Open Standards Board Meeting 16 September 2021

[KoalaGeo]

Minuets available at - https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1030759/Open_Standards_Board_Meeting_-_16_September_2021__1_.odt

Copy of text:

4. Discussion on UK GEMINI (Geo-spatial Metadata Interoperability Initiative)

• Kieran Millard gave a brief overview of the UK Geo-spatial Metadata Interoperability Initiative (GEMINI).
• Ed raised a couple of points regarding the custodianship of the standard itself and who owns it at this point. Is it AGI?
• Kieran - That is correct. It's governed by AGI. There is an open board. So basically, it can be maintained. You don't have to be an AGI member to be part of the standards board which is governing that as the UK profile. It's also then back to back with the ISO standards, which are governed through the UK's representation to ISO through the BSI.
• Ed - If I wanted to use it, how does that work? And if I am using it, how is that licensed?
• Keiran - It's on an open license. Anyone is free to download and use it. The implementation guidance is published on the AGI website.
• Ed - Having looked at the AGI’s website, there was no membership list of the working group.
• Kieran - No, I think you have to contact the AGI to get the list of the working group. But if you wanted that to be available, I don't see any reason why it can't be published.
• Ed - One final question: it’s a broader issue with these types of data standards. They're reliant on the existence of a catalogue and a data input process.
• Kieran - There are a range of software implementations for catalogues to support this. There are proprietary ones available in software packages.
• Ed - If I was someone in a government department or a citizen wanting to find, for example, some hydrographic data, where would I go? What would I be expected to use?
• Kieran - This is a standard for the metadata. It doesn't prescribe an implementation pattern for how you should deploy catalogs that implement the metadata standard. I think those two things are different.
• Ed - I’m worried this is basically only solving half the problem. You’re creating a mechanism to collect metadata in a consistent way. But there’s a missing piece.
• Kieran - I think that's the same for most metadata standards around the world. This is a standard for the specification of the data, not a specification for the metadata catalogues themselves. And it's also not a mandate for who should deploy metadata catalogues.
• Jacqui - I have lots of concerns, particularly around user needs. I see this as a proprietary approach. It doesn't appear to me to be an open standard because it has proprietary components to it. To get minutes of some of the meetings you need to be a member. I don't consider it to be at the stage of taking the next steps.
• Keiran - Standards should be voluntary. It has over 20 years of development. This is already recommended for geospatial metadata by GDS and widespread implementation already exists. You shouldn’t confuse the purpose, aims and objectives of the standard with the ways people might implement incorrectly. Your concerns are around implementation details. From impact, being adopted and used voluntarily.
• Jacqui - I'm the Flying Binary geospatial specialist too. We have our own offering that doesn't depend on these proprietary routes. It has barriers to the wider market. I have concerns.
• Keiran - I'm sorry Jacqui, I think you are mistaken. There are no proprietary elements. The standard itself is an open document under a Creative Commons license. Probably much more open than that. There are proprietary implementations of that, of course, but there are also open source implementations of this. It doesn't stop anybody developing their own services if they want to.
• Ed - It is a standard. Concerns why AGI is managing it. Perhaps the Geospatial Commission should become custodian of it.
• Kieran - It’s not controlled by AGI.
• Ed - It doesn’t appear too open right now. If I can't find out the membership of the working group, and all the history of the development up to this point, how do we know how we got here?
• Matthew - No problem with the ISO standard. They all require funding and time. What is specific in the Gemini bit? Because that seems to be the bit that's controlled by AGI, as opposed to the ISO agency.
• Keiran - AGI provides an umbrella for the GEMINI group to operate. AGI facilitates the Gemini Working Group to work and operate. It provides a secretariat.
• Matthew - Ok. What does the profile do over the base standard?
• Kieran - ISO 19115 is a very large standard. The Gemini profile makes clear what are the appropriate mandatory and non-mandatory elements for implementation. It reduces the implementation choices. So anyone building services against the Gemini profile doesn't have to worry about how many different organisations have decided which elements they're going to take as mandatory etc. It provides consistency at the UK level and allows pan-UK government services to work across and between the standard.
• Matthew - We need, as a board, to have a fundamental discussion about what we mean by an open standard. Some members have different views. We did go through this a long time ago. Do we include organisations like ISO where there is a paywall to buy the standard? AGI and BSI also. I think we agreed they would be, but we need to come to a consensus as to whether this constitutes an open standard. I do think that with the new board, we need to come to a consensus before we put people like Kieran under the spotlight, where we've got that sort of disagreement as to what we understand by "open standard".
• Linda - The definition was set in government policy and that is the remit within which this Board operates. It would have to go back to public consultation to change the definition. ISO is included in our definition of open standards.
• No decision is taken on GEMINI by the board due to no more time available for this agenda item. The board agreed to continue the conversation on GEMINI after the definition of an open standard had been circulated to all board members and are quorate.

[nmtoken]

I'm surprised that there was no mention of INSPIRE in this conversation

[KoalaGeo]

Key issues:

• [ ] Clarify standard "ownership"
• [ ] Clarify current WG
• [ ] How to join WG
• [ ] Requirements to join WG
• [ ] List some implementations (create metadata & serve metadata)
• [ ] Meeting minuets to be public
• [ ] History of development
• [ ] Clearer user requirements - ease of use over 19115/39

[nmtoken]

• Meeting minuets to be public

  Isn't that what we have now on this platform?

• Clearer user requirements

For some this is a legal requirement (INSPIRE).

ease of use over 19115/39

bad phraseology perhaps? GEMINI isn't over/instead of 19115/39 it is (now) those things +

[PeterParslow]

There was further discussion on the Open Standards Board Slack channel, in which I resolved Ed's concerns - but then I discovered that the channel has a 7-day retention period! So he & I will have to remember what we concluded. I've used my memory of it here.... For example, I think he accepted that AGI has a better record of maintaining GEMINI than any Cabinet Office team has had (e.g. of maintaining the eGovernment Metadata Standard)

I think we can ignore much of what Jacqui said, in that it was resolved during the meeting. However, some things may return. We should also ignore Matthew's point about 'what is an open standard' - you can see that was answered in the meeting.

In my view, the actions for us are:

• To have a link from the AGI GEMINI "home" to this GitHub home - it's try that we're pretty open here, but that can't be found from the "official" AGI GEMINI pages.
• Make the declaration of CC-BY more obvious (it's currently "hidden" on the page that used to be seen as the introduction)
• Ed's (or Jacqui's) broader concern is that GEMINI "only solves part of the problem" of discoverability. Aside from the fact that any standard only solves part of the problem, once we have clearer GEMINI guidance on how to use it in such a way that it improves 'organic search results' as well as portals, I think we'll have done quite a bit of this. See e.g. https://github.com/agiorguk/gemini/issues/42.

I have asked the Open Standards Board if the discussion will continue, or only if the Geospatial Commission brings it back. Parallel to that, Geospatial Commission is in conversation with the Open Standards Board about getting a sort of delegated authority to decide which are the appropriate geospatial standards - they'd use IST/36 to make those decisions. For those of you who aren't IST/36 members, watch this space!

[PeterParslow]

There is now a link from https://www.agi.org.uk/uk-gemini/ to this GitHub repository, so the minutes are now (more) public. I think the site is quite clear how to join this group (https://www.agi.org.uk/agi-uk-gemini/ - needs updating re chair & secretariat). I think the fact that AGI owns GEMINI is quite clear (never understood their idea that it wasn't!)

Leaves:

• list some implementations: see https://github.com/agiorguk/gemini/issues/58
• history of development: lots of detail on this repository, which I summarised for them to address their patent concerns (attached: UK GEMINI Open Standards Board proposal - patent evidence.docx
• clearer user requirements: see https://www.agi.org.uk/why-uk-gemini/?

Should we close this as an issue?