Why isn't LiveView popular for mobile app startups? Or is it?

Chat / Discussions
, ,
1

I am relatively new to Elixir (creating my first hobby app and learning on the go) and I wonder about where Elixir and particularly LiveView fit well.

Costs of supporting several platforms
A large share of startups that I see are fully or mostly mobile app based (think early Uber, Yelp or even AirBnB if it was started today). It involves having somebody to work on the backend, iOS, Android, web frontend (if there’s website) + somebody to design protocols/API, control their compatibility and at least half of a project manager to coordinate. Some people merge iOS/Android efforts via ReactNative or web tech running on the phone, but the amount of native nuances (e.g. trying to debug Android-specific layout troubles) is usually high and it’s rare that you save much while extra complexity is added.

Some teams just wrap web sites into app with tech like Cordova or Capacitor and then you just develop a single web site, but in order to make it look fast you still need to have a Single Page App that forces you to create and maintain a proper API.

LiveView opportunity
Now meet LiveView: an automagically fast web app that could be wrapped into a mobile app, no API to maintain, instant upgradeability (as it’s run on server side) and just one developer to create it all in a single language. Okay, plus somebody to help with the design, but such a person is needed for any option anyway.

It might be not exactly an ideal long-term solution as latency will exist, web pages will never feel 100% native, offline work is not really possible, using native functionalities (camera, gps, file upload, contacts, etc) might be tricky, but for many-many apps that don’t work offline anyway and are still searching most for product-market fit it seems like a killer option: just one developer to do it all, no caring of APIs + you get BEAMs reliability and scaling for free with one cheap(ish) machine possibly supporting many thousands of simultaneous users.

Is it actually happening?
Have you heard of Elixir/LiveView working well for mobile startups? Or have you considered it and rejected the idea?

Am I missing something?

  • Maybe creating a decent UI with native-like navigation is tough specifically in LiveView? E.g. maybe the way LiveView patches web page DOM is poorly compatible with mobile web frameworks such as Ionic and it’s super hard to implement typical mobile app view stacks (sliding in from the right)?
  • Maybe integrating with mobile capabilities via Cordova/Capacitor wrappers is hard and apps can’t really use camera, gps react to push messages easily?
  • Or is it about founders not willing to do even significantly cheaper-faster prototypes in the technologies they are unlikely to continue using if prototype succeeds?
  • Or maybe it’s all about the talent pool and elixir people just happen to be more backend-minded, telecom-banking oriented and don’t meet startup people much?
2 Likes
2

I think there at least 2 things here:

  1. LiveView is still very new, and has had a pretty high rate of churn over the last year in terms of major changes to how it works and how you use it.
  2. LiveView is fundamentally unable to do an “offline” mode for apps. If you’re a mobile startup it’s pretty likely that this is a deal breaker. If you’re a web startup and you want to have a good mobile experience, probably you can just have a good mobile website and don’t need to do an app.
1 Like
3

I will not go that far. Its important for some, but not for the majority. How many apps do you have installed on your phone that support offline mode?

Yes, the new LiveView Native :slight_smile:

https://liveviewnative.github.io/liveview-client-swiftui/tutorials/yourfirstapp/

1 Like
4

I honestly don’t understand the “offline mode” argument.

Internet connectivity is getting better as time goes on, not worse.

Admittedly this is anecdotal evidence, but my most used applications are:

  • Uber Eats (I am fatty)
  • Apollo (Reddit client, I am nerd)
  • Slack
  • Discord
  • Airbnb

I have zero expectation for any of those to work well with spotty internet connectivity.

5

What about PWA? Like NextJS ?

6

PWA works with any programming language and framework, therefore you can do it with Phoenix, and to start you just need to add to the root of the site the manifest file:

Example for https://apibaas.io:

{
    "name": "API BaaS",
    "short_name": "API BaaS",
    "icons": [
        {
            "src": "/favicon/android-chrome-192x192.png",
            "sizes": "192x192",
            "type": "image/png"
        },
        {
            "src": "/favicon/android-chrome-384x384.png",
            "sizes": "384x384",
            "type": "image/png"
        }
    ],
    "theme_color": "#ffffff",
    "background_color": "#ffffff",
    "display": "standalone"
}

If you go to the website on your phone you can then add the website to your home screen and when you tap to open it it will look like a mobile app on your device, but will not have all the native feeling.

Learn more about the manifest file:

Learn more about PWAs:

3 Likes
7

Well, LiveView Native is super-fresh to rely on yet (and probably will stumble into all the challenges ReactNative got into where you try to really work with native controls-code). Traditional LiveView is just web, so almost guaranteed to work within known limitations, and it is already here.

I didn’t think that maybe LiveView is just too young, so can’t be popular in all the niches just yet.

8

You need to watch the talk to understand their choices. They will work with native components for each target, not a virtual DOM that they then try to translate to each native platform.

1 Like
9

No offline support.