Ask HN: Is MiniApp W3C standard required? | Hacker News Reader

Ask HN: Is MiniApp W3C standard required?

w3.org

4 pointsjsdeveloper5 comments

I came across this new web standard in progress called MiniApp. The use cases it presents , can all be covered with the existing PWA + Service worker ecosystem. One can create a PWA which saves web pages (basically other mini apps): either in cache or indexeddb.

It's orthogonal to that. It's about the app being embedded in some other app that handles concerns such as authentication.

From my point of view, PWA is mostly a nothingburger. I mean who needs another way for web sites to spam me with spamifications about the spamifications they'd like to spam me with if only I am so foolish to consent to more spamification? Similarly the service worker approach is like a Rube Goldberg machine for building Rube Goldberg machines. It's like nobody learned from Netscape Netcaster.

Then some stooges for Google go around and say that Apple is ruining the web by not supporting that nonsense when really Apple is just stopping the insanity.

Apple considers PWAs a threat to their App Store cash cow. They would support them in an instant if they thought it would be profitable.

I am unsure about how/which experience of yours made you conclude Service worker as Rube Goldberg machine, as otherwise with just few lines one can write a service worker.

Push Notification has its own use case, a user who wants instant stock updates and don't want to install a whole big native app for just this, can simply accept notification from a PWA and starts getting customized notifications. No need to install the PWA itself, juts subscription to notification is required.

Regarding push service and its usage for spams, the user has full control to stop those notification any instance. Personally PWA seems to me as the greatest thing that has come out of web, which gives full control in user's hand. User can very easily see what data an app collects and control what data user wants to share.

This is really wild. Thanks for sharing.

Interesting seeing a bunch of huge Chinese Bug Tech companies starting to "standardize".

Right now this feels more like an idea than a full built out spec. It's quite a grab bag of concerns. It feels like there's a ton of relevant specs they could have drawn from like Web Bundles, rather than loosely declaring vague intent to build yet another packaging system as a part of this mega-multi-spec. At least we some logical extensions: they use the existing Web App Manifest work as a basis, and throw a ton in top, which could be worse.

Overall I'm still really confused what this really represents. There's such a weird new UI paradigm, married to new bundling, with new linking, and a new life cycle... And more. It doesn't feel web. It looks like a native app toolkit that loosely references the web.

I don't think it will have any of the charitable rfc8890 "the Internet is for end user" benefits of the current web. It feels like it takes & I can't see advantages over what we have. Distribute some Custom Elements on your proprietary platform, if that's what you want, which seems to be the case. Don't build an entirely parallel non-web system, gross.

Split it up! I'd love to have a new widget spec! Widgets rocked! But this macro framework feels bad. It feels so native-platform like, such a large unit, which isn't how the web won.

Reply on news.ycombinator.com