. So the UML is broken here. You should always trust the PDF ;-)

Thanks for bringing up these errors. In general, the PDF UML (although not even built using a tool!) is 99.5% dependable, with most errors found by now (actually prior to 1.0.2).

- thomas

_______________________________________________
openEHR-technical mailing list
openEHR-technical@openehr.org
http://lists.chime.ucl.ac.uk/mailman/listinfo/openehr-technical