Spatial reference system: encoding example with authority - is this still valid? #28
Comments
|
@Sgaff could you label this? That should help us as we begin to resolve issues |
Sgaff
added
the
Guidance
|
I suggest we remove the example that is not valid w.r.t. INSPIRE |
|
2021-07-01: Increasingly, we are considering GEMINI being used for things other than INSPIRE, but we do say that a GEMINI record is INSPIRE compliant. Delete the example (and therefore renumber some others). |
|
Email discussion with Sean after the meeting (copied to all): I think it concluded that the example does actually meet the INSPIRE requirements: it provides the code, using an Anchor (TG Requirement 2.1), and does not give a codeSpace (TG Requirement 2.2 second paragraph 'shall not'). Providing the 'authority' is extra to this, allowable under https://www.agi.org.uk/gemini/40-gemini/1051-uk-gemini-v2-2-specification-for-discovery-metadata-for-geospatial-resources/#extended It would be good to change the description of the example to state this: i.e. change
to Example Three - encoding example with ISO 19115 authority, as generally described Extended metadata/additional elements |
|
2021-07-16: we realised that the element Guidance doesn't actually say what to do if the CRS is one of the INSPIRE faults. Add a bullet item between Add an element Comment mentioning the optional additional "ISO 19115 authority, as generally described Extended metadata/additional elements" Move element Comment 1 ("Only coordinate reference systems identifiers specified in a well-known common register shall be used") up to the Domain, as it is a rule. |
|
Doing this has highlighted to me that the actual wording of INSPIRE TG requirement metadata/2.0/req/isdss/crs-id differs from its example & the approach everyone uses. The requirement actually says: If the coordinate reference system is listed in the table Default Coordinate Reference System Identifiers in Annex D.4, the value of the HTTP URI Identifier column shall be used as the value of gmd:referenceSystemInfo/gmd:MD_ReferenceSystem/ gmd:referenceSystemIdentifier/gmd:RS_Identifier/gmd:code element. But the example takes the sensible approaching of putting the HTTP URI in the gmx:Anchor link attribute, and the "well-known" EPSG code as the value of the code element. Taking the general approach that a URI is more authoritative than a character string, I'll stick with that. Perhaps I'm just interpreting 'value of ... element' in a stricter XML sense than the INSPIRE TG! I have raised this as INSPIRE-MIF/technical-guidelines#7 |
|
Fixed on my laptop/Onedrive |
|
related agiorguk/gemini-schematron#7 ? |
Same element, but I don't think it's a similar issue. That one was about accepting URNs instead of HTTP URIs. This one was originally about providing the additional "authority" information - which would be possible with either encoding. The related INSPIRE-MIF/technical-guidelines#7 is about the "clash" between the wording of their requirement when referencing one of the CRSs in their default list (using its HTTP URI), and the example. |
|
INSPIRE-MIF/technical-guidelines#7 fixed in August 2022. As at October 2023, The examples had already been relabelled in the live GEMINI text. Other changes in #124 |
* Update dev-build.yml include patch-styling-code branch in dev action * Update build-adocs.yml * Update dev-build.yml To include dev-agi branch as push trigger * Massive PR to get main updated with changes to dev branch (#105) * Update build-adocs.yml testing adding a personal deploy key (JO_PAT) * Update build-adocs.yml Reverting previous change * WIP: creating dev deployment workflow * WIP deploy to dev repository * Update README.md A test commit to the dev branch * Update dev-build.yml WIP- changing to new deploy key for dev * Update README.md Force a change to dev branch * Update dev-build.yml switch to PAT * Update datasets.adoc Forcing a change to datasets.adoc to test dev workflow * Clearly state "DEV BRANCH" in Readme * Try floating the menu To bring the content nearer the top * Update README.md Remove the misleading stuff * Update README.md Use relative links so that they go to files in the same repository * Link to the GitPages Allows developer to see what they have achieved! * Update abstract.asciidoc https://docs.asciidoctor.org/asciidoc/latest/blocks/collapsible/ suggests that we can provide our own 'clickable text' instead of the default "Details". I have only done that to this element as an experiment to check it works. * Update README.md Link to where you should see the output of changes * Reducing size of Index table Remove "float" trial; make as two cells, like in the current GEMINI * Update README.md Changed title to say this is the dev branch, while working through step by step guide to editing in place * Update README.md * Update README.md Changed the two absolute URLs to the correct domain * Update dev-build.yml * Changed gemini css to handle asciidoc collapsible blocks more like links, and changed the metadatadate partial to show how it works * Update 1055-uk-gemini-major-changes-since-1-0.asciidoc Added a colon to make the list styling more consistent throughout * Updated static URLs in docs readme * Updated all the partials asciidoc files with the new style for the collapsible elements * Update README.md Made a minor change to the readme to check that the updated personal access token works correctly * Update 1048-uk-gemini-encoding-guidance.asciidoc Some minor corrections: section numbering leads to duplication of table of contents; I removed the "manual" one - on preview, this loses both, so I'll have to wait to see what happens on publication (to dev) According to #2, the sample links should already have been to GitHub, although as of 2023-03-08, that is only true of one of each link on the live site. To save migrating the invalid XML samples from the live site (!), I've changed the second of each pair of links to point to GitHub (and removed the 'series' example link). Need to migrate 1046-xml-order, or remove a lot of links to it (as it is no longer possible to automatically generate it as we could with the old "massive XML -> HTML" process.) Fixed "end note" links to go directly to the relevant page (i.e. not [4] -> page 1047 "bibliography" under a different name -> an old edition of GEMINI!) - in fact, I removed that link as basically circular! The source (live) page included a duplicate of "Figure 2" and the following explanatory paragraph; the duplicate figure had been migrated (but not the following paragraph). I've removed it. Similarly for the explanation of Figure 22a. Corrected 2.4.2 reference to "Figure 23" to the correct Figure 25. How to avoid URI namespace identifiers being converted to links in ASCII Doc? Fortunately, some of them have landing pages... Manually numbered the "figures". They (the code fragments) look good in preview, less good in HTML) * internal links should be to .html, not .htm! * Update 1049-metadata-guidelines-for-geospatial-data-resources-part-2.asciidoc Fix links mainly (including bizarre one which has for a long time pointed out to a quite different page on AGI's site & one which was pointing to 1244 but should have been to 1051!) Also fixed headings and removed quite a few instances of {nbsp} * Check & update 1051 Fixed links e.g. to be relative, to avoid going via the 'bibliography' Removed link to 1040 (see #64 (comment)) Fix the two links which said they were to 1250-summary-table but weren't (even in the live page!) Removed {nbsp}s, some of which were preventing ASCIIdoc "constrained bold" formatting See #96 - there's some duplication between this page & 1051. * Update 1051-uk-gemini-v2-2-specification-for-discovery-metadata-for-geospatial-resources.asciidoc Remove odd link that claimed to be to some summary of INSPIRE IRs, but never has been. And one formatting issue. I don't know why the "sectnums" bit doesn't work; it looks the same pas page 1049, where it does! * tidying * Update 1052-metadata-guidelines-for-geospatial-data-resources-part-1.asciidoc Change internal links to link:, and manually full-URL index to sectnums Fix data.gov.uk CSW link - now at ckan.publishing.service.gov.uk Tidy heading styles Remove mention of bibliography on the assumption we won't migrate it (see #64 (comment)) Fix link to "Part 3", which had gone to the bibliography - even in the live pages! * Update 1052 - fixed one heading w/ migrated format * WIP PR to address discussion issue #91 (comment) around how to better style code fragments. This branch uses pygments for code highlighting, with some styling overrides in assets/gemini.css. It's not perfect but at least demonstrates how to get the styling to work. So far, abstract, additionalinformation, boundingbox, characterencoding and metadatadate have been done * added pygments-default.css to the WIP patch-styling-branch * added highlighttest to see if pygments is actually working in the github version of asciidoctor * Update dev-build.yml * Update dev-build.yml * WIP test relative URLs in readme page * WIP- made changes to encoding guidelines for conformity, coupledresource, dataformat, dataquality, datasetlanguage, equivalentscale and extent * 1056 Glossary Page checked- plus tweak to stylesheet to add some padding to the left and right of the content and header divs * finished the encoding changes for the remaining partials. Renamed datasets and services to the correct name. Added ulist.html.slim to improve line-spacing for ordered lists. Edited gemini.css to change list-style inside disclosure blocks to remove linebreak between number and list item * corrected urls in readme and added additional parameters for using slim template in command examples * Update dev-build.yml * Update dev-build.yml * Update dev-build.yml * corrected URL for datasets page * Update dev-build.yml * Made some corrections to services doc and partials * Checked and tweaked 1037 * added a page for 1046-xml-element-order, made a few editorial changes to make it look neater * added 1250-element-summary * added 1053 common metadata errors * added first pass at 1054 operational guide * added additional pages to README * created a separate branch to hold version of docs that have agi URLs rather than archaeogeek * added corrected dev-build.yml * Added pages that were previously missing * Finished formatting guidelines part 3 and corrected typo in conformity --------- Co-authored-by: Peter Parslow <[email protected]> * Fixed readme and workflow actions for main branch * Fixed title of datasets and series page * Fixed title of docs readme page * Fixed links in docs readme * fixed workflow action in main branch * Styling fixes for 1051 page * Styling fixes for 1051 page * Fixes to references in 1054 * additional fonts needed (#108) * Css changes for additional fonts (#109) * additional fonts needed * additional fonts needed * Css changes (#110) * additional fonts needed * additional fonts needed * additional fonts needed * Update README.md * Update README.md * Update 1062-gemini-datasets-and-data-series.adoc * Update spatialreferencesystem.asciidoc See #28 --------- Co-authored-by: Jo Cook <[email protected]> Co-authored-by: Jo Cook <[email protected]>
|
@archaeogeek to change description for example 3 to Example Three - encoding example with ISO 19115 authority, as generally described Extended metadata/additional elements |
The guidance text for https://www.agi.org.uk/gemini/40-gemini/1062-gemini-datasets-and-data-series/#17 Spatial Reference System states the class comprises two elements, a code and a codespace. This is in keeping with the guidance from INSPIRE (p70 of inspire-tg-metadata-iso19139-2.0.1%20(1).pdf)
However, in the encoding guidance we provide for this element, we display an encoding we state is valid which is spatial reference with authority, This uses CI_Citation for the authority part and, while valid in ISO, does not meet the INSPIRE requirements.
Should we keep this example, or, as we have decided to merge GEMINI and UK INSPIRE, should we remove it?
Sean Gaffney
MEDIN
The text was updated successfully, but these errors were encountered: