[notsapiensatall]

Very cool!

I remember about a decade ago, I wanted to analyze US Congressional votes to look for evidence of a possible third-party voting base in across-the-aisle votes.

The votes were generally published as XML documents, which was sort of annoying but a fun way to learn Python.

Nice to see things getting more standardized - I've been pleasantly surprised by how the US government's digital services have advanced in the past 10-20 years.

(Except for the contracted-out disasters like recreation.gov and V1 of healthcare.gov, of course.)

[aaronbrethorst]

I was about to say "recreation.gov, I haven't seen that one before" — and then I noticed it was in my browser history and thought 'oh shit, this disaster of a website...'

[kevml]

Disaster? I am personally very impressed that site works so well! It’s a unique and complicated system, and I’m impressed the government built that. I’m certainly not a daily user, but every experience I had has been good.

[jonas-w]

As a non-US citizen, what's wrong with recreation.gov?

[notsapiensatall]

It's a website set up to manage camping and hiking reservations for national parks.

It just didn't turn out very well. Difficult to navigate, bloated JavaScript interfaces, poor discoverability.

It works, and worse websites exist. But it feels like a buggy early-2010s web app, and serious outdoorsy types have bots snipe the best reservations as soon as the system opens for the season. It's hard on people who don't use a computer every day and just want to plan their family's vacations.

[aaronbrethorst]

The one and only time I had to interact with it was while I was in a US national park trying to pay a parking fee for my car. I had one bar of service on my phone, and it just sat there trying to download JavaScript.

Lots to be said for considering where and on what devices your users will be interacting with your website.