April 7, 2021
6 min read
After 4 long months and 475 hours of development I have finally deployed my first mobile app to their respective app stores. In the grand scheme of things I am not sure if this is a long time for developing a mobile app or if it's a relatively quick development lifecycle. To get closer to a definitive answer, i'll break down what the app is and what's its scope, what technologies I used to create the app, the motivation behind the app and what I hope to get out of creating the app.
The app is business to client (b2c) management platform for a dog walking business. Our dog goes on walks twice a week with a group of other dogs to help him socialise and to help his confidence. He absolutely loves it 🐶 ♥️. The dog walking business managed all payments manually, used Apple notes to keep track of client's information and used WhatsApp to communicate to its clients. This was highly tedious and the dog walker would spend up to 4 hours every night doing admin work. The app was designed to streamline all the business processes so that the dog walker could focus on growing their business.
The app's functions include:
Firebase authentication handles the login tokens, signup and password reset functions for the app. This takes the legwork out of worrying about the app user's security.
Firebase's Firestore is my favourite NoSQL database to use for web and mobile applications. The response times are super quick and have the ability to implement listeners that can update the UI based on database document changes. This is especially useful for anything that requires real time updates such as chat messages or notifications.
Tying in nicely with the rest of the Firebase ecosystem, running a serverless back end gives me the ability to call function only when I need them rather than having an actual backend server on idle. These are similar to AWS Lambdas.
Also keeping with the theme of Firebase, all profile pictures, chat messages media and news feed media are stored on Google Cloud Storage buckets. The permissions system plays nice with Firebase Authentication and Firebase Firestore to ensure that all resources are protected on the app.
One of my biggest unknowns moving from web development into mobile development was how I was going to handle push notifications on mobile. I know Expo has their own push notification system however I wanted to steer clear from the Expo ecosystem. Firebase Cloud Messaging paired with
@react-native-firebase/messaging made handling push notifications a breeze.
Without setting up my own SMTP server I wanted to use a cloud service such as Twillio to handle the transactional emails for the app (such as account verifications, password resets, invoices etc.). Twillio offered 100 free email a day, however since the budget was quite tight for this project, SendInBlue was the alternative option which offered 300 free emails a day. Their API and user interface to compose email templates are super simple and easy to interface with via the Nodejs SDK.
The ability to pay and automate invoices was super important to the client. This task of taking payment and manually adding credits was very time consuming in the past. Much like the email server, the most popular option was not the most cost effective one. I had initially looked at Stripe. They charged 1.75% + 0.30c per transaction which ended up costing more than Square's offering of 2.2% per transaction (for most the transactions in this app which were less than $100).
There are many outcomes that I am anticipating as a result of creating the app. Firstly, it gives my toolbox a new set of diverse tools that can cater to a new set of devices. If I am able to offer an application that can run on a web browser, natively on mobile devices and also natively on desktop operating systems, I broaden my chances of finding work. Secondly, I hope to get a better understanding of UX design and how to affectively create applications that are appealing to the eyes and are a pleasure to use. I was lucky enough to have my partner, which is a UI/UX designer, design this app and critically think about the user journey. This is just something that developers do not think about enough. Lastly, I hope that this is my gateway into native app programming (Swift for iOS and Kotlin for Android). Although most things can be done with React Native, I still do believe a better experience can be programmed using native code. Writing code once and deploying to both platforms sometimes feels like you are programming for the lowest common denominator instead of harnessing the full potential of the platform's capabilities.