Meeting 1 (2014, Oct 14) chat and notes

Owned by Erik Sundvall
Last updated: 14-Oct-2014
5 min read

The meeting was conducted via the meeting room https://connect.sunet.se/openEHR

A recording of the meeting (53 minutes) is available at https://connect.sunet.se/p2xmsvfxshu/ 

  • You may need to unblock popups for the site https://connect.sunet.se (the conference service of the Swedish university network)
  • Note that the replay-popup-window is somewhat interactive so you may need to scroll in chat windows to go back and forward. Also note that you can jump in the recording based on chat events listed in the "Events Index" pane (if you expand the "Collaboration" heading in that pane). Sadly enough you can not kill the hand-waving video of me (Erik) during the start of the session, but I paused the camera after a while...

Misc. notes

PETER: What is 1.5? I thought there what we used to call 1.5 is now known as 2.0.
Erik: Yes 1.4 vs 2.0, sorry
PETER:

  • Ian is saying that 1.5 would be backwardly compatible with 1.4. 2.0 would have breaking changes.
  • Bostjan is saying that 1.5 would differ internally from 1.4 (AOM etc.), whereas with 2.0 there are external changes in the ADL, etc.

IAN: Bostjan's explanation was correct IMO .

IAN: To answer Peter - the source ADL 1.4-> 1.5 does change but by the time it is 'compiled' back to operational template format, the adl.14 .opt is very similar to the adl1.5 .opt i.e there is a natural junction point.

Chat transcript

  • This is a somewhat edited version of the written chat (things about microphone/speaker-testing etc removed)
  • A spoken conversation occured simultaneously (available in recording above)

----------------------------------------------------------------------- (10/14/2014 11:16) -----------------------------------------------------------------------
Diego: Hi all! I'd love to help with this, but I'm really focused on writting my PhD thesis :( so I won't be able to be as involved as I'd like
----------------------------------------------------------------------- (10/14/2014 11:43) -----------------------------------------------------------------------
Erik Sundvall: Diego, I know the feeling. Focus on the PhD. But you are welcome to give some feedback when you need a break/distraction...

[removed hello/mic/speaker/test etc by Sebastian Iancu, Thomas Beale, Choi Dong Won(Republic of Korea), Alex, Erik Sundvall, Ian McNicoll, Diego]

----------------------------------------------------------------------- (10/14/2014 12:17) ----------------------------------------------------------------------- 

Thomas Beale: I'll add some things to the wiki re: APIs, from ADL Workbench
----------------------------------------------------------------------- (10/14/2014 12:23) -----------------------------------------------------------------------
Sebastian Iancu: i agree

[...]

----------------------------------------------------------------------- (10/14/2014 12:27) -----------------------------------------------------------------------
Choi Dong Won(Republic of Korea): I'm sorry to interrupt your conversationI don't have mic...Hi, this is choi dong won. in south of korea, interesting in openehrI am a member of nousco.co.,ltd.(www.nouco.com)We are developing the EHR platform based on openEHRI already developing a template designer for the Web version.this project is very interesting for meI will be observing today because I can't speak very well

Sebastian Iancu: 1.4/1.5 is still very important
[...]

Bostjan: i think from adl 1.5 we could easily produce regular 1.4 opts?

Bostjan: so we might not need tool to natively support 1.4?

Thomas Beale: I would suggest that the internal tooling works based on ADL 2

Thomas Beale: ADL 1.4 outputs need to be an input and output mode

Thomas Beale: reasoning properly with ADL 1.4 internally is hard/impossible

Thomas Beale: I would suggest so

Thomas Beale: right

Ian McNicoll: Agreed

Thomas Beale: we still need to see if we want to define a new 1.5, 1.6 etc

Thomas Beale: example things for a new 1.5 = adding annotations

Thomas Beale: etc

>> spoken question here regarding how stable the ADL 2.0 spec is right now... 

Thomas Beale: it is very stable

Thomas Beale: we are looking at final changes due to Mayo / Intermountain review

Thomas Beale: that's the team converting it to AML for OMG standard

Thomas Beale: and they are relying on stability as much as us

>> TODO: Listen to recording and add spoken question here 

Thomas Beale: one way to do that might simply be to reprocess the current ADL WB OPT XML

Thomas Beale: e.g. with Xslt or Xquery
----------------------------------------------------------------------- (10/14/2014 12:39) -----------------------------------------------------------------------
Thomas Beale: the current workbench can be made headless and wrapped

Thomas Beale: but I suggest just looking at the flattened XML output and seeing what conversion is needed from there
----------------------------------------------------------------------- (10/14/2014 12:40) -----------------------------------------------------------------------
Diego: Is there an XML schema for the 2.0 AOM?

Thomas Beale: good Q ;-)

Thomas Beale: not yet!

Diego: ;D

Thomas Beale: but Iguess it isn't a lot of work

Diego: I have tried to generate 1.4 AOM from other models via XQuery

Diego: in theory was possible

Diego: so AOM 2.0 to 1.4 seems feasible

Thomas Beale: you might want to look at the flattened XML output in the ADL Workbench

Bostjan: great diego, thanks
----------------------------------------------------------------------- (10/14/2014 12:44) -----------------------------------------------------------------------
Bostjan: i think now we need only one tool for both templates and archetypes

Thomas Beale: we do

Sebastian Iancu: ... and demographic-RM

Thomas Beale: next ADL WB release can create new template on single click

Thomas Beale: Bostjan - this is a very good point

Sebastian Iancu: agree

Thomas Beale: Q to group: importance of 'clinical friendly archetype editor'?

Thomas Beale: I try to keep up;-)

Sebastian Iancu: this is quite a good platform

Thomas Beale: platform Q: def ned to separate pres layer from back-end

Sebastian Iancu: (first time user)

Diego: just that if multiple RM are considered from the start then this is a critical issue to be taken into account at the beginning

Thomas Beale: also true
----------------------------------------------------------------------- (10/14/2014 12:49) -----------------------------------------------------------------------
Thomas Beale: BTW in CIMI, a Enterprise Architect => BMM converter was built

Thomas Beale: it means you can build archetypes based on any RM

Diego: yes

Thomas Beale: e.g. FHIR, CDISC, 13606

Thomas Beale: even openEHR ;-)

Diego: node identification, not interesting classes

Diego: etc.

Diego: agree

Thomas Beale: @Ian - with RM-independence, you need to create a whole layer on top that maps from 'tree' structures to friendly view

Diego: Maybe a question before any other is if we want a modelers tool or a implementers tool :)

>> TODO: Listen to recording and add peters spoken question/statement here 

Thomas Beale: agree with Peter's statements

Bostjan: me too

Choi Dong Won(Republic of Korea): later... gdl editor also plugged in tool?

Ian McNicoll: me too - I thikjn we need ot start with the more technical layer

Thomas Beale: GDL is another good Q

Ian McNicoll: @Choi - good suggestion

Diego: Yes they are

Diego: haha
----------------------------------------------------------------------- (10/14/2014 12:53) -----------------------------------------------------------------------
Diego: been trying to generate a openEHR specific clinician view for archetypes

Diego: not so easy at it seemed at first look :(

Thomas Beale: it's worth looking at CKM - it actually abstracts even openEHR in subtle ways

Diego: yeah, problem is abstracting in the creation

Bostjan: sometimes you want/need RM view as well - i like TD 2.7 which has that option

Thomas Beale: agree

Sebastian Iancu: no i'm done also

Thomas Beale: that's why ADL WB has those controls on right hand side

Sebastian Iancu: ok

Ian McNicoll: @Erik - thanks for hosting and organising

Diego: thanks! :D

Alex: Thanks. Good night

Thomas Beale: thanks

Diego: bye

Choi Dong Won(Republic of Korea): bye
----------------------------------------------------------------------- (10/14/2014 12:58) -----------------------------------------------------------------------
Erik Sundvall: Stopping recording now. I will transfer chat etc somewhat edited to the wiki.

Agenda + interactive speaker-list

This list/agenda was used part of the time and may help identify timepointe in the recording.

Welcome - Erik Sundvall
Short presentations: Who are you? Main interests in this tooling project? (add your name here)

  • Erik Sundvall
  • Bostjan
  • Peter
  • Sebastian
  • Ian McNicoll
  • Thomas Beale
  • Seref
  • Alex
  • Diego
  • +have a look in at presentations in the chat transcript

Thoughts about the approach of first experimenting with modules/apis  (and a lot of other things, 

  • Erik
  • - Tom
  • - Ian (about scope) - backwards compat.
  • - Seref (1.4 vs 2.0)
  • - Sebastian
  • - Erik
  • - Ian (Upcoming 1.5 and 2.0 at the same time)
  • - Tom (2.0 has breaking changes)
  • - Bostjan 1.5/2.0/1.4
  • - Erik
  • - Bostjan
  • - Ian
  • - Peter
  • - Ian & Peter
  • - Erik
  • - Bostjan & Ian
  • - Bostjan (is 2.0 fixed)
  • - Ian
  • - Peter

Open discussion

  • - Ian - COnsider including (collaborative?) mindmapping at an early stage (xmind?)
  • - Erik
  • - Ian - COuld we do a backwards approach ok? Focus on template editing first, before archetypes?
  • - Bostjan , yes templates first
  • - ...
  • - Ian
  • - Erik
  • - Peter