Support reconciliation (e.g. via nomenklatura)

[pudo]

Suggest we set up some kind of autopush to nomenklatura ...

[rufuspollock]

What about just loading the data to nomenklatura and redirecting there?

[pudo]

I like where this is going. You realize that nomenklatura supports a random data={} dict on all entries which could be used to store template input for a static site generator?

[rufuspollock]

I don't think you want the site itself in nomenklatura - that can stay in a dedicated site - but all the reconciliation can be nomenklatura. I think this is a nice model of one dedicated tool for one dedicated function :-)

So e.g. publicbodies site front page would say for reconciliation against this dataset visit nomenklatura/publicbodies (or whatever the url is!)

[pudo]

Yeah, thats what I understood. I really like this model. However, I'm not sure about the lack of FTS for a site like publicbodies. What do you think: can the JS solr clones handle this?

[rufuspollock]

@pudo on FTS I've booted a new ticket #3

[rufuspollock]

This should definitely go to nomenklatura (IMO).

@pudo if you could give me simple instructions (or point me to them) I'd like to have a go at doing this so i can learn more about nomenklatura - so don't do it for me ;-)

[davidread]

Ha I've put the UK body names in Nomen: http:https://nomenklatura.pudo.org/uk-public-bodies and will be adding the DGU ones too, which are reasonably reconciled already.

[rufuspollock]

@davidread perfect - and it makes me think that it may be worth splitting by country.

One super minor think for consistency: let's use iso codes (which, confusingly, for the uk is "gb"). That would suggest this be gb-public-bodies in nomenklatura :-)

[davidread]

I've now added DGU to Nomen and will move on to merging the fields into gb.csv using a Refine recipe that can be done again in the future.

I hear you about calling it gb-public-bodies - @pudo is it easy for you rename the project on nomenklatura with a cheeky SQL UPDATE?

[rufuspollock]

@davidread @pudo @markbrough @... anyone else. Have some questions here:

• Should we have one big consolidated entry on nomenklatura or should we partition by jurisdiction?
• What's the best way to automate the process of getting this stuff into nomenklatura and then regularly syncing back (my sense is that nomenklatura is probably best for being the authoritative location for creating new entries (though not for adding info to entries ... - cf Support for sending corrections / additions #8))

[davidread]

I like the partitioning by jurisdiction. Bodies do this very well naturally. And no doubt the schemas and tags are going to have a lot of local conventions. Why, do you have any use-cases for looking at bodies cross-jurisdiction?

I agree on using Nomenklatura for just reconcilliation. I put the UK ones into Nomen with a script which could be run regularly - should I check it into publicbodies/scripts or is this repo just for the web app?

[rufuspollock]

@davidread I'm very happy with partitioning - and it means we can have per-jurisdiction maintainers.

Re Nomenklatura definitely put this into publicbodies/scripts (need to recreate the directory). Also if you have any instructions or a short tutorial on how to do this it would be great to have so others can follow it too - just add it to the README or other suitable place!

[rufuspollock]

@davidread any suggestions for best way to push to nomenklatura (including your current practice) very welcome ...

[todrobbins]

@rufuspollock @pudo @davidread any interest in reviving work on Nomenklatura integrations? What is the current status of Nomenklatura project?

Cheers!