2020-07 Minutes

[sarahcleary]

2020-07 Minutes

Date

Tuesday 5/5/20; 9:30 AEST

Members: Video/Teleconference:

@peter-devoil; @LouisAK; @sarchontoulis; @sarahcleary; @EnliWang; @MarkLieffering; @jbrider; @kchenu; @Keith-Pembleton; @sno036

Apologies/Not in attendance

@JulianneLilley; @yashvirchauhan; @HamishBrownPFR;

1.1 Welcome/Apologies

Welcome from @peter-devoil

1.2 Review Minutes

Review Minutes: 2020-05 Minutes
Taken as read and correct.

20200602_RP Strategy Minutes and Actions.docx
Taken as read and correct

2. Science/Software

2.1 Check for new APSIM Major Improvements

GitHub MAJOR science issues
GitHub ALL science issues
Issues requiring RP Review and/or Oversight - Classic
Issues requiring RP Review - NG
Issues requiring RP Oversight - NG

No additional discussion

2.2 Updates from RP on models and reviews

Models under development

• No new models under development

Models currently under review

• Sorghum model
  progressing with minor amendments - no further update

• Soybean Model
  awaiting final review

• New PMF Gliricidia Model
  Ready for release. A number of minor issues will be fixed as part of this process. @hol353 to progress

• New Stock Model
  @m8harrison will commence review of the Model this month. @peter-devoil is seeking an additional reviewer.

Models requiring review

No new models requiring review

Models/Updates in release since last RP meeting

N/A

3 Software

--Labels for APSIM Initiative Issues
@APSIMInitiative/reference-panel to progress as part of standard processes. Agreed to close the issue

-- Naming of APSIM Next Gen - APSIMInitiative/ApsimX#4690
@APSIMInitiative/reference-panel agreed that although renaming the extension from .apsimx to .apsimng would help with communication, it may cause some errors/issues which will require fixing. At this stage, agreed for the extension to stay as .apsimx and utilise the NG in ALL other communication. Agreed to look at renaming the Repository for communication purposes

-- Guidelines for Development

Discussion around whether retrospective review of models should occur. Agreed that this would be preferable and encouraged but noted availability/capacity to do this, may not be feasible for all models. Mention made that without consistency, it could be seen that APSIM is a number of disparate applications under a single banner. Retrospective review will reduce these inconsistencies. Noted that CLEM won't comply due to the monthly time steps. Guidelines are being set by the @APSIMInitiative/reference-panel and new modules will be required adhere unless there is a valid reason for the module not to comply. An example would be MicroMet which would not require to interact with solutes.

Discussion on examples - what should be expected? Agreed that there should be more effort put into example simulations, especially to assist with training new users to the model. Reduces the learning curve with utilising a new model. Noted that sometimes data isn't available. Agreed that examples are beneficial as it demonstrates functionality. If it isn't 'real' - need to ensure documentation is clear as users may assume validity.

(1) There should be an equivalent of iSow, iKill, iDamage, iHarvest - add to required implementation list.
(2) Include a generic management component to utilise this feature.

ACTION Include a few more required interfaces - sowing and killing the crop. Question is to how to generalise this across all plants.

ACTION @hol353 to finalise with input from @APSIMInitiative/reference-panel if required.

4 Outstanding Actions

To be discussed by exception - Actions to be discussed.

-- Concept Note - #57
No update.

-- Invite colleagues and others to GitHub Org - #58
Discussed above as part of the "Labels for APSIM Initiative Issues" issue

-- Upcoming AI SC Meeting

--- Projects for potential funding

Noted that Implement new version of SWIM is being internally funded by AgR and CSIRO. Outcome should be SWIM into release. May also include work on Port the old apsim SoilTemp2 model. No need to include either as a project for funding from the AI.

Discussion on draft project proposal for Anaerobic capability for nutrients & ponds including the scope and structure. Would be ideal if the SWIM model is completed in NG prior to this work being undertaken. However, Ando's time in constrained to this financial year, so would need to occur concurrently. Agreed to provide to AI SC for review after amendment. Retain one face to face workshop but suggest moving others to virtual. ACTION KP to discuss with Don Gaydon re: required amendments.

Noted that New Chickpea Model - work going to be undertaken internally by CSIRO. No need to include on this list.

Agreed for two more projects to be included on this list - Fixing the Autogeneration NG documentation - @HamishBrownPFR to provide AND NG documentation and training material - @sarchontoulis to provide

It was also noted by @hol353 that CSIRO needs to decommission it’s APSIM build machines. As a result it is time for the APSIM Initiative to cover the costs of building APSIM. This will be an ongoing cost that covers virtual machines running somewhere in the cloud. It might be on NECTAR or somewhere in UQ or CSIROs cloud compute infrastructure in which case the cost might be minimal ($0). Or if that doesn’t work then it might cost somewhere around $15-20K / year. We won’t know until later this year when we investigate the options. ACTION @sarahcleary to include on AI Budget

5 Training

Update on APSIM Week

Discussion on timing/delivery options - Reviewed email trail. Agreed to postpone the event in November. Aim for the week commencing 12/4/21 to provide APSIM Symposium and Advanced Training workshops.
ACTION - confirm dates with key presenters and facilitators.
Discussion on providing an interim 'update' - format? value? Agreed that value from an update - webinars, podcast. Noted that podcast is a low investment level. Webinar could be live but also provided as a recorded file on the website/youtube. Agreed that any of these would be useful and encourage anyone to progress these. Preferable option would be to provide an 'update' webinar to cover some of the new developments.
ACTION - @hol353 and @Keith-Pembleton to progress
Reference to above issue and action for @sarchontoulis to provide a half pager for the AI SC to review. @sarchontoulis to work with key trainers/RP reps to review best way of developing NG documentation/training materials.
ACTION to provide an overview of the ISU workshops/format/structure in the August RP meeting

6 RP Roles and Responsibilities

Discussion as per emails. Key discussion points:
-- 2 members often difficult for an organisation to provide
-- RP representatives - active developers, high end users and expected to have cross domain interest in APSIM.
-- no of members should be directly related to the function of the panel.
-- Review of decision making model. More much of the procedural decisions of the RP are now being made via GitHub and should continue to.
-- Sufficient numbers are still required to provide diversity of opinion
-- No. of meetings to be reduced.
-- The more complicated the issue, the greater number of people required - does not necessarily need to be RP reps but relevant people to the issue.
-- RP role includes building the APSIM community. RP to bring more active users to the GitHub organisation. Noted that a lot of peer review occurs now and much of it by non-RP representatives.
-- Should there be an FTE commitment? Discussion with the AI SC
-- Is there a way to better fund projects/sprints - both time and travel.
-- Recommend
--- Drop down to only requiring 1 member per organisation (can be more) who has a broad interest in APSIM, not just focus in one area.
--- Attempt to reach consensus.
--- No comment indicates agreement
--- Online decisions require time limits to comment. If decision can't be made within the time limit, a meeting would be called, or part of the agenda of the next RP meeting
--- Minimum of 3 or 4 times a year; preferably a month prior to the AI SC meeting.
--- Would require clear GitHub decision making processes and active involvement from the RP reps. At least weekly review of the GitHub issues and comment on RP relevant issues.
--- Face to face (ad-hoc) meetings to include science/development presentations/discussions from the model developers.

ACTION: @sarahcleary to provide a revised RP roles and responsibilities for the AI RP and SC to review.

7 Other Business

No further issues discussed due to time.

8 Next Meeting

4 August 2020 - 9:30am