I’ve seen a lot of positivity surrounding login.gov on HackerNews. I’ve never used the service and am unfamiliar with the quality of its implementation. Many commenters here point to login.gov as an example of the US government shipping good software. [1]
1. From the end-user’s perspective, what makes this a quality service? Is it simply better than other government alternatives, or does it compete with equivalent modern services from the private sector?
2. From the technologist's perspective, why is this considered quality software? I see it's an open-source Ruby on Rails app[2] with basic documentation, tests, and monitoring. As a non-RoR developer, I'm curious where this project falls on the spectrum from merely adequate to exceptional, and why.
[1] e.g., in this comment section: “login.gov is one of the few government services that as a private sector techie I'm in awe of”
> 1. From the end-user’s perspective, what makes this a quality service? Is it simply better than other government alternatives, or does it compete with equivalent modern services from the private sector?
I couldn't even login to ssa.gov before it was integrated with login.gov. Every year or two I'd give it a shot, it told me my account was locked, I had to visit a Social Security office to get it unlocked. I tried that once; the local office wasn't able to help. Fast forward a few years and the login process has been delegated to login.gov. I was able to prove my identity in the normal way (asked a bunch of questions from my credit report) and finally login.
So let's start with: it works.
But it's at least as good as any SSO that I use elsewhere (Okta, Apple, Google). It supports multiple factors (security key, passkey, TOTP, etc), something that, e.g., Fidelity only barely offers.
Besides that, it's visually appealing, having a nice modern look.
I'm happy that it sounds like Login.gov is better than the broken solution we had before. At the same time, I do not think a basic functional login system is something that should be celebrated as a success.
Why would this not be a success? The previous system didn’t work, the govt created a specialized team that built a great functional product, and now it works... they fixed a problem? Many would consider that a success.
Maybe don’t look at this through the lens of a tech company or normal business (bc it’s not), but look at I from the perspective of how shite govt tech is. Not sure if you live in the states but you should try apply to unemployment in somewhere like Florida and then report back to me how having a functional login page isn’t a success.
It's funny you say that. The NC DMV used to have a decent site for renewing registration. It was basic but functional. No bling. Took me like 2 minutes to use.
A few years ago they replaced it with a vendor solution (PayIt[1]). It's terrible. The renewal process easily takes 5x as long. The old site was two steps and a couple forms. The new site is this stupid chat bot interface that pretends like it's thinking between the half dozen or so steps it now is. On top of that, I get to pay $3 or something for the privilege of using it.
Annoys me just thinking about it.
I have a whole rant about our local private toll road's web site too. Easily in the bottom 5 sites I have to deal with. I may switch back to MA's EZ Pass just out of spite.
As a Canadian end user with a nexus card, it just works, and it works with less errors and issues than my revenue canada login which has poor UI and steers me towards sign in partners with by making that the natural login with dark patterns. Come to think of it, I'm surprised the nexus program hasn't been cancelled yet.
1. From the end-user’s perspective, what makes this a quality service? Is it simply better than other government alternatives, or does it compete with equivalent modern services from the private sector?
2. From the technologist's perspective, why is this considered quality software? I see it's an open-source Ruby on Rails app[2] with basic documentation, tests, and monitoring. As a non-RoR developer, I'm curious where this project falls on the spectrum from merely adequate to exceptional, and why.
[1] e.g., in this comment section: “login.gov is one of the few government services that as a private sector techie I'm in awe of”
[2] https://github.com/18F/identity-idp/