Kudos to @andertonsmusic support for a swift and informative reply to my order status request. Good stuff.

RT @policyatkings: EVENT: Failed State: Why Nothing Works and How We Fix It. Join us and @fairnessfdn as @Samfr discusses his bestselling….

In a meeting where various folks didn't know each other. My lead asked for suggestions for an ice-breaker, and I suggested "favourite time zone" - we went with it, and it was awesome! #LifeGoals.

And it's over - for now. An absolute joy to see the "regular" cast celebrating the swings; really in keeping with the themes of the show. Now I need to work out when I can see them sing it again (and again).

Interval. Everyone is *smashing* it. @melanielabarrie seems extra playful today - maybe it's just me, but it's fab.

Last Sunday at @HadestownUK I was fortunate enough to see Tiago Dhondt Bamberger as Orpheus; today I get to see Miriam Nyarko as Eurydice, and @L_R_Buckley as one of the Fates. All the regular leads are fab, but I do enjoy seeing the subs/swings shining.

(There's a lot more to do here - in particular when the results are incomplete; the last radio button option is terrible at the moment. But as ever, I've got nearly 5 years to fix that. I want to use query parameters in a similar way on the maps page too. ).

It's the political bombshell *they* don't want you finding out about. have you ever wondered who the winners and losers would be if in 2029 the UK voted based on the first letter of the name of their constituency? Wonder no more!

Less cynically: I'm sure CS has test environments, etc. My guess is that the cause was an unfortunate process bypass rather than a lack of technical testing facility. We'll see over time. (At least, I hope we do. I really hope a comprehensive report is published eventually.).

My (highly amateurish) election 2029 site now has a staging environment (deployed on push to main) and a prod environment (deployed on tag creation). Does that put me one step ahead of CrowdStrike?.

Ooh. contrary to earlier indications, ServiceBasedPageModelActivatorProvider *does* work, so long as my view models extend PageModel. Not sure how I feel about that though.

It's easy to go down rabbit holes in ASP .NET Core. My desire to use Razor Pages but have DI-provided models has led me to IPageModelActivatorProvider, and it feels like I may have taken a wrong turn. I'm really hoping that ServiceBasedPageModelActivatorProvider just works tho.

Normally I'm a huge procrastinator. Today, I have bought a Christmas present for Holly *and* pushed the first "real" features to (albeit only the 2024 results per constituency). I suspect this is the first site dedicated to the UK 2029 election. Yay me.

Blogged: "Lessons from election night" - - with thanks to @Samfr for the seat previews that got me started on this in the first place.

What a night - obviously for the country, but I had a fabulous time doing election tech. Many, many lessons learned - will try to find time to blog over the weekend. Apologies for the 11 minutes of downtime 5:42-5:53. Next order of business is scraping majorities from the BBC.

So far, in terms of seat accuracy, the Survation MRP from June 2nd and the exit poll from tonight are "joint worst" on only 90% accuracy.

Now done - and realised I hadn't actually pushed the previous version. Doh.

Have worked out that I *do* have the data for "too close to call" but will need a bit of time to munge it first. .

First result flowed into my site seamlessly - hooray! I've added the exit poll as another prediction, but I don't have the data for "too close to call" (so don't expect the numbers to look the same, apart from anything else).

Anyone else tapping their fingers waiting for the next batch of seat previews from @samfr to land? Young habits die hard as well, apparently. .