Bad word: phonegap

[udg]

Hi all,

most of you know how much I like B4x tools (I often refer to them as BFD Blazingly Fast Development tools), but recently someone insisted on how wonderful is to program against a single framework (cordova/phonegap) instead of separate native tools.
Although I partially know what phonegap is (I mean its roots in HTML/CSS/JS), I find some difficulty in argue in favor of B4x, so please can you help me collect some useful points about why a software developer should go native versus framework?

BTW, I think that you can easily find programmers for both camps, so from a sw-house perspective the programming language behind the tool is non that significant. Maybe I am wrong.

Thanks in advance for your sharing of comments and experiences.

 

[Peter Simpson]

The thing about PhoneGap if my memory serves me correctly is you need to know CSS, JavaScript and HTML 5 to be able to create anything at all, that put you in the same category as a website developer IMO. It's a great idea but I'm okay in HTML 5 and CSS, but my JavaScript needs a hell of a lot of improvements.

At least with B4x tool you only need to know BASIC.

Both have pros and cons...

 

[Beja]

Hi,
In Phongap you need to submit your source code to someone (the web service) to compile it for you. a cloud technology that I don't want to trust. No Liko

 

[Erel]

PhoneGap doesn't create native applications. It uses an embedded WebView for the UI.

 

[imbault]

And about Xamarin?
C# code, a runtime embedded in the APK, can we say it produces natives apps?

 

[Erel]

As far as I know, Xamarin (on Android) runs your app in their own runtime engine however eventually it does call the native APIs so you can consider it a native app.

 

[KMatle]

With b4x you just code. For a simple app you don't need to know much. No external source. What you have works fine and easy.

On other languages or frameworks you have to understand a complete world to get an app running.

 

[udg]

Thank you all for your comments.
@Beja : submitting to a remote webservice will happen even if you install PG on a server of yours? If yes, I agree with you: too bad!
@Erel : so it's almost like having an app solely based on a WebView with everything done by javascript/HTML? Would it be useful to have a single code-base both for a web site and an app? I mean reusing most of what was coded for a web site in an app?
Well, we could do almost the same with B4J (did I tell you I simply love it? It's from Turbo Pascal 2.0 days that I don't feel like this.. !).

Thanks again to all.

 

[lemonisdead]

submitting to a remote webservice will happen even if you install PG on a server of yours

Yes because the developer version is allowed to test your app only and not to sell it

so it's almost like having an app solely based on a WebView with everything done by javascript/HTML?

You are right + some mobile functions built into the PhoneGap framework

Would it be useful to have a single code-base both for a web site and an app?

It could be interesting for some people only as you will never go deep in the system enough to make a really useful app. What about the services ? Will that "HTML5" app be able to send SMS without the user pressing the "send" button for example ?

In my opinion, PhoneGap is useful only for some uses (books, really light apps, etc) because you write once and publish to all the major platforms. But take a look at the prices, they are not cheap at all.

 

[udg]

you write once and publish to all the major platforms

B4J let me write code to be run wherever a Java VM is installed. And for light apps I could use the smartphone's browser to have those on mobiles too.

So far we found that PG buys you:
- a dev version that sends your code to a remote service
- a costly pro version
- a whole framework to learn even for basic apps
- a unique codebase with an eventual website just for simple apps

Is it correct/shareble to say that PG could be a choice only for pure website devs that look for a way to publish something on the mobile camp?

 

[socialnetis]

I'll say phonegap is a good approach just for light apps. What is more, with just one code you can deploy to all major plattforms.

But if the app is big (lets say facebook app) it would definetly crash. This is history, it happened in real life to facebook, twitter an linkedin when they started with html5 apps for smartphones. They had to went all the way native to stop people complaining. I dont know if you had facebook installed back to 2012-13 in your android device, it was just terrible and painfully slow (I had a Motorola defy back then, a very common device in south america).

Here is a very intersting answer in Quora:
http://www.quora.com/?digest_story=5866052
The question is "What is the most valuable programming language to know for the future and why?", some answers are Javascript, and the best best voted answer is an answer that goes a little offtopic saying why Javascript for mobile is not good, gives statistics, and tells why you should go all the way native.

If right now you are a little more convinced why native is better than hybrid apps. Then here is were B4X kicks in. You get all the power of native development with a very and easy to use framework. What is more, right now you can share almost the same code between iOS and Android apps! Why go hybrid when you can go all the way native in a more easy way with a powerful tool like B4X!?

 

[udg]

Why go hybrid when you can go all the way native in a more easy way with a powerful tool like B4X!?

Speaking about myself, I am more than convinced about my two-years old choice in favor of B4x tools. My knowledge about JS (and its world) is very basic and as long as I can avoid Eclipse, Java and the whole bunch I'll be happy.
I initiated this thread solely because I met someone using PG (he has a website dev backgroud) and I didn't seem to have enough arguments in favor of my choice, mainly due to my ignorance about PG that I never used so far (and, from what I learned here, I'm inclined not to use anyway).
So my request to collect your experiences and comments.

That said, it's time to complete my next B4A library: PushBullet Easy Pushes (I somewhat privately promised to @DonManfred to take care of it during Easter days but I've to shamefully acknwoledge that instead I took the family to the beach...). Anyway, it's coming...

Umberto