DwC "picklist" extension to MIxS IRIs
Dear TG, Dear Ramona (GSC CIG), Chris, Bill, In our DwC-MIxS mapping activity, we noted a practice to support the IPT to generate "picklists" (for drop-down menus etc) by adding extensions to IRIs. This was suggested as a means to link the values to keys in terminology-focused fields. You can see the link to the "picklists" in the tracker: https://github.com/tdwg/gbwg/tree/main/dwc-mixs/dwc#vocabularies An example is available in [1] and looks like: dc:URI='https://w3id.org/gensc/terms/MIXS:0000005/contigs' I'm not quite sure about this and would like your input. To me, it seems like a very strange way to modify an identifier *issued by another standards group* in order to help a tool generate dropdown menu. It essential creates a false impression that there's something standardised on the other end of that IRI that may or may not be the case. Best, Pier Luigi [1] https://github.com/tdwg/gbwg/blob/main/dwc-mixs/dwc/vocabulary/contam_screen... -- https://orcid.org/0000-0002-4366-3088
Thanks, Pier Just for a bit of background - you're assuming these are IRIs, but the DwC-A work was all designed at a time when URNs were common and expectations of resolvability weren't as high as today. The original design was really just to ensure a globally unique URN was used. I think the concerns you raise are valid, and especially with the current intent to achieve resolvability for all the w3id.org/... terms. We do need a URN though, and I understand there isn't an intention to make resolvable terms for these concepts. The rs.gbif.org namespace has been used for this kind of thing in the past, which might be an alternative option to lower expectations. Thanks, Tim On 22/06/2021, 16.02, "dwc-mixs on behalf of Pier Luigi Buttigieg" <dwc-mixs-bounces@lists.tdwg.org on behalf of pier.buttigieg@awi.de> wrote: Dear TG, Dear Ramona (GSC CIG), Chris, Bill, In our DwC-MIxS mapping activity, we noted a practice to support the IPT to generate "picklists" (for drop-down menus etc) by adding extensions to IRIs. This was suggested as a means to link the values to keys in terminology-focused fields. You can see the link to the "picklists" in the tracker: https://github.com/tdwg/gbwg/tree/main/dwc-mixs/dwc#vocabularies An example is available in [1] and looks like: dc:URI='https://w3id.org/gensc/terms/MIXS:0000005/contigs' I'm not quite sure about this and would like your input. To me, it seems like a very strange way to modify an identifier *issued by another standards group* in order to help a tool generate dropdown menu. It essential creates a false impression that there's something standardised on the other end of that IRI that may or may not be the case. Best, Pier Luigi [1] https://github.com/tdwg/gbwg/blob/main/dwc-mixs/dwc/vocabulary/contam_screen... -- https://orcid.org/0000-0002-4366-3088 _______________________________________________ dwc-mixs mailing list dwc-mixs@lists.tdwg.org http://lists.tdwg.org/mailman/listinfo/dwc-mixs
Thanks Tim, this clarifies a great deal. Measures to make it clear that the URNs (wherever they appear, e.g. on documentation pages, lists...) used by legacy systems are not to be understood as resolvable or under the control/custodianship of the GSC would help. While technically valid, prefixing URNs with HTTP does indeed set a false expectation - anything that can mitigate that would be appreciated. On the GSC and TDWG side however, it would be very interesting to find out how to do this more sustainably by, e.g., sharing vocabulary files with all the metadata needed to, e.g., map to ontologies, qualify terms with comments, include provenance data, etc On 22/06/2021 17:18, Tim Robertson wrote:
Thanks, Pier
Just for a bit of background - you're assuming these are IRIs, but the DwC-A work was all designed at a time when URNs were common and expectations of resolvability weren't as high as today. The original design was really just to ensure a globally unique URN was used.
I think the concerns you raise are valid, and especially with the current intent to achieve resolvability for all the w3id.org/... terms. We do need a URN though, and I understand there isn't an intention to make resolvable terms for these concepts. The rs.gbif.org namespace has been used for this kind of thing in the past, which might be an alternative option to lower expectations.
Thanks, Tim
On 22/06/2021, 16.02, "dwc-mixs on behalf of Pier Luigi Buttigieg" <dwc-mixs-bounces@lists.tdwg.org on behalf of pier.buttigieg@awi.de> wrote:
Dear TG, Dear Ramona (GSC CIG), Chris, Bill,
In our DwC-MIxS mapping activity, we noted a practice to support the IPT to generate "picklists" (for drop-down menus etc) by adding extensions to IRIs.
This was suggested as a means to link the values to keys in terminology-focused fields.
You can see the link to the "picklists" in the tracker: https://github.com/tdwg/gbwg/tree/main/dwc-mixs/dwc#vocabularies
An example is available in [1] and looks like: dc:URI='https://w3id.org/gensc/terms/MIXS:0000005/contigs'
I'm not quite sure about this and would like your input. To me, it seems like a very strange way to modify an identifier *issued by another standards group* in order to help a tool generate dropdown menu.
It essential creates a false impression that there's something standardised on the other end of that IRI that may or may not be the case.
Best,
Pier Luigi
[1] https://github.com/tdwg/gbwg/blob/main/dwc-mixs/dwc/vocabulary/contam_screen...
-- https://orcid.org/0000-0002-4366-3088
_______________________________________________ dwc-mixs mailing list dwc-mixs@lists.tdwg.org http://lists.tdwg.org/mailman/listinfo/dwc-mixs
Thanks, Pier We'll clearly state expectations in the description, and this will be kept as internal as it can be. Using the example of the vocabulary of controlled terms for lib_layout[1] currently we have: https://w3id.org/gensc/terms/MIXS:0000041/vector This could become something like the following e.g. https://rs.gbif.org/vocab/dna/lib_layout/vector Would this be preferable to the GSC folk to avoid confusion, please? Note that the source from where it is derived will still be cited. Thanks, Tim [1] https://rs.gbif.org/sandbox/vocabulary/mixs/lib_layout.xml On 22/06/2021, 17.41, "Pier Luigi Buttigieg" <pier.buttigieg@awi.de> wrote: Thanks Tim, this clarifies a great deal. Measures to make it clear that the URNs (wherever they appear, e.g. on documentation pages, lists...) used by legacy systems are not to be understood as resolvable or under the control/custodianship of the GSC would help. While technically valid, prefixing URNs with HTTP does indeed set a false expectation - anything that can mitigate that would be appreciated. On the GSC and TDWG side however, it would be very interesting to find out how to do this more sustainably by, e.g., sharing vocabulary files with all the metadata needed to, e.g., map to ontologies, qualify terms with comments, include provenance data, etc On 22/06/2021 17:18, Tim Robertson wrote: > Thanks, Pier > > Just for a bit of background - you're assuming these are IRIs, but the DwC-A work was all designed at a time when URNs were common and expectations of resolvability weren't as high as today. > The original design was really just to ensure a globally unique URN was used. > > I think the concerns you raise are valid, and especially with the current intent to achieve resolvability for all the w3id.org/... terms. > We do need a URN though, and I understand there isn't an intention to make resolvable terms for these concepts. The rs.gbif.org namespace has been used for this kind of thing in the past, which might be an alternative option to lower expectations. > > Thanks, > Tim > > > > On 22/06/2021, 16.02, "dwc-mixs on behalf of Pier Luigi Buttigieg" <dwc-mixs-bounces@lists.tdwg.org on behalf of pier.buttigieg@awi.de> wrote: > > Dear TG, Dear Ramona (GSC CIG), Chris, Bill, > > In our DwC-MIxS mapping activity, we noted a practice to support the IPT > to generate "picklists" (for drop-down menus etc) by adding extensions > to IRIs. > > This was suggested as a means to link the values to keys in > terminology-focused fields. > > You can see the link to the "picklists" in the tracker: > https://github.com/tdwg/gbwg/tree/main/dwc-mixs/dwc#vocabularies > > An example is available in [1] and looks like: > dc:URI='https://w3id.org/gensc/terms/MIXS:0000005/contigs' > > I'm not quite sure about this and would like your input. To me, it seems > like a very strange way to modify an identifier *issued by another > standards group* in order to help a tool generate dropdown menu. > > It essential creates a false impression that there's something > standardised on the other end of that IRI that may or may not be the case. > > Best, > > Pier Luigi > > [1] > https://github.com/tdwg/gbwg/blob/main/dwc-mixs/dwc/vocabulary/contam_screen... > > > -- > https://orcid.org/0000-0002-4366-3088 > > _______________________________________________ > dwc-mixs mailing list > dwc-mixs@lists.tdwg.org > http://lists.tdwg.org/mailman/listinfo/dwc-mixs > -- https://orcid.org/0000-0002-4366-3088
Thank you both for working this out and explaining the details. I do think it is less confusing to use the GBIF namespace for URNs, rather than append the GSC URIs. Citing the MIxS URI then allows people to find the original term. Tim, would it help if we assigned URIs to the values in the CVs? That is our eventual plan. Ramona ------------------------------------------------------ Ramona L. Walls, Ph.D. Data Scientist, Ontologies, Semantics, and Metadata, Critical Path Institute Assistant Research Professor, Bio5 Institute, University of Arizona Part-time artist (https://www.instagram.com/omnivoresgratitutude/) On Tue, Jun 22, 2021 at 12:21 PM Tim Robertson <trobertson@gbif.org> wrote:
Thanks, Pier
We'll clearly state expectations in the description, and this will be kept as internal as it can be.
Using the example of the vocabulary of controlled terms for lib_layout[1] currently we have: https://w3id.org/gensc/terms/MIXS:0000041/vector
This could become something like the following e.g. https://rs.gbif.org/vocab/dna/lib_layout/vector
Would this be preferable to the GSC folk to avoid confusion, please? Note that the source from where it is derived will still be cited.
Thanks, Tim
[1] https://rs.gbif.org/sandbox/vocabulary/mixs/lib_layout.xml
On 22/06/2021, 17.41, "Pier Luigi Buttigieg" <pier.buttigieg@awi.de> wrote:
Thanks Tim, this clarifies a great deal.
Measures to make it clear that the URNs (wherever they appear, e.g. on documentation pages, lists...) used by legacy systems are not to be understood as resolvable or under the control/custodianship of the GSC would help.
While technically valid, prefixing URNs with HTTP does indeed set a false expectation - anything that can mitigate that would be appreciated.
On the GSC and TDWG side however, it would be very interesting to find out how to do this more sustainably by, e.g., sharing vocabulary files with all the metadata needed to, e.g., map to ontologies, qualify terms with comments, include provenance data, etc
On 22/06/2021 17:18, Tim Robertson wrote: > Thanks, Pier > > Just for a bit of background - you're assuming these are IRIs, but the DwC-A work was all designed at a time when URNs were common and expectations of resolvability weren't as high as today. > The original design was really just to ensure a globally unique URN was used. > > I think the concerns you raise are valid, and especially with the current intent to achieve resolvability for all the w3id.org/... terms. > We do need a URN though, and I understand there isn't an intention to make resolvable terms for these concepts. The rs.gbif.org namespace has been used for this kind of thing in the past, which might be an alternative option to lower expectations. > > Thanks, > Tim > > > > On 22/06/2021, 16.02, "dwc-mixs on behalf of Pier Luigi Buttigieg" < dwc-mixs-bounces@lists.tdwg.org on behalf of pier.buttigieg@awi.de> wrote: > > Dear TG, Dear Ramona (GSC CIG), Chris, Bill, > > In our DwC-MIxS mapping activity, we noted a practice to support the IPT > to generate "picklists" (for drop-down menus etc) by adding extensions > to IRIs. > > This was suggested as a means to link the values to keys in > terminology-focused fields. > > You can see the link to the "picklists" in the tracker: > https://github.com/tdwg/gbwg/tree/main/dwc-mixs/dwc#vocabularies > > An example is available in [1] and looks like: > dc:URI='https://w3id.org/gensc/terms/MIXS:0000005/contigs' > > I'm not quite sure about this and would like your input. To me, it seems > like a very strange way to modify an identifier *issued by another > standards group* in order to help a tool generate dropdown menu. > > It essential creates a false impression that there's something > standardised on the other end of that IRI that may or may not be the case. > > Best, > > Pier Luigi > > [1] > https://github.com/tdwg/gbwg/blob/main/dwc-mixs/dwc/vocabulary/contam_screen... > > > -- > https://orcid.org/0000-0002-4366-3088 > > _______________________________________________ > dwc-mixs mailing list > dwc-mixs@lists.tdwg.org > http://lists.tdwg.org/mailman/listinfo/dwc-mixs > -- https://orcid.org/0000-0002-4366-3088
Thanks Ramona
would it help if we assigned URIs to the values in the CVs? That is our eventual plan
Yes, longer term this makes sense. I anticipate we’ll probably arrive at SKOS or similar for these kind of vocabularies.
…it is less confusing to use the GBIF namespace for URNs
Thanks. We’ll change to use GBIF namespaces unless someone brings new information and will update the TG report accordingly. Best wishes, Tim From: Ramona Walls <rlwalls2008@gmail.com> Date: Thursday, 24 June 2021 at 18.18 To: Tim Robertson <trobertson@gbif.org> Cc: Pier Luigi Buttigieg <pier.buttigieg@awi.de>, "dwc-mixs@lists.tdwg.org" <dwc-mixs@lists.tdwg.org>, Chris Mungall <cjmungall@lbl.gov>, Bill Duncan <wdduncan@gmail.com> Subject: Re: [dwc-mixs] DwC "picklist" extension to MIxS IRIs Thank you both for working this out and explaining the details. I do think it is less confusing to use the GBIF namespace for URNs, rather than append the GSC URIs. Citing the MIxS URI then allows people to find the original term. Tim, would it help if we assigned URIs to the values in the CVs? That is our eventual plan. Ramona ------------------------------------------------------ Ramona L. Walls, Ph.D. Data Scientist, Ontologies, Semantics, and Metadata, Critical Path Institute Assistant Research Professor, Bio5 Institute, University of Arizona Part-time artist (https://www.instagram.com/omnivoresgratitutude/) On Tue, Jun 22, 2021 at 12:21 PM Tim Robertson <trobertson@gbif.org<mailto:trobertson@gbif.org>> wrote: Thanks, Pier We'll clearly state expectations in the description, and this will be kept as internal as it can be. Using the example of the vocabulary of controlled terms for lib_layout[1] currently we have: https://w3id.org/gensc/terms/MIXS:0000041/vector This could become something like the following e.g. https://rs.gbif.org/vocab/dna/lib_layout/vector Would this be preferable to the GSC folk to avoid confusion, please? Note that the source from where it is derived will still be cited. Thanks, Tim [1] https://rs.gbif.org/sandbox/vocabulary/mixs/lib_layout.xml On 22/06/2021, 17.41, "Pier Luigi Buttigieg" <pier.buttigieg@awi.de<mailto:pier.buttigieg@awi.de>> wrote: Thanks Tim, this clarifies a great deal. Measures to make it clear that the URNs (wherever they appear, e.g. on documentation pages, lists...) used by legacy systems are not to be understood as resolvable or under the control/custodianship of the GSC would help. While technically valid, prefixing URNs with HTTP does indeed set a false expectation - anything that can mitigate that would be appreciated. On the GSC and TDWG side however, it would be very interesting to find out how to do this more sustainably by, e.g., sharing vocabulary files with all the metadata needed to, e.g., map to ontologies, qualify terms with comments, include provenance data, etc On 22/06/2021 17:18, Tim Robertson wrote: > Thanks, Pier > > Just for a bit of background - you're assuming these are IRIs, but the DwC-A work was all designed at a time when URNs were common and expectations of resolvability weren't as high as today. > The original design was really just to ensure a globally unique URN was used. > > I think the concerns you raise are valid, and especially with the current intent to achieve resolvability for all the w3id.org/.<http://w3id.org/>.. terms. > We do need a URN though, and I understand there isn't an intention to make resolvable terms for these concepts. The rs.gbif.org<http://rs.gbif.org> namespace has been used for this kind of thing in the past, which might be an alternative option to lower expectations. > > Thanks, > Tim > > > > On 22/06/2021, 16.02, "dwc-mixs on behalf of Pier Luigi Buttigieg" <dwc-mixs-bounces@lists.tdwg.org<mailto:dwc-mixs-bounces@lists.tdwg.org> on behalf of pier.buttigieg@awi.de<mailto:pier.buttigieg@awi.de>> wrote: > > Dear TG, Dear Ramona (GSC CIG), Chris, Bill, > > In our DwC-MIxS mapping activity, we noted a practice to support the IPT > to generate "picklists" (for drop-down menus etc) by adding extensions > to IRIs. > > This was suggested as a means to link the values to keys in > terminology-focused fields. > > You can see the link to the "picklists" in the tracker: > https://github.com/tdwg/gbwg/tree/main/dwc-mixs/dwc#vocabularies > > An example is available in [1] and looks like: > dc:URI='https://w3id.org/gensc/terms/MIXS:0000005/contigs' > > I'm not quite sure about this and would like your input. To me, it seems > like a very strange way to modify an identifier *issued by another > standards group* in order to help a tool generate dropdown menu. > > It essential creates a false impression that there's something > standardised on the other end of that IRI that may or may not be the case. > > Best, > > Pier Luigi > > [1] > https://github.com/tdwg/gbwg/blob/main/dwc-mixs/dwc/vocabulary/contam_screen... > > > -- > https://orcid.org/0000-0002-4366-3088 > > _______________________________________________ > dwc-mixs mailing list > dwc-mixs@lists.tdwg.org<mailto:dwc-mixs@lists.tdwg.org> > http://lists.tdwg.org/mailman/listinfo/dwc-mixs > -- https://orcid.org/0000-0002-4366-3088
participants (3)
-
Pier Luigi Buttigieg -
Ramona Walls -
Tim Robertson