1.40am. Why not?

So the game is, find a new appointment listing on .gov.uk

https://www.gov.uk/government/news/lord-lieutenant-for-dunbartonshire-12-august-2020

Do we have that position on wikidata?

Yes we do.

Although we say LL of D, they say LL for D. Alias required.

Alias also required 'cos spelling.

https://www.wikidata.org/wiki/Q6679560 

Do we know any of the position holders?

Add {{PositionHolderHistory|id=Q6679560}} to the talk page, hit the button.

Umm. No.

None of our items have LL of D as a P39 position held.

https://www.wikidata.org/wiki/Talk:Q6679560

Can wikipedia help us out? Is there a list?

(You'd better believe this exercise would not have taken place had there not been a list.)

Yes. There is a list. It has dates. We can throw that into a spreadsheet and separate out the distinct values.

https://en.wikipedia.org/wiki/Lord_Lieutenant_of_Dunbartonshire

But we need QIds to make progress. Can wikipedia help, part II?

Yes it can: it's got a category - Category:Lord-Lieutenants_of_Dunbartonshire in which many of those on the prior list, are also listed

https://en.wikipedia.org/wiki/Lord_Lieutenant_of_Dunbartonshire

So we can ask Petscan (pay attention at the back, @watty62) to give us a list of articles and their QIds

We can throw this list into the same spreadsheet and line up the two lists, keyed on name, using the eye, or sort, or vlookup, or whatever.

http://petscan.wmflabs.org/?psid=17072743 

Here's the wikipedia list of... in the spreadsheet.

Step 1 was to extract three columns, of Name, From and To. The data is in a regular pattern; we can search for "(" and "-" as anchors for some formulii which do what's wanted.

And then break the from & to dates down into their DD MM YYYY components - using leading zeroes when required - so we have the data needed to build a quickstatements date string - which is in the form +2020-08-13T01:15:00Z/11

Throw in the quickstatements list, which regrettably is only 8 rows long, against the 17 rows of the wikipedia table. So we'll have to hunt for QIds for 9 people, and/or coin new items for them. In doing so, keep a waether eye out for biog articles missing the LLofD category.

So quickly, by eye, we get this - and can see that the likes of Michael Gregory, Angus Cunninghame Graham, Alexander Telfer-Smollett have articles but are not in the LLofD category. So I'll off & fix all that, get their QIds, and then dig into six who have no article.

Adding QIDs for articles now categorised we're missing QIDs for 3 people, who may or may not be on wikidata ... too little info to tell.

I'll do a more searching, and if I draw a blank, create new items.

The LLofD has a website. How cool is that?

https://www.lordlieutenantdunbartonshire.co.uk/index.php 

We were missing a QId for Donald Ross. Found him in, of all things, an "Articles that were deleted in Wikipedia" on a Wikia site.

Anyway, enough info to check that he's absent from Wikidata (unless hidden amongst the 1,000s of The Peerage items)

https://military.wikia.org/wiki/Donald_Ross_(Lord_Lieutenant)

Here's the missing Robert Arbuthnott, or, at least, enough info to work out if he is or is not on wikidata.

I drew a blank on James Cassels Robertson; and, conclude that all three need new items.

http://www.greatwarbritishofficers.com/index_htm_files/Arbuthnott,%20J%20Research.pdf

Might as well use quickstatements to make three stubby items

And then the big QS - give each of them an LLofD P39 position, with start and end date, replaces, replaced by, and an imported from wikipedia source.

No series ordinal right now - not convinced wikipedia's list starts with the first LL.

Slightly cocked up the three dates which were YYYY only, and spotted one which was YYYY MM only; but Quickstatements sensible hung when it encountered the first mistake. So. Fixed, and the last four P39s go in.

All done. 90 minutes including digressions like, err, this thread & its many screengrabs.

Probably longer and more faffy than the approaches @tmtm has been documenting recently, but it works for me, bear of very tiny brain.