Hacker News new | past | comments | ask | show | jobs | submit login

They directly address this point. In the article they say they considered adding sign in with Apple everywhere. Besides the fact that it's more code to write and test, they say the documentation is very poor for other platforms, as it's not even great for iOS.






Absolutely not. Compare the native ios documentation[1] with their "other platforms" documentation[2]. Their native documentation has code snippets, helpful links, and explains in depth what is happening.

The "other platforms" documentation is "make this request, store some data, follow redirects". No code, no helpful links on how you might accomplish these things, nothing. You get the bare minimum.

I'm not saying its impossible, and neither is the article. I'm saying that Apple clearly doesn't care about supporting a platform as huge as Android, and that clearly signals to multi-platform developers that they are on their own.

[1] https://developer.apple.com/documentation/authenticationserv...

[2] https://developer.apple.com/documentation/sign_in_with_apple...


The nature of "other" platforms means that example code could be in any language at all.

The fact that their iOS documentation is so much better than average doesn't mean their "other platforms" documentation is inadequate. It just means there's plenty of room for third parties like indie bloggers to documentation their own approaches in JavaScript, Python, Ruby, Rust, or whatever.


I think you're missing the point.

Part of the problem is that android is an "other platform" in the first place. Sign in with apple is supposed to be a cross platform feature, but Apple can't be bothered to even write out some decent documentation for a platform with over 2 billion devices. Compare, for example, the Google sign in for iOS[1]. They provide a working example project and full documentation.

If you are a developer supporting a cross platform app, you're not getting much help from Apple. That's what the article is saying: integrating this feature is going to be more work and more risk than it's worth. That's the point.

> It just means there's plenty of room for third parties like indie bloggers to documentation their own approaches in JavaScript, Python, Ruby, Rust, or whatever.

We are talking about signing in. This is one of the most fundamental features you need to have. This is not something that you just copy paste from some half baked blog post. It is amazing to me you think that's acceptable.

[1] https://developers.google.com/identity/sign-in/ios/start


> This is not something that you just copy paste from some half baked blog post. It is amazing to me you think that's acceptable.

Have you ever actually implemented a proper sign-in process e.g. with OIDC, JWT, SSO etc ?

Because half baked blog posts is the industry standard.


Hardly a surprise that they're so crap, then.

Not surprised the documentation that highlights the Apple Platform is better, however, give.

That said, it’s a REST API you query and you get a well defined payload:

> A successful response contains the following parameters: code A single-use authorization code that is valid for five minutes. id_token A JSON web token containing the user’s identity information. state The state contained in the Authorize URL. user A JSON string containing the data requested in the scope property. The returned data is in the following format: { "name": { "firstName": string, "lastName": string }, "email": string }

I could implement this using curl really it’s that straightforward. If you have any experience consuming REST APIs


To repeat myself, I know that this is more than possible to implement. But you're also hiding a lot of complexity about redirecting from your app to a browser, managing state, custom url schemes, etc. If you want to turn your curl request into an actual app, there's some nontrivial code you have to write and test yourself. And this code is important - if a user can't sign in, your entire app is broken.

And to what benefit? This is the point of the article. Sign in for apple is extra work and extra complexity for no benefit (to the developer, at least). It's an immature project and the fact that Apple is putting in the bare minimum effort into the docs does not encourage me to adopt this feature.

Google, in comparison, has a working sample project and step by step guide for implementing Google sign in on iOS[1]. Google sign in is just as much a "curl request" as apple sign in, but they put in the effort to give a high quality, well integrated, and native example.

Apple can't be bothered, which discourages people like OP from adopting the feature.

[1] https://developers.google.com/identity/sign-in/ios/start


And even with that - they are not going to use Google.

Even though there is way more value in supporting Google, than Apple. Google SSO is widely used in small businesses, unlike Apple ID.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: