Hacker News Reader

Comment by PedroBatista | Hacker News Reader

Obviously I feel the pain of the OP but I think he blames the wrong thing.

Flutter is a ( viable ) technology that’s being growing for the last few years. A very significant percentage of new mobile apps are programmed with it these days.

It seems his problems have everything to do with the local job market and not the technology. With the added bonus that you can be a single man operation and release your own apps for both iOS, Android and the desktops.

I’m not a fanboy and even less of Google but Flutter really offers the ability to make lemonade if the job market gives you lemons.

View on HN

Flutter is pretty old if you think about it and still has too many rough edges and not a lot of big name adoption. The project is too ambitious, and the features seem more like PoCs than something you can rely on in production, IMO.

This is what has prevented me from diving into Flutter for a personal project. Although it supports desktop and web, it is first and foremost a tool for building mobile applications. For desktop, too much of the API is built around the assumption that you are on a mobile device with a single window and a finger as the pointing device. Feature requests for desktop, such as proper support for mouse-based drag and drop and multiple windows, are either ignored or take forever to develop.

Any platform that tries to support web, mobile, and desktop all at once, of which there are many, are destined to this fate of being spread too broadly and thinly.

What do you mean by mouse based drag and drop? As far as I can tell Draggable and DragTarget work fine on windows and macos at least.

The API assumes you are on a relatively small screen and using a finger to drag and drop. Thus, it doesn't provide functionality that you would expect if you were on a desktop or laptop monitor with more space and using a mouse and keyboard where there is a wider variety of things that you would like to happen with a drag and drop interface. The API is primarily assuming you're taking a widget and dragging it with a finger to somewhere just an inch or two away on a phone or tablet screen. Yes, drag and drop "works" on desktop as it is, but the API is not well-designed for desktop-based usage. A lot of ad-hoc type of behavior using global state has to be used.

As long as you don’t include in that number - most apps built by Google.

Is that actually true though? Most major Google apps predate Flutter. I was under the impression that a lot of their apps are either custom web view wrappers or native with shared C++ code for business logic.

Stadia and Google Pay are the only consumer facing Google apps I’m aware of that are made with Flutter.

Family Link [1] also looks like it's made with Flutter (I checked the DevTools in the browser).

[1]: https://familylink.google.com

Most of their apps are native. They learned the same thing that Facebook learned years ago. Cross platform GUI apps have always sucked - or at least since Java tried to popularize them in the late 90s.

Ads is all flutter on mobile and Dart on the web.

Fuchsia is an OS that is also using Flutter for their entire UI layer.

I think there’s more a case of there isn’t an argument for doing but rewrites just because Flutter is now a thing but in a greenfield scenario it would be a serious contender

So you can invest your time I learning Fuschsia and Flutter. What could possibly go wrong?

But that’s not demonstrating the promise of a great cross platform experience.

You made an argument saying that even Google won’t use it for cross platform.

I have some context as to why it’s both a wrong and misleading statement.

When it came to the biggest money maker of the entire business this is what they picked. When it came to a new greenfield project for an entire OS… this is what they picked for the UI.

The statements you made about “the lessons they learned” is just some stuff you made up and isn’t actually supported by facts.

Ads is their biggest money maker. But it doesn’t have to be usable compared to their consumer facing software. Even though it’s used by ad buyers, it’s more akin to internal tools that at most large companies don’t get near the care of consumer facing tools (looking at you Amazon where the internal review process website broke last April).

I used to feel the same way but Flutter is hardly a new technology anymore yet still seems to be used very little. I've dabbled with it on hobby projects here and there and it seems "fine" but not sure I'd bet the farm on it.

It’s been a thing outside the context of iOS and android for maybe less than 2 years? It’s hardly in the “it had its chance but failed” category.

Dart only survived thanks to Ads business unit, naturally they have a Flutter app.

Google can hire separate developers for each platform and pay them, investing millions of dollars into their mobile app.

Can you afford to spend money developing same thing twice forever?

I worked for a midsized health care company a few years ago (around 300 people including customer service). They had three mobile developers who supported the iOS app and the Android app.

Similarly I worked at a 1000 person financial institution for years and represented 50% of the Android mobile developer division there.

Similarly there were 3 iOS devs.

The feature set was enormous but we did fine.

Reply on news.ycombinator.com