10:01:38	 From Shelley Stall : Notes:  https://docs.google.com/document/d/1foNeWiep57leu-JEDyIG0S3Gyuvk30WS3x9PT1JOzFg/edit
10:03:24	 From Deb Agarwal : folder with the materials for the meeting https://drive.google.com/drive/folders/1UOW7oq-hCYzWhgsTLV6aTHNpyEd_oMKB?usp=sharing
10:13:55	 From Caroline Coward : I like the image…
10:19:36	 From Maggie Hellström : I added the link to the folder at the top of the collaborative notes - see https://docs.google.com/document/d/1foNeWiep57leu-JEDyIG0S3Gyuvk30WS3x9PT1JOzFg/edit (also sign in if you didn't already ;>)
10:31:12	 From Maggie Hellström : An issue possibly related to Deb's Ameriflux use case: probably many reliquaries would be including also older legacy datasets. Many of these do not (in my experience) have complete metadata on "authorship" (and other roles) included in their (DataCite) catalog entries. I see a need for repositories to revisit these older datasets and, if possible in collaboration with the original data producers, enhance/update the missing/outdated information.
10:41:09	 From Caroline Coward : DOI vs Handle…how to reconcile the two in a single document or repository.
10:41:54	 From Caroline Coward : There are reasons to use one or the other, but for some items it could go either way.
10:42:48	 From Bruce Wilson : Internally, where we are tracking citation, we have fields for persistent identifier and identifier type.
10:43:02	 From Kirsten Elger : @Caroline: if you use the DataCite relatedIdentifier, you Need to define identifierType (DOI, Handle, IGSN..) and relationType (e.g. HasPart, IsCitedBy). DataCite doesn't make a difference between PID types as Long as they are described via the valid metadata xml
10:43:05	 From Mark Parsons : Arks are very useful for finer granules
10:43:46	 From Elisha Wood-Charlson : Naive Q, maybe - are all DOIs handles, but not all handles  are DOIs?
10:43:56	 From Mark Parsons : Correct Elisha
10:44:03	 From Elisha Wood-Charlson : 🙏🏼
10:44:24	 From Caroline Coward : I'm enjoying this particular rabbit hole...
10:45:09	 From Maggie Hellström : Of course also "datacite dois" are Handles; I think there may be a needed to distinguish between "basic Handles" and "DOIs". In my RI, we assign basic Handles to all digital objects - this allows workflows to both find & use data appropriately, and e.g. record that usage in associated provenance records.
10:45:56	 From Elisha Wood-Charlson : And DOIs are “gold star" mostly b/c they have agreed upon standardized metadata to describe the object, yes?
10:46:29	 From Kirsten Elger : yes, absolutely, Elisha
10:46:48	 From Bruce Wilson : And because of the social contact for the long-term persistence of a DOI, even if the underlying object is no longer available.  There should at least be a tombstone page to which the DOI resolves.
10:47:46	 From Caroline Coward : Just like link ROT, DOI ROT is a thing.
10:47:49	 From Maggie Hellström : (Sorry, hit return before I was done typing :>) At the same time, we assign DataCite DOIs to finalized datasets - i.e. the ones normally used (and cited) by our end user communities. But we are still struggling with adding really  rich metadata about these "published" datasets - both in the DataCite catalogue records and in our own extended metadata store.
10:49:18	 From Mark Parsons : Amen. PID require institutional maintenance. See Klump et al.
10:49:34	 From Martina Stockhause : We also have tools creating provenance records based on HandleIDs of the datasets. This provenance record could be published and referenced in the reliquary. We need good tools which take care of this important information on product creation.
10:50:24	 From Elisha Wood-Charlson : @Maggie, yeah, we have rich metadata for things like samples that doesn't fit into "DOI”.  So I wonder if we should consider how to converting particular "handles" to DOI-quality instead of pushing the other way.
10:50:59	 From Maggie Hellström : @Caroline: very true! In fact, there are many calls for a deepened discussion about what "persistent" should mean - both for the identifiers (and associated metadata at e.g. registry level) and the digital objects that need to be PID:ed. Expectations seem to differ between the various stakeholders involved. This will have an impact also on the reliquaries, both at concept level and in practice.
10:52:14	 From Caroline Coward : @Martina Self organizing AND siloed within organizations. The left hand is not aware of the right hand and the work they are doing.
10:54:13	 From Mark Parsons : Again RDA Rec is good for this.
10:54:55	 From Mark Parsons : Y’all really need to read: Rauber, A., B. Gößwein, C. M. Zwölf, C. Schubert, F. Wörister, J. Duncan, K. Flicker, K. Zettsu, K. Meixner, L. D. McIntosh, R. Jenkyns, S. Pröll, T. Miksa, and M. A. Parsons. 2021. “Precisely and Persistently Identifying and Citing Arbitrary Subsets of Dynamic Data.” Harvard Data Science Review 3 (4): https://doi.org/10.1162/99608f92.be565013.
10:55:21	 From Bruce Wilson : There is a difference in changing the use case where there is an error correction versus adding something to the collection.  What happens with a reliquary, for example when there is a name change for one of teh authors of one of the component pieces.
10:55:48	 From Martina Stockhause : @Caroline: Sometimes it is just too much work under time pressure, where the relevance of data documentation does not have the highest priority.
10:56:34	 From Maggie Hellström : To me, the reliquary content description is actually a kind of provenance in that it provides a record of items that someone at some time decided belonged together. But I wonder what happens later on, after that first paper citing it was published. Other end users might continue to reuse the reliquary definition (and cite it) even if only a subset of the items were used in these (follow-up) papers. Is that OK? Do publishers have to go through the details and ask authors to make sure that the cited reliquaries are all relevant for the paper?
10:57:12	 From Maggie Hellström : "...the cited reliquary's contents..."
10:58:51	 From Kelly Stathis : I have to drop off at the hour - very interesting discussion, looking forward to following this work as it progresses!
10:59:42	 From Martina Stockhause : Sorry that I have to leave. Versioning and recommendations for dealing with reliquary versions is missing and important,
10:59:51	 From Elisha Wood-Charlson : same, and yes thanks all - super interesting and important.
11:01:41	 From Bruce Wilson : Need to drop off.
11:03:34	 From Matthew Cannon : would there need to be a specific reference type made, so it would be clear in a paper's references that someone was citing a reliquary rather than a static or dynamic data set?
11:04:29	 From Shelley Stall : Matt, I’m thinking it would be a bracketed description of “[Collection]”
11:04:55	 From Matthew Cannon : okay, thanks!
11:05:46	 From James Ayliffe : Hi, sorry or missing the start, dropped in on James's zoom link so the called me James too (but it is Justin)
11:06:00	 From Caroline Coward : Hi Justin!
11:06:30	 From Mark Parsons : To be clear, Maggie is talking about a different RDA Rec than I was
11:07:40	 From Madison Langseth : But isn't the RDA dynamic data citation intended for systems to create whereas the reliquary is intended for an individual author to create?
11:10:49	 From Maggie Hellström : Sorry Mark! To be clear, I'm referring to DOI: 10.15497/RDA00022 from the RDA Research Data Collections WG that ended in 2017
11:10:56	 From Mark Parsons : @Madison the human makes the query, the system then makes the dOI
11:15:15	 From Maggie Hellström : I think there should be a Type in the equivalent of the PID Registry of the Handle system that allows to indicate if a digital object is a Reliquary (or other types of collections) - this would flag it to indexers, and they can take appropriate action.
11:16:51	 From Caroline Coward : @Mark 🥯☕
11:17:24	 From Maggie Hellström : What is much more challenging IMHO is the "reverse lookup" - i.e. to be able to see in which reliquaries a specific data object has been included!
11:17:48	 From Caroline Coward : @Maggie Yes! It has to work from both directions.
11:21:19	 From Maggie Hellström : I've been told some time ago that if one limits oneself to only Handle-based PIDs, this could be achieved relatively easily per se, but would require a new layer of automated analyses and caching that will be quite costly from a IT resource point of view. I'm not sure how true this is today - but unless there are very well-thought-out use cases proposed my major stakeholders, probably nothing will be implemented
11:23:40	 From Caroline Coward : We wear our print publication model culture like a comfortable pair of slippers.
11:24:06	 From Maggie Hellström : Personally, I'm more interested in how these reliquaries can be leveraged in data processing contexts.
11:24:45	 From Mark Parsons : Those slippers have worn out, Caroline :-)
11:24:50	 From Andreas Czerniak : OpenAIRE Resrach Graph link objects together -- https://graph.openaire.eu/ and would like to give the credits.
11:25:02	 From Mark Parsons : +1 Andreas
11:25:31	 From Shelley Stall : +1 Andreas
11:25:55	 From Maggie Hellström : @Mark: https://ae01.alicdn.com/kf/Hd8efeb51449e4fb0a2d81b08f2911661f/Women-Cute-Animal-slippers-Girls-Rabbit-Home-shoes-Big-size-42-Non-slip-Flat-with-Winter.jpg_640x640q80.jpg.webp?1000
11:26:20	 From Mark Parsons : {Giggle}
11:27:59	 From Maggie Hellström : @Andreas: exactly. So we need tools that can "open up" the reliquaries and populate a composite graph that includes all linkages for each of the relics contained within.
11:33:29	 From Melissa Cuthill : I'm from ONC - I'll tell Reyna you want to talk to her
11:33:39	 From Justin Buck : For IDW the time is tricky for remote attendance from the UK, will be watching/viewing the content after the event
11:33:44	 From Mark Parsons : Thanks Melissa
11:34:03	 From Maggie Hellström : Reyna is of course also very much involved in the RDA Granularity WG,!
11:37:14	 From Caroline Coward : We need a good analogy for when we try to explain this to friends and family.
11:37:24	 From Caroline Coward : Or get comfortable with the eye roll and glaze...
11:38:12	 From Andreas Czerniak : I'm happy to contribute also
11:43:46	 From Maggie Hellström : I would be happy to join that second group you mentioned, Deb, to look more deeply into real use cases
11:45:28	 From Kirsten Elger : I Need to leave, great discussion and see you all soon!
11:46:13	 From Mark Parsons : Hopefully RDA can help with that. Will y’all have a session in Seoul?
11:46:42	 From Mark Parsons : cool
11:48:21	 From Mark Parsons : There could be millions, but they only need to be indexed when cited. See MAST example.
11:48:37	 From Andreas Czerniak : Is there an list of existing, open endpoints of reliquaries available and ready to share?
11:52:04	 From Andreas Czerniak : +1 Mark
11:53:07	 From Caroline Coward : I need to jump to the next meeting. Thank you everyone for a great discussion today. We have some exciting directions in our future. Onward!
11:56:40	 From Christine Laney : Thank you everyone!
11:56:52	 From Maggie Hellström : Yes, great discussions!
11:57:02	 From Andreas Czerniak : Thank you all , very interesting meeting and happy to contribute -- good bye