I was able to build and run a corona test app with Android Studio. No problems. A native build is not the easiest thing to do. You have to configure each plugin you use. In the case of Appodeal you’ll have to configure each of the networks that you decide to add.
Does this indicate a Corona build issue, as you have successfully built the Appodeal plugins natively, targetting the Amazon market?
I did a native build of one of my apps with the exact same plugins that Corona uses to build from the simulator. I didn’t create my own plugins. The only difference is that I did not add the dependency for Google Play Services from this plugin:
['plugin.appodeal.beta.base'] = { publisherId = 'com.coronalabs' },
The process is sort of simple but tedious:
-
Copy the entire “Apps” folder from the native sample.
-
Replace the corona folder in this folder with your own app folder. Make sure to keep the name as Corona.
-
From within Android Studio open the project in the Android studio.
-
Open a command line and navigate to the android studio and run the .sh or .bat script. This will download all the plugins.
-
Go to the plugin folder. Copy the jars in each folder to the libs folder in Android folder.
-
For all the pure lua plugins copy the lua files into a “plugin” folder in the corona folder.
-
Each plugin that downloads comes down with a “metadata.lua”. Here you’ll see the depended .jar (which you have to copy to the libs folder), permissions, extra activities and other random manifest stuff (that you copy to the manifest file). If permission or activity is a duplicate don’t copy it again.
-
After that, all you have left is to modify the manifest with your app name, supported version and build.
So to answer your question it is not a build issue, it is a problem with how the plugin for Appodeal is currently built. And the word problem might be a strong word, Appodeal might have their reasons for including that dependency that might not have come up with my limited testing.
Thanks for the great info, agramonte.
Is this an issue with the Appodeal plugin, or Corona’s build system then? Appodeal owns Corona so at some point I wish they would just resolve the issue between them. Seems like they are two different teams when I am communicating with them…
It is an issue with the Appodeal Corona plugin not with the build system.
As of right now, there is no Amazon store Appodeal SDK modular build, so it’s not a Corona plugin issue, sadly, we can’t fully support Amazon store builds on a modular plugin, since Appodeal doesn’t have an SDK for that atm.
Appodeal team is already working on Amazon store Appodeal SDK modular build for quite some time, but they can’t give an ETA for now, I will update you on this.
I’m really hoping that you can hear me out on that subject, guys, I thought it was quite simple to understand what exactly this issue is about, let me elaborate.
Before said SDK for Amazon builds is ready, all we can do is put a stub for Amazon builds, so that you can’t build an app for Amazon store with the Corona Appodeal plugin at all. To avoid further misunderstanding, we are considering doing that, the only thing stopping us is that some users, like agramonte, can find a way to bypass said restrictions, although it’s definitely not recommended to not include dependencies, on which Appodeal SDK is relying.
If you have any more questions, feel free to ask them, it seems that subject may be hard to understand.
Btw, using a Corona Native is basically means that you’re building an app with Corona, but take a responsibility of native development, so it’s up to you to decide, what native functions, APIs, third-party SDKs and dependencies you’ll be using.
Thank you both. Looking forward to when Corona + Appodeal have a plugin available that targets Amazon.
Although the work on the said SDK has been going on for quite some time, as I’ve said, I’ve pointed out the importance of the said SDK for Corona users to responsible Appodeal employees, so that the SDK should be prepared asap. It may be hard to prepare all adapters and base of modular system to be Amazon store compliant, though, hope for your understanding, because I can’t really comment on that.
I hope you don’t stub it out. It works for me. Maybe you can put it on the documentation “AMAZON BUILD IS NOT SUPPORTED” in red, bold and all caps.
And Amazon is a misnomer. I am trying to get my app into other 3rd party stores that don’t necessarily need the Google stuff (Yandex and Aptoid to name 2).
Actually, the docs were updated with the warning of that kind, the web page should be updated soon.
Hi;
I’m adding a post to this thread because we have published quite successful word games on Kindles for 4+ years and we have ALWAYS been challenged in finding solid & stable advertising alternatives.
Of course I voted on the Features Request (years ago) to have Corona offer an “Amazon Ads” plugin. (Link - http://feedback.coronalabs.com/forums/188732-corona-sdk-feature-requests-feedback/suggestions/5646802-add-support-for-amazon-mobile-ads-api))
I’d even pay the $199 “premium” for it if it was available – as long as the plugin supported the three main app store types (which it does in its SDK from Amazon).
But I also realize that Appodeal is the parent here . . . so getting things set so that the Appodeal Base with the “Amazon Ads” feature works properly and fully should be the most direct path to a solution.
An interesting article ran in the WSJ about a week ago. The title was:
Amazon, With Little Fanfare, Emerges as an Advertising Giant
Its push challenges the big ad sellers, including Facebook, Google and TV networks
I’m not including a link to it as you have to be a subscriber to read the whole thing (easy to Google the title if you are a subscriber). But here are three key paragraphs to convey the gist of the article’s message:
Amazon.com Inc. handles nearly half of all online sales in the U.S., giving it a popular platform and a wealth of consumer data. Now it’s on track to become the next juggernaut of online advertising, and its rise threatens to upend Silicon Valley’s ad titans and change the way business is done on Madison Avenue.
The online retailer has ascended to the No. 3 spot in the U.S. digital ad market behind the dominant players, Alphabet Inc.’s Google and Facebook Inc. Though Amazon has just 4% of the market now, the company is expanding its avenues for marketers and hiring aggressively for its ad unit.
Amazon is expected to collect 15 cents of each new dollar spent on U.S. digital ads in 2020, up from 5 cents last year, according to an analysis of data from research firm eMarketer.
Better monetization options on Amazon Kindles would make for better overall monetization for many of us that depend upon Corona’s great SDK for building advertising-based apps for all 3 major app platforms.
Steve
Hi Steve,
Thanks for your feedback!
I as a person that has native apps also want to mention that although Amazon Ads sounds good on paper and their eCPM is sort of competitive their fill rate is fairly low. In many countries, it is 0% and even in the USA, it is not great. I just looked at my data and in the recent month (October 2018) when I was using them: 17K impressions, fill rate of 12.38% and an eCPM of $1.61. All interstitials. All USA. All on iOS. Appodeal usually gets me at least 2x that in the USA with very close to 100% fill rate. Although I don’t have large numbers for Android devices they are not much better (actually worse).
Hi Agramonte;
Thanks for the info. Good to have input on actual numbers for “Amazon Ads”.
Is the fill rate 0% on Kindles (Amazon App Store Builds)? That is my primary concern.
We use our own little waterfall on iOS and Android (Admob, Chartboost, AppLovin, Facebook) so we probably wouldn’t even do “Amazon Ads” in the waterfall if the economics and fill rate is that bad.
Our focus is on whether the economics are good on Amazon-specific devices.
Steve
If I remember correctly it is less than 10% fill rate in the USA on Amazon devices. Go to the Amazon forums under services and search for "fill rate date:“2018-08-05 TO 2018-12-05”. You’ll notice many responses from people mentioning the fill rate being very low or 0% (in Europe). If you want to do a native build for Corona, I believe I have a plugin on my machine that I can put on github.
Thanks for the additional info. Certainly important to ponder.
And I appreciate your plugin offer for a native build. I’ll keep it in mind going forward.
Steve
Is there any fix to this issue? Talking to the Appodeal people it looks like they will never fix Appodeal + Corona + Amazon error about the Google Play Services. Excerpts of my chat with the Appodeal people in the past few days. Is there anything that Corona can do to help out with this?
Appodeal plugin for Corona for Amazon store are restricted at the moment.
Hello. Unfortunately, we won’t release new version of corona plugin. We won’t support it anymore. It’s so because is’is not so popular as rest of frameworks and we decided to concentrate on other frameworks to provide better quality of our product.
I wish they would open source it. I could easily maintain it.
Funny you guys posted here just now. Yesterday I received two followups from Appodeal stating the same thing you did: they are no longer supporting Corona SDK. Which is truly bizarre given that Appodeal BOUGHT Corona SDK.
Appodeal bought Corona out and then the paid ad plugin model showed up, basically pushing everyone towards Appodeal. But then Appodeal stopped supporting Corona? Was this just to kill Corona?
Sucks because we have a large portfolio built on Corona and this was just the latest issue with Corona that is leading us to migrate to a different platform. We simply can’t afford to be invested in a platform that may not exist. Sucks because it’s going to be a LOT of work to migrate.
Good luck guys.