OpenWIS Steering Committee 2020 December
7th December 2020 - MS Teams video-conference
1. Welcome and introductions
- RGr – introduced the meeting.
2. Attendance
2.1 Current members that were present
- Rémy Giraud (RGr), Météo France [MF] - SC Chair
- Jeremy Tandy (JT), Met Office [UKMO] - SC Vice Chair
- Loïc Le Gallou (LG), Météo France International [MFI]
- Sungsoo Do (SSDo), Korea Meteorological Administration [KMA]
- Kari Sheets (KS), National Weather Service [NWS]
- Mikko Visa (MV), Finnish Meteorological Institute [FMI]
- Weiqing Qu (WQ), Bureau of Meteorology (BoM)
2.2 Members not present
- Mr. Y K Reddy, India Meteorological Department (IMD)
2.3 Other attendees
- Steve Olson (SO), National Weather Service [NWS]
- Jude Anthonisz (JA), OpenWIS Association Secretariat
3. Approval of the agenda
- All - No changes or addition to the agenda.
4. Approval of previous minutes
- All - No changes. Therefore approved.
5. OpenCDMS
Update on discussions with AWS.
- JT - I had intended to begin discussions with AWS. However, I was unable to have the intended discussion with AWS in November 2021 due to WMO commitments and changes in MO. So, I expect things to begin soon, but no update at this point.
- JT - The changes to the Articles of Association are published on the Association site.
- RGr - But not in Belgian Gazette
- JT - Need to be clear about the definition of a Contributing Partner that is then reflected in the Internal Rules
- RGr - My definition of a Contributing Partner is someone giving money and not having a vote.
- JT - We will need to understand how this would be of benefit to AWS.
- RGr - Philanthropic approach of any Contributing Partner without a role in the running of the Association.
- JT - That would be like an Associate Partner who wouldn’t vote at the SC.
- WQ - So a Contributing Partner is someone who signs a CLA?
- RGr - Not really, the intention was to enable a partner to provide funding and share common goals, backed up via an MOU. So, shouldn’t attend the SC.
- WQ - Make sense to me.
- JT - It’s clear, but not sure what are protecting if they can listen but not vote. Maybe we need a longer discussion.
- RGr - I can update on work with Belgian lawyer re formalising the changes in the By-laws. She has suggested that a clear statement that everything can be done electronically. Other changes are proposed, so I am reviewing the document. When a final version is available, I will share with Michael Robbins before finalisation (cost less than 2K). The goal is to be able have a final version in early 2021. So, the decision on what a Contributing Partner can be made by March 2021.
- JT - So talk at the Annual Meeting?
- RGr - We can draft beforehand and endorse at the Annual Meeting
- JT- Six month’s may be a long time for AWS. So, we need to think about how to engage with AWS beforehand.
- RGr - Agreed. Maybe we have an MOU but not an Internal Rule for all contributing partners, perhaps attending a PMC.
- JT - Agreed. So, we can talk (RGR and then play back to SC)
ACTION – JT ad RGr to agree approach to engage AWS between now and May 2021
- RGr - What are other views?
- SO - PMC activity is relevant.
- WQ - Can contribute like any other non-members but not influence.
- RGr - Probably the key difference is whether can be an observer in the SC or not?
- WQ - I would say no as they will be a privileged compared to an Associate Member.
- LL - Like a sponsor. Code and development?
- RGr - No just money.
- LL - Well if a sponsor then another option to participate.
- MV - agreeing with WQ.
- JT - More privilege, what were you thinking?
- JT - seeking to avoid a commercial organisation take control. But cannot see how an Associate Partner can influence as they cannot vote.
- WQ - it’s about the motive when money is given.
- JA - we could learn via starting with an MOU and then decide the and evolve Internal Rules later, as there could be other partners in the future.
- JT - OK
- RGr - JT and I need to consider all aspects when defining the Internal Rules.
- JT - We should also look at what other organisations do and see what contractual vehicle AWS needs/expects and look at all options to make a consensual decision.
6. OpenWIS 3.x
- SO - Progress made on Keycloak and meeting set up to get an update on Singapore code progress and resources available. So, based on this, the timeline is as per in the issue 593.
- JA - What is the extent of code that would covered by the Singapore code compared to the existing RFP.
- SO - about 80%.
- LL - So, the RFP should be lighter than previously?
- SO - Yes.
- RGr - Anything in the budget for 2021?
- SO - I think a certain amount was allocated, based on the July 2020 SC meeting.
- LL - Checking figures now.
- JT - From previous experience the quality of code previous generated a lot of remedial work. So, can we quality a quality review of the Singapore code ?
- SO - Yes. We are looking at the code and the implications for maintenance and patching.
- JT - Very interested in a short report.
- JA - When would we need this report in relation the envisaged timeline ?
- JT - I believe that we need this report in March 2021 (before a contract is agreed)?
- SO - Agreed
- ALL - Agreed.
- JA - SO do you have availability of people to do this in the TC?
- SO - We have 14 contributors.
- LL - 10k was allocated in 2020 to start the project. During last SC this was shifted to 2021 but without a defined total amount. In 2019 at an SC, 20k was indicated. We need to analyse the RFC to be more precise.
- JT - Clearly, we need to assign more to the budget. I had envisaged in being 80-100k, but this seems a lot lower now. Is there any indication of what the charges would be for a similar work?
- LL - Not sure.
- SO - We have an indication of 40-60k Euros.
- JA - total or addition?
- SO - Total.
- JT - Suggest 50k with variance of plus or minus of 10k Euros.
- RGr - Budget is for Board to approve.
- JT - Yes but what is the process for dealing with variance.
- RGr - Based on information known, we should not go more than that. When we have more information in late April 2021. So, we could have special meeting or wait until the Annual Meeting for an update to the SC from the TC and with subsequent approval by the Board.
- JT - I am comfortable with the process. Conscious that we won’t get a fixed price as FMI’s charge via its supplier will be for time and material.
- RGr - We should aim to work with a fixed price.
- WQ - Plus additional fixed charge from FMI’s supplier?
- RGr - Yes.
- LL - From a budget point of view its about agreeing a value.
- JT - So the question is when the SC should make a recommendation to Board.
- RGr - When we have more information, say mid to late April 2021, rather than a figure now.
- JT - Agree, it would be good to make a note for the budget so that we are clear about previous assumptions when the proposal comes to the SC and Board for approvals.
- All - Agreed ACTION – LL to make note of the current estimate in the 2021 budget
- JT - Any more questions for SO?
- All - None.
7. Procurement routes and partners
- JT - mentioned the three options being pursued, i.e. potential procurement services via either - FMI, Belgian, or UKMO.
- RGr - Daniel (Director) says OK to work with Belgium but need to publish tender in French and Dutch. So first in English and pay for translation. Practically, answering/translating from Dutch could be difficult. Could be done by IRM.
- JT - So, a possibility with some challenges.
- MV - Following a competitive procurement, two framework agreements are in place that could be used (CGI
and EWork Group). So, should be possible to sub-contract directly with any company. E-work has fixed price rates. - JT - A few questions around how this would work. So, we wouldn’t be able to tell e-work to choose a company?
- MV - It would be easy to choose a supplier.
- JT - this makes it easy for the Association. What everyone’s view?
- RGr - if we choose a supplier directly without a procurement, then a problem. So, using FMI means that we can demonstrate that we are following EU rules and no risk.
- MV - should be no problem as competition has been done. MV to check with FMI
Action – MV to check with FMI whether it is possible to choose a supplier directly via Ework or CGI
- JT - Internal Rules means we need to demonstrate due diligence. Price may be fixed. But we should assess capability. How do we manage that ?
- RGr - Mikko, how would you proceed normally MV for FMI, what is your process?
- MV - We describe what we need and what skills we need. We then get a bunch of CVs and may have an interview. Mostly we buy man-hours and not fixed price.
- RGr - So, talk with many parties and then ask EWork?
- MV – Yes
- RGr - Talk with people who do the work?
- MV - Yes
- RGr - So if we work with FMI. We issue the ITT to CGI and E-work. You should speak to a contractor?
- MV - No we can choose EWork.
- RGr - for due diligence can approach both.
- MV - Yes
- JT - Yes that make sense, but to nuance slightly, we use EWork as an agent to look at their suppliers and then select sub-contractor.
- MV - Yes
- JT - So, we could also ask CGI for to do the same. How much more effort?
- MV - Not much effort. Just a Statement of Works (SOW).
- JT - Pending final confirmation. This seems reasonable.
- JA - Any language translation needed?
- MV - No all use English
- RGr - Can ask and receive questions in English.
- JT - Any more questions?
- SO - If we go via FMI. It would be good to have the necessary templates for RFP etc. Can you send the templates that would be needed?
- MV - at this is point no templates . Just basic info, such as what needs to be done, when etc which can be emailed.
- JT - any more questions
- All - None
- JT - UKMO update. Our Procurement team do not have capacity. So, I am working with our Technology Department colleagues to see whether work for the Association can be done instead of something else. Not a promising situation.
- JT - feels like an SC activity to decide the procurement partner, so no need for Board.
- RGr - Agreed.
- JT - So the decision we need to make is whether FMI to takes on procurement of middleware.
- RGr - We should check whether Ework, FMI and the people who developed the Singapore code are happy work together.
- LL - Is a contract in place with EWork already and does it need to be based in Finland?
- MV - No and No.
- JT - What kind of timescales? How long does it normally take once there is a SOW?
- MV - Mostly depends on sub-contractor, about a month typically.
- JT - How long does it take to start work (i.e. agree a contract)?
- LL - Likely to February or March 2021?
- SO - Planning for end of May 2021 as a kick-off.
- JT - Let’s say you had a new piece of work. What is the turnaround?
- LL - Generally, about one month.
- JT - So, the four-way meeting doesn’t need to be in 2020 and suggest having the meeting by end of January 2021.
- RGr - That would be OK and have the agreement by end January 2021, i.e. so we know viability by end January 2021.
- JT - So re the SC Decision, do you want to defer until agreement in place or agree in principle now?
- RGr - Vote now in principle subject to agreement being reached.
- JT - Seems OK unless there is a significant variation.
- RGr - So, the conclusion, pending confirmation that Ework and the developers of the Singapore code are happy to work along the lines discussed. SC could approve to proceed with procurement via FMI, Ework and sub-contracting with the developers of the Singapore code.
- JT - Agree that this is the decision to be made by the SC. Vote – unanimously agreed.
RESOLUTION RESOLUTION the SC agreed to proceed with using FMI as the procurement partner of the middleware development requirements for OpenWIS 3.x
Action LL to explain approach to the developers of the Singapore code before end December 2020
8. CLA agreement with NOAA/NWS R
-
JT - I chased MR. KS any info from Fred - KS - No response from Fred on leave. Will chase - JT - Not being able to progress due to capacity of Michael Robbins and myself, a similar position in NOAA/ NWS. However, Michael has a better understanding of the issues and is working with Derrick. - JT - I also asked Michael Robbins whether everything should be on same CLA. In summary, Michael’s view is:
• Avoid a proliferation of CLAs if possible.
• Hope to agree a version with NWS/NOAA that can be implemented that is universal, i.e. something that all
new contributors can be asked to sign up to from that point forward.• For transition to the new agreement, the preference is to leave existing CLA alone until receiving new contributions are received (i.e. “From x date there will be a new CLA. Anyone contributing after that date should sign it”). This is easier to administer than having several old CLAs laying around and trying to work out who’s on which version.
- JT - My summary is that there is ongoing work with NOAA/NWS, so can we encourage Fred and Derrick ‘move faster’. JA and I will do the same in UKMO, with a view to completing by June 2021.
- JT - Questions and thoughts?
- RGr - OK with timeline. Just a concern as OpenCDMS may need CLAs in place sooner that the timeline we are proposing.
- JT - I will take an action to confirm with the OpenCDMS team.
- RGr - I think OpenCDMS will need a CLA in place according to Denis Struber. So, I propose that we confirm with OpenCDMS whether a new one is required or not (i.e. to cover software development; database structure; and documentation).
- JT - We could do an amended CLA for OpenCDMS and then sign a new one when there is an agreement with NWS in June 2021.
- RGr - If we have many that need to sign then may be an issue but if only a few, then may not be an issue of OpenCDMS.
- JT - So get everyone to the latest version between June and September 2021?
- RGr - Yes, no need to get everyone to the interim version. Ask OpenCDMS but do not mandate.
- RGr - Also ask Peer Hechler whether a CLA is expected for OpenCDMS. We need to be not blocking OpenCDMS. JT - Work is going ahead, so not a blocking situation
[ACTION – JT to update clauses referring with documentation and data assets?] (https://github.com/OpenWIS/openwis-documentation/issues/494#issuecomment-740509976)
- JT - Is a decision needed by SC?
- RGr - I don’t think so. SC would approve the interim version once created. Peer would confirm whether an interim version is needed and when. So, at present a formal vote is not needed.
- All - None
- RGr - My concern is if the final or interim version is US specific, then do we need one specific for US people and another for everyone else?
- WQ - What is the problem for the Association for having two different CLAs?
- JT - For example, if BoM and NWS jointly developed code and a dispute arises, the Association needs to be clear on which court is used to resolve the dispute.
- WQ - Understood.
- JT - Any more questions ?
-
All - No. 9. Summary of recommendations to the Board
- None.
10. Next meeting
- JT - Tentatively schedule a meeting for late March 2021 to confirm arrangements with FMI and contractor and the Internal Rules for Contributing Partners.
- All – agreed.
ACTION JA carry out a doodle poll 22 March plus two weeks to firm up the date for the next SC.
21. Closure of the meeting - 1425 UTC