nmtoken
commented
3 months ago INSPIRE download sevrices don't have to have static content (predefined in download services TG), they can be dynamic.
Open coalsh opened 3 months ago
nmtoken
commented
3 months ago INSPIRE download sevrices don't have to have static content (predefined in download services TG), they can be dynamic.
coalsh
commented
3 months ago Hi @nmtoken, thank you for responding so quickly. There isn't any problem with the ‘Download Service (download)’ Spatial data service type per say. I believe INSPIRE's description is adequate for near real-time data streaming services:
Download Service (download): Service that enables copies of spatial data sets, or parts of such sets, to be downloaded and, where practicable, accessed directly.
The issue lies with the requirement for Coupled resource to be mandatory for all Download services, necessitating the drafting of a dataset record to associate with the Service record. While this aligns well with static data, where downloading involves associating with a dataset record, it may not always be applicable to near real-time data streams. The dynamic nature of streaming data makes it more difficult to describe in detail within a dataset record. As a consequence, both the dataset and service records end up containing similar levels of detail, essentially becoming duplicates of each other. This makes including a Coupled resource somewhat redundant and inefficient, resulting in a waste of time and resources.
coalsh
commented
3 months ago I should also note that Cefas currently describe the service using a dataset record https://data.cefas.co.uk/view/21682. What would make more sense is to simply use a service record instead having to create a dataset to associate with it.
For data access see the Data tab which for view and download, see the map server links on the map tab and also see if the API service https://data-api.cefas.co.uk/api/recordsets/12651/data?page=1&resultsPerPage=100000
More description of API services as a whole are here https://www.cefas.co.uk/data-and-publications/cefas-data-hub-apis/ and here https://data-api.cefas.co.uk/index.html
One of our MEDIN stakeholders have a data streaming service which provides a near real-time data feed for WaveNet data which is limited to only the past 48 hours. The service allows users to view the data, download the data, connect to it via an API and/or via a WMS. For this reason, ‘Download Service (download)’ appears to be the appropriate Spatial data service type (#37) classification for this service.
The Coupled resource (#38) element is where things become somewhat problematic. Given that Coupled resource is mandatory for View and Download services, to maintain compliance with INSPIRE/GEMINI/MEDIN, it necessitates drafting a dataset record to associate with the Service record. This of course makes sense for conventional download services that provide static precompiled data. However, this is redundant for near real-time streaming data services, as both the service metadata record and the dataset record would essentially be duplicates of each other (given the dynamic nature of streaming data compared to precompiled data) which is an inefficient use of time and more of a workaround.
As real-time data streaming services are becoming increasingly common, it's likely that this issue will be raised more frequently in the future. It would be ideal if INSPIRE added a new ‘Data Streaming Service’ Spatial data service type to their list (and make Coupled resource optional for it) to accommodate streaming services but that is obviously outside the scope of this group.
What would the group thoughts on this scenario and what would the group advise for near real-time data streaming services? Perhaps ‘Other Service (other)’ would be a better Spatial data service type to use given the dynamic nature of streaming data compared to precompiled data? Or does ‘Download Service (download)’ need to be used in this scenario along with an associated coupled resource?