Setting the scene - 2017-11-29

Brainstorming...

Picking on one bullet on slide 10 although it does relate to bullet 1 as well:

• Maintenance of donated content

• Scaling/QA/Gateway/approval

From an authoring point of view, technical requirements notwithstanding, posit: could we have a separate release branch? I.e. One which is published in real-time containing urgently needed content i.e. I am implementing yesterday and I am missing this and need it today.

For this, we could organise the content team into those who are on the "urgent request desk" triaging requests or perhaps reviewing and approving content modelled by NRCs and submitted for publication.

Could we have an author seconded to an NRC to oversee content creation?

A separate team could work on the larger more complex big pieces of work e.g. as we have now LOINC, Drugs etc. 

At the same time, other improvements, updates work requiring more analysis are managed in a branch which is published or rebased to the "DAILY/WEEKLY/MONTHLY" release when ready?

Thanks for starting this off Rory, I went through the recording last week but only getting around to providing some comment now (trying to make sense of my notes now..)

You kind of cover two problems here (though it depends on how you choose to slice it)

1. Distributed authoring with a range of folk authoring, and SI providing QA of the content before release.

2. Promotion of content already authored in extensions. (Though you could frame this the same as 1.).

It hints at a future where there is only one monolythic authoring platform - the managed service provided by SI. Is that the endgame? There's pros and cons to that..

I've got a note about "RF2 is not incompatible with high frequency releases"... RF2 can at least be published as frequently as daily. I'm not sure if time is allowed in the effectiveTime field (it seems like it's viable ?) if it is, you could release every second..

I think we agree the technical solutions reasonably straight forward. It's the actual content QA that is tricky..

Jim (I think) made a really valid point that SI is responsible for the content it publishes. However if SI staff are the only ones who can QA content published, that's remains bottleneck/scalability.
And there's nothing wrong with that, but if we wanted to scale beyond that there needs to be some minimum requirements defined, that extension builders have meet E.g. qualifications, authoring processes, dual review etc.

I've mentioned elsewhere that extension builders are somewhat like franchisees of the SNOMED CT franchise. And the quality of a given extension is a reflection of SNOMED CT in a given jurisdiction. There's currently limited quality oversight of extensions (but I think this is where the RVF has a role).

Notwithstanding the request for a small WG, wondering if I could join please - as a representative of a country just starting a national extension, wanting to keep the extension contained and using the International Edition as much as possible.