Looking for input on a new project

[+Biscuits Brown MVC]

OK, let me start by saying I've been a developer for nearly 2 decades. Professionally I've primarily been involved with line of business tools and services split between VB.NET and C#. Before anyone besmirches VB.NET, that's irrelevant for this discussion. I used the language and tools dictated by my various employers over the years usually having no input on the matter.

 

My side projects however have been a bit different. I published my first astronomy based app in the Windows store in the Windows 8.1 time frame. This app was C# and XAML (what else...) Since then I've published an Android app (still astronomy based but unrelated to the Windows app). The app was a tool for last year's eclipse and was originally written in Xamarin. I eventually recoded it in Android Studio just for the fun of it long after its Play Store release. That code sits quietly in a repo someplace. Most recently I was working on a larger Windows Centennial targeted app that I scuttled after God knows how long having become the victim of endless scope creep. You'd have thought I'd have figured out how to NOT do that after all these years... For the record, both original store apps are unpublished at this time.

 

This brings me to where I am now. I'm looking to recreate and slightly enhance the original Windows Store app but I'd like it to be available on my Windows based desktops, laptops and tablets as well as my Android based tables and phones so I have some cross platform requirements. While these apps are primarily geared toward personal use, I'd publish the app to both the Windows Store and the Google Play Store because why not.

 

There are many here who's input I greatly value as you represent a cross section of developers that have used a great many more tools and frameworks than I have so I'm looking for a bit of guidance, warnings or suggestions.

 

Here are some app specifics to consider

 

• Once installed, it works completely offline. There is never a need to reconnect. If fact, in most places the app is used, there is often no connections available.
• It includes a large data set of over 2 million records. These records need to be searchable and sortable on many fields (some lending themselves to indexing better than others).
• Precess records based on the current Epoch date. This means accounting for the fact that everything you see in the sky is moving and the Earth is wobbling on it's axis means an object that is listed at being at a set of coordinates in the year 2000 (the most recent full Epoch) are no long accurate as the result of 18 years of movement. While the difference is overall small, it still impacts positional calculations for a given location/date/time.
• For selected records, once precession has been calculated for each item, items will undergo real time coordinate conversions between equatorial and azimuthal will occur with perpetual coordinate updates.
• The UI is self is rather simple and does not require plotting of stars or object, its strictly a text/data based tool. Mostly search criteria and a sortable results grid.
• Since this tool is a personal project, I have no timeline to work towards. If this takes a year or more to complete, that's OK.

 

Here are my options as I see them

 

Xamarin Forms app targeting Windows and Android

I like this option as it fits my background but while offering a familiar set of tooling it feels a little heavy handed, like using a sledge hammer to hang a picture. Also, it provides the least amount of learning so I get a new app but little else.

 

PWA

Providing no native functionality is needed (and it really isn't) this seems like the better route on paper but seems to bring its own blend of negatives. First is the question of performance issues managing a multi-million row data set and the real-time coordinate conversion. Second is the baggage a PWA may include to monitor/handle being connected or not. Since this would be a completely offline app once installed, this seems to be adding superfluous code. This requires almost requires a start from scratch learning experience that brings a considerable learning curve that could impact my dedication.

 

Some kind of hybrid

This would be a hybrid of React Native for Android and Electron for the Windows app. I've never used either so I can't speak to the benefits or pitfalls of this option. It would appear to check all the boxes but will come with a very steep learning curve that could take a toll on my confidence to complete the project.

 

So there it is. What route would you take? What are the gains and pain points from your experience that would lead you down one path or the other. I honestly have no real gut preference at this point so I'm 100% open on this.

 

 

[LaP]

Quote

This would be a hybrid of React Native for Android and Electron for the Windows app.

I'm not familiar with Electron but if you gonna maintain two version of the code and the UI is simple then you might as well use Android Studio or IntelliJ IDEA for the Android version. I made my first "native" Android app last fall using IntelliJ IDEA and frankly it was one of the most pleasant experience i had with an IDE. In term of the quality of the experience as an IDE it's very close to Visual Studio and the learning curve is very small. The doc for Android is very well written and the api is very good. Gradle makes managing Java libraries and dependencies a lot easier too and while there's some learning curve in using Gradle once you are familiar with Gradle it makes building the project a lot easier.

 

I would personally lean toward PWA. I never did a full project with Xamarin but i fiddled with it and was not overly impressed to be honest. I would probably rather build native Android and IOS apps than use Xamarin specially if the UI is very simple. There's a learning curve to PWA (less if you are an experimented web dev) but i don't think PWA are going to die so might as well learn it. Also pretty sure there's React Native support for Windows 10 (edit : https://github.com/Microsoft/react-native-windows).

 

I never used React Native but i used ReactJS and it's very good.

 

Here's an interesting link about the different technologies : https://belitsoft.com/react-native-development/react-native-vs-xamarin-vs-ionic

 

 

Edited September 12, 2018 by LaP

[firey]

Everything I do now is PWA. Using a framework like ionic will do the job you want.  You can use WebSQL for storage so large records aren't an issue. You can process what you need to on the fly.  Can compile native for android/ios/windows using Cordova, and can build for web with service workers for the true PWA Feel.

 

I've done many apps using Ionic with full offline capabilities. Javascript is at a point now where there is a lot of native functionality, so there's really no need to 'fluff' or over-code aspects to handle offline. It just works.

 

https://ionicframework.com/

 

That's just what I would go with. Been using it professionally for 2 years now. It's great in that you have one code base, one project, but can have full cross platform support.

[+Biscuits Brown MVC]

Thanks both, good input - and faster than I expected. I'll certainly look at ionic and WebSQL. 

[Seahorsepip Veteran]

Personally I've recently done a project in React Native.

 

It's gives complete freedom but at a large cost, it's rather complex when things don't work as expected and you end up writing more native code then expected.

 

I'm starting on another project with React Native that also requires a web app, for that I've opted to use React Native Web so everything is still largely one codebase.

 

React Native Windows/MacOS/Desktop are other projects I'm interested in to work with in the future but for now I plan on bundling React Native Web with Chromium when a desktop app is required.

 

Flutter is another framework I'm planning to use in the future.

 

Biggest difference between React Native and Flutter is how things are rendered, React Native uses the view system of the underlying OS and Flutter renders everything itself like a game engine.

 

One of the main issues with React Native is that things render sometimes differently between platforms, this is not the case with Flutter.

 

Besides that are the obvious language differences between the different framework but those are expected.