Firebase

Chris Oryschak
 Product Manager

Last year, we launched the beta of Cloud Functions for Firebase to help you build backend functionality for your app without worrying about managing servers. It lets you write in JavaScript or Typescript, deploy to Google's Cloud infrastructure, and execute in response to app events. Today we're pleased to release version 1.0 of the Cloud Functions for Firebase SDK - along with new, frequently requested improvements for the development, testing and monitoring of Functions.

What Cloud Functions for Firebase can do

The Firebase SDK for Google Cloud Functions enables you to extend other Firebase products, giving users the features they need in your apps. One of our earliest customers, Posse, makers of the Hamilton app, needed to create a lottery system to give fans a chance to win tickets to the massively popular Broadway play - without the overhead of managing their own servers. When a user enters the lottery, Functions powered logic runs in the cloud with results stored directly to Cloud Firestore. Lottery results are pushed automatically to lucky fans' Hamilton mobile app.

Posse also needed to scale the Hamilton app. When the lottery opens, the app's usage can suddenly surge multifold before normalizing a few minutes later. Functions automatically scales without needing any action from Posse, and their client only pays for the resources they use.

In addition to ticket lotteries, Posse also used Functions to integrate with other Firebase products to:

  • Automate creation of user profiles with Firebase Auth
  • Send push notifications with Firebase Cloud Messaging
  • Auto-resize user photos with Cloud Storage
  • Process in-app payments using Stripe via HTTP webhooks

There's a great deal of things you can do with Functions. For more use cases and sample code, see our popular GitHub repo.

Today's v1.0 release comes with a number of new features based on your feedback - here's the lowdown!

Mobile client SDKs

One of your biggest requests was for an easy way to directly call Functions from your client apps. To facilitate that, we're adding a new type of function, along with SDKs for iOS, Android, and web clients. The new type is a Callable HTTPS function that manages an HTTPS request from the client SDK. Callable HTTPS functions drastically reduce the amount of boilerplate code you have to write to make HTTPS requests to Functions.

Here's how it works:

  1. First, write a Callable HTTPS function that accepts a data object from the client. The function returns another data object to the client.
  2. Then, in your mobile app, you use the client SDK to invoke the function. The client SDK manges the HTTPS connection and automatically serializes the input and output data objects. In the request, the SDK will include a user ID token from Firebase Authentication if a user is logged in. Also included is an instance ID token used by Firebase Cloud Messaging.

For more, check out our docs!

Easier unit testing

While Functions Shell gave you the ability to interact with your functions locally, it was not optimal for writing unit tests such as those that run in a CI system. That's why we are launching a new firebase-functions-test npm module that simplifies writing unit tests. firebase-functions-test takes care of the necessary setup and teardown, allowing easy mocking of test data. Now, you can either write completely offline tests that have no side effects, or tests that interact with a development Firebase project - where you can observe the success of actions such as database writes.

Official support for functions emulation

Last fall, we introduced the ability to emulate your functions locally via the Firebase CLI through both the "firebase serve --only functions" and "firebase experimental:functions:shell" command. These were experimental features, and now we are officially supporting them. "firebase experimental:functions:shell" has been renamed to "firebase functions:shell", and "firebase serve" will emulate all HTTPS functions by default without the need for the "--only" flag.

Improved monitoring tools

After deploying functions, you may be wondering: "What's happening with my functions? How is my app performing? Is anything broken?" With today's release, you can now keep tabs on any errors in your functions via the new health monitor:

Additionally, you can track your functions' performance, latency and memory usage:

API updates

The APIs for writing Cloud Functions with the Firebase SDK have changed in this v1.0 SDK release. This means that upgrading to the new SDK will require some updates to any code you've already written during the beta. But don't worry - this doesn't affect any functions you have already deployed. To see what has changed and how to adapt to the new format, just follow our migration guide!

Getting started with Cloud Functions for Firebase

If you're just getting started with the Firebase SDKs for Cloud Functions, try following our step-by-step Codelab and visiting the documentation. There is also a video tutorial to help get you set up using TypeScript as the recommended programming language.

We hope you find these new features helpful. While the Firebase SDKs for Cloud Functions are fully released, the Cloud Functions platform is still in beta. Stay tuned for more updates!

Labels: , , ,

Michael Hermanto
 Software Engineer
Originally posted on the Google Developer Blog

We launched the Google URL Shortener back in 2009 as a way to help people more easily share links and measure traffic online. Since then, many popular URL shortening services have emerged and the ways people find content on the Internet have also changed dramatically, from primarily desktop webpages to apps, mobile devices, home assistants, and more.

To refocus our efforts, we're turning down support for goo.gl over the coming weeks and replacing it with Firebase Dynamic Links (FDL). FDLs are smart URLs that allow you to send existing and potential users to any location within an iOS, Android or web app. We're excited to grow and improve the product going forward. While most features of goo.gl will eventually sunset, all existing links will continue to redirect to the intended destination.

For consumers

Starting April 13, 2018, anonymous users and users who have never created short links before today will not be able to create new short links via the goo.gl console. If you are looking to create new short links, we recommend you use Firebase Dynamic Links or check out popular services like Bitly and Ow.ly as an alternative.

If you have existing goo.gl short links, you can continue to use all features of goo.gl console for a period of one year, until March 30, 2019, when we will discontinue the console. You can manage all your short links and their analytics through the goo.gl console during this period.

After March 30, 2019, all links will continue to redirect to the intended destination. Your existing short links will not be migrated to the Firebase console, however, you will be able to export your link information from the goo.gl console.

For developers

Starting May 30, 2018, only projects that have accessed URL Shortener APIs before today can create short links. To create new short links, we recommend FDL APIs. FDL short links will automatically detect the user's platform and send the user to either the web or your app, as appropriate.

If you are already calling URL Shortener APIs to manage goo.gl short links, you can continue to use them for a period of one year, until March 30, 2019, when we will discontinue the APIs.

As it is for consumers, all links will continue to redirect to the intended destination after March 30, 2019. However, existing short links will not be migrated to the Firebase console/API.

URL Shortener has been a great tool that we're proud to have built. As we look towards the future, we're excited about the possibilities of Firebase Dynamic Links, particularly when it comes to dynamic platform detection and links that survive the app installation process. We hope you are too!

Labels: , ,

David East
 Developer Advocate

All too often I become frustrated when building an authentication component. The asynchronous flow is hard to manage and it makes my template a mess of ngIfs and Elvis operators: 

{{ (user | async)?.uid }}

The problem is always same. I'm mixing too many concerns in one place. Organizing the right code in the right place is what makes a happy codebase.

The solution: Store route specific code in Angular's Route Guards.

Step 1: Identify Route code in your component

The AngularFireAuth service is a great match for routing code. It contains information about a current user, whether they are coming from a redirect, and all sorts of other things.

Take a simple signInWithRedirect example: 

import { Component, OnInit } from '@angular/core';
import { AngularFireAuth } from 'angularfire2/auth';
import { Router } from '@angular/router';
import * as firebase from 'firebase/app';

@Component({
  selector: 'app-login',
  template: `
    <button class="btn" (click)="auth()">
       Login
    </button>
  `,
  styleUrls: ['./login.scss']
})
export class LoginComponent implements OnInit {

  constructor(
    private afAuth: AngularFireAuth,
    private router: Router
  ) { }

  ngOnInit() {
    this.afAuth.getRedirectResult().then(result => {
      if(result) {
        this.router.navigateByUrl('profile');
      }
    });
  }

  auth() {
    const provider = new firebase.auth.GoogleAuthProvider();
    this.afAuth.auth.signInWithRedirect(provider); 
  }

}

Here's what's going in this example:

  1. The user is logged in with the auth() method.
  2. The signInWithRedirect() method triggers a redirect to sign in with Google.
  3. After the user selects their account, they are taken back to this component.
  4. The getRedirectResult() promise in ngOnInit() resolves the logged in user.
  5. The router navigates the user to the profile page via the navigateByUrl() method.

This isn't a problem with one component. However, imagine having to manage redirects in multiple components.

The code in ngOnInit belongs in a Route Guard. This will clean up the component of routing logic and remove the Router dependency entirely.

Step 2: Create a Route Guard

A Route Guard is used for all sorts of useful things. In this case a guard can handle the redirect code. 

@Injectable()
export class RedirectGuard implements CanActivate {

  constructor(
    private authService: AngularFireAuth,
    private router: Router
  ) {}

  canActivate() {
    this.authService.auth.getRedirectResult().then(result => {
      if(result.user != null) {
        this.router.navigateByUrl('profile');
        return false;
      }
      return true;
    });
  }

}

The canActivate method determined if the user has access to the route. Here's what happens:

  1. The user visits the route.
  2. AngularFireAuth checks if a redirect result exists.
  3. If a user comes back from a redirect, route to the 'profile' route.
  4. Else, go to the current route.

This code is now reusable. The Angular Router lets you specify which routes you want to listen for a redirect. 

import { Routes, RouterModule } from '@angular/router';
import { LoginComponent } from './login/login.component';
import { RedirectGuard } from './redirect.guard';

const ROUTES: Routes = [
  { 
    path: '', 
    component: LoginComponent, 
    canActivate: [RedirectGuard] 
  },
  { 
    path: 'profile', 
    loadChildren: 'app/profile/profile.module#ProfileModule'
  }
];

const AppRoutes = RouterModule.forRoot(ROUTES);

export { AppRoutes }

Now the LoginComponent is free of routing code: 

@Component({
  selector: 'app-login',
  template: `
     <button class="btn" (click)="auth()">
       Login
     </button>
  `,
  styleUrls: ['./login.scss']
})
export class LoginComponent {

  constructor(private afAuth: AngularFireAuth) { }

  auth() {
    const provider = new firebase.auth.GoogleAuthProvider();
    this.afAuth.auth.signInWithRedirect(provider); 
  }

}

No more async flow mess. The right code is organized in the right place. You can add the RedirectGuard to more routes as the app grows, which keeps the complexity out of your components.

Labels: , ,

Tyler Rockwood
Software Engineer

Dashboards are great, but what if you're not checking them? Wouldn't it be great to know when you have a huge spike in traffic, or if you're about to hit your concurrent connections limit for a single database? Don't worry, Google Stackdriver Alerts have you covered!

Back in December, we announced that the Firebase Realtime Database has integrated with Google Stackdriver to let you create dashboards and graphs of powerful new metrics. We also have more in-depth descriptions of these metrics available in our documentation, as well as other monitoring and debugging tips for your databases. Now we've made all this data actionable through Google Stackdriver Alerts!

You can set alerts that will trigger actions whenever particular conditions are met. For example, if the metric io/database_load goes over 90% for more than one minute, an email will go out to your engineering team. It's also possible to include documentation in your alert, which can help if you haven't yet had your morning coffee. To learn more about conditions that can trigger alerts, see the Condition details documentation.

In addition to the plethora of available conditions, Google Stackdriver also lets you choose how to be alerted. On the basic tier, you can be notified via email or the Google Cloud Console Mobile App. On the premium tier, you're able to create an incident via PagerDuty or have a message posted to your Slack Channel. One of my personal favorite premium notifications is the Webhook option, which you can even use to trigger a Cloud Function to help you diagnose the issue. For the load example, when your database is nearing 100% load, you could start a Cloud Function, have it profile your database, then report the findings to your Slack Channel.

Setting up a custom alert in Stackdriver is a straightforward process. Start by heading over to the Stackdriver Monitoring Console. Once you've signed up for an account, navigate to Alerting > Create a Policy. You should see a panel like this:

For this alerting policy, I've specified that when my database utilization is over 90% for more than a minute, I'll receive an email at alerts@you.com with the name of the policy, plus the documentation I specified. Now all I need to do is hit the Save Policy button, and my alert is live!

Labels: , , , ,

Darin Hilton
 Art Director

Firebase is returning to the Game Developers Conference this year in San Francisco! We've been hard at work making Firebase more valuable for all you game developers out there, and would love to spend time talking with you about how we can make it even better.

While we'll be available to chat at booth 823 for the rest of the conference, we also dropped by the Google Developer Day on Monday to share a few exciting announcements with everybody. In case you missed it, here's the quick summary:

Google Play Games Services support in Firebase Auth

Firebase Auth already lets your players log in safely and securely through a variety of ways -- everything from using their Facebook or Google account to a custom email and password -- but we've now added support for players to sign in using Google Play Games Services (GPGS).

If you're currently using GPGS in your game, you can use the credentials you already have to authenticate your users' accounts with the Firebase service. And if you haven't used GPGS in your Android games yet, you really should because it's an easy -- and often frictionless -- way to get your players signed in. Either way, this means your game can start using Firebase to store user-specific data in Realtime Database or Cloud Storage as soon as your players are signed in through GPGS.

Desktop support for C++ and Unity development

While many of you are developing games for mobile devices, we know that it's often much more convenient to build and test your game on a desktop machine. With that in mind, we've added desktop support so that Firebase will work right from your development environment (Unity player on a Windows machine, for example). This support includes products like:

  • Authentication to manage sign-in
  • Realtime Database to sync and store data to the cloud
  • Cloud Storage for Firebase to store and share large files like images or movies
  • Remote Config to easily change elements of your game from the cloud

So if your game relies on one of these products, you can continue to use the same development process you always have, without the hassle of deploying your game on a device every time you want to test it.

Note that desktop support is intended to make development and testing easier, but it's not for publishing Firebase-enabled games to end users on desktop machines.

Firebase Predictions and A/B Testing

We debuted Firebase Predictions and A/B Testing at the Firebase Developer Summit in Amsterdam a few months ago, but we think both of these features are so useful to game developers that we want to mention them again. :)

Firebase Predictions brings the power of machine learning to Firebase. By analyzing your players' behavior, we can predict which ones will be most likely to churn out, spend money, or perform any other conversion event that you've defined in Analytics. This allows you to customize your app's behavior (using Remote Config or Notifications) specifically to that group of users. Want to engage players who are likely to stop playing your game? Give them extra lives, or send them a free gift they can redeem inside your game! Want to target players who are likely to spend money on in-app purchases? Notify them about a big sale, or change the layout of your in-app store!

In fact, Rockbite Studios did just that. They used Remote Config to feature certain items at the top of their in-app store for players who were predicted to spend money in their game, while also adding new ad-viewing options for players who were predicted to not spend money. As a result, they boosted revenue by 25% for their spenders, and 24% for their non-spenders!

Firebase A/B testing is great way to improve your game by making data-driven decisions based on experiments that you can run inside your app. Want to know if increasing the difficulty of your game will increase or decrease player engagement? Curious to learn if increasing the amount of premium currency your players start off with will increase or decrease your overall revenue? You can run experiments to uncover answers to all of these questions (and much much more) through Firebase A/B testing.

Firebase A/B testing is easy to get started with, as it runs on top of Remote Config and Analytics, which you might already be using. Experiments can be set up entirely through the Firebase console, which means your game designers or analysts can build, publish, and analyze experiments on their own without needing too much time from your engineering team. And it's all available at no cost to you, which is always nice.

Come find out more!

If you want to learn more about how Firebase can make it easier for you to build successful games, come talk to us! We'll be available at booth 823 all during the expo, and we're giving a special presentation on what's new with Firebase and the Google Cloud Platform on Wednesday, March 21st, at 3:30 in the South Hall.

And of course, you can always find out more at https://firebase.google.com/games/

Game on!

Labels: , , , , , , , , ,