My SED-ML Level 2 Proposal

[Lucian Smith]

(Note:  I'm posting this both here and on the SED-ML Issue Tracker at https://github.com/SED-ML/sed-ml/issues/217  Feel free to reply either there or here!)

Hi, everyone!  At COMBINE, I presented a proposal of what SED-ML Level 2 could look like.  If you weren't there, or if you want a recap, we have:

* The talk itself:  https://www.youtube.com/watch?v=GvJKdROtILA (the playback speed option is your friend!)

* The slides:  https://docs.google.com/presentation/d/1X8bhK8eFxUInoV1tksGaFPmn3NB913-w/edit

* The UML diagrams:  https://drive.google.com/drive/folders/1RFAGw6eL5JhSR4XgdQSI6wAap6fplWnv

* Examples: https://docs.google.com/document/d/1jZkaNhM_cOqMWtd4sJZ9b0VGXPTLsDKsRNI5Yvu4nOA/edit#heading=h.gbwg6eltvdph

Some of those examples are in the slides/talk, but many are new; I was asked at COMBINE to come up with some more examples, and they're there.

What happens next depends on you!  I need to know:

Is this approach worth pursuing further?

If you don't have an answer to that, then I have a different question instead:

What additional information do you need to make that decision?

For the purpose of these questions, 'this approach' is:

* Moving to a procedural/script based way of storing the simulation experiment.

* Keeping a lot of the trappings of existing SED-ML.

* Storing information in a way that could remain as XML, or could be switched to Yaml or a database like JSON:  this proposal is agnostic on that front.

If you think the approach in general is OK, but disagree/have opinions about some of the specifics, I definitely want to hear that, too, but those discussion are moot if we don't want to go this direction in general.