WFS - some remarks

[filipmahieu]

On the development-site a (new) wfs was made (see issue 65): https://dev-alert.riparias.be/api/wfs/observations .
We wish to add some remarks:

• Can the dutch name of the species be added (necessary for us)
• Can the scientific name of the species be added (necessary for us)
• Is the 'date' the date of observation?
• Are the observations only the 'present' observations (not the 'absent' observations).

It is thus stil the development-site. Can there also be created a wfs on the production-site?

[damianooldoni]

I answer the question I think I can know the answer:

I will add soon the Dutch name of species. I will try to do it on Admin panel of production. In this way @niconoe can add it.

Absences are not harvested from GBIF so we don't have them. What we still have are observations which are actually management actions, so the species is present but then removed/eradicated.

[filipmahieu]

Please don't forget to add the scientific (latin) name. Thanks

[niconoe]

@filipmahieu: to answer your last question: the WFS server is also available on the production site, the URL is: https://alert.riparias.be/api/wfs/observations

Hopefully I'll find time soon to implement your other remarks.

[niconoe]

Hey @filipmahieu, I've been able to make some progress on this issue today. The results is already visible on https://gbif-alert-demo.thebinaryforest.net/ and will soon be on the RIPARIAS prod site.

Changes/answers:

• The scientific names and the vernacular names (in 3 languages) have been added
• The "species_id" field has been renamed to an embedded field (species.id)
• Yes, the date field contained the date of the observation. It has been renamed to observation_date to avoid any ambiguity.
• GBIF alert will return all observations present in the system. In the specific case of LIFE RIPARIAS, the configuration was done so only the observations considered as "present" by GBIF are integrated in the nightly download. So for your use-case at least, the answer to your question is yes.

Details about the fields after the update:

Can you tell us if this works for you? Also I'd be interested in knowing which fields you actually don't use. I initially included everything but later realised some things should probably be kept internal. But I don't dare removing data, knowing you might use it in your own scripts.

[filipmahieu]

Ok, I cannot see the changes yet in the WFS's (dev or production).

[filipmahieu]

Don't worry about 'our scripts' because the automatic download at night downloads it all from scratch.

[niconoe]

Thanks for the feedback, @filipmahieu!

• About the non-visible changes: I'll deploy it very soon (hopefully today)
• About "don't worry about our scripts": that's great, I've take the freedom to improve the response further (do not leak internal data, better field names, more details about the dataset). See image below to see the new format of the response:

[filipmahieu]

We see it now in production.
We see in QGis the fields under 'Species' as one field (content '95
Baccharis halimifolia
struikaster
Eastern baccharis
séneçon en arbre
'
Is it possible to give these fields separately so we can visualise them on e.g. id or vernacular_name_nl

[filipmahieu]

We thank you allready for the efforts made to have a WFS!

[niconoe]

Hi @filipmahieu: indeed, we use the "complex fields" possibilities of WFS to provide details about the species and the dataset, apparently this feature is not well supported in QGIS, making things impractical. I'll update the WFS server to avoid this issue.

[filipmahieu]

Hi, referring to #272 , we'd like to test if the asked not-use of 'complex fields' would solve our issue that the WFS does not work in our automatic download tool. So solving this issue could use another one too here.
@damianooldoni

[niconoe]

I just updated the format, here is it:

Not that this is not yet deployed/visible on the site, but I plan to do that later today!

[filipmahieu]

Thanks,

could it be possible to add the field id that I saw in the previous version: here a screenshot

I suppose that's an id referring to a certain list thats's used inside Riparias with a stable id per taxa. Can I somewhere see that list? Would be very interesting to make a legend upon in QGis.

[niconoe]

I just added the field and deployed the changes to both servers (test and production). I'll send you the species list by mail immediately.

Tell me if all that works for you now!

[filipmahieu]

We are able now to download each night the data and implement a legend (with speciesid). Thanks.