This is our code repository for our project submission for FreyHacks
- Zac: Full Stack Developer and Project Leader
- Luke: Frontend and Integration Developer
- Greyson: Frontend Developer and UI Designer
- Kellan: Frontend Developer and UI Designer
- Matthew: Backend Python Developer
- Scott: Backend Python Developer
- Hargun: DevOps Engineer
Project Description 📝:
As the weather gets warmer and more activties become accessible, it becomes increasingly difficult to chose what to do in your past time. Our team has developed an app that can solve that problem. our carefully formulated software is capable of processing local weather data and reccommending weather dependant activities based on current and future weather conditions. Not only is the schedule capable of reccomding activities based on weather conditions, the app can also be used to plan events for you and your friends all within the app.
Living in Muskoka, we see alot of tourists that come from the south. A fair majority of these people have trouble knowing what weather conditions are ideal to go sailing, paddleboarding etc... Our inspiration stemmed from wanting to make an app that can aid new cottage-goers to have more fun when visiting the cottage. This idea then stemmed out to not just watersports, but other sports such as basketball and archery.
What it does 📱:
The main objective of our app is to allow people to find the best time to do the activities they love. The app takes in your location and is able to provide a score from 0-10 based on how well the current and future weather conditions fit a certain activities.
Another feature of our app is meant to get people together, we did this by creating "events". events are scheduled activities that people can RSVP to and do activities together. People are able to find their friends within the app through the friends section. Companies can also make use of the events section by advertising kayak tours, sports tournaments, beer leagues etc..
Our main objective of the app is to get people active for the summer, we encourage users to do their favourite activities with or without friends at the most optimal time.
How we built it 💻:
- We built our mobile application in react native with TypeScript
- We used Bottom Tab Navigation as our primary navigation, then Stack Navigation as secondary navigation
- We used graphql to integrate our mobile app with our server
- We built our server in python with flask
- For graphql, we used Ariadne and SQLAlchemy
- We used PostgreSQL from our database
Custom Github DigitalOcean CI/CD Pipeline 🤯 :
- Automated by Github Actions on each push/pr to the main branch
- Builds docker image, and pushes to Docker Hub
- Authenticates by SSH into Digital Ocean droplet linux server for deployment
- Pulls image from Docker hub into remote server and runs container
- GraphQL domain is live: sum-fun.xyz/graphql
Challenges we ran into 💥:
The first intial roadblock our team ran into was acquiring a suitable API. Many APIs required a key, and a majority of the time the key requries a payment. Our team then found a suitable API that was free. However, the key could not be activated till hours later. This left us unable to make any progress on the weather aspect of our code until we acquiried a working key.
When designing a function for the event aspect of our app, we thought it would be a good idea to include a validity score. When trying to code this function, we intially realized that any event that was more than 48 hours away, we could not provide a validity report due to limitations in the API. The API can only provide an hourly report up to 48 hours. Another problem that we ran into was our API provided times in epoch time and convertering the users event date into epoch time became very time consuming, and we ended up scraping the event validity ideas.
Creating the data pipeline was quite challenging, since our DevOps engineer had taken a an unfamiliar route in its development, with new technologies. There was a lot of debugging and difficulty especially when trying to deploy the docker image to the remote server, as port configuration and credentials were finicky to set up. In the end, a robust pipeline was made that updates with each push/pr.
The GraphQL Library, Ariadne, had one suuuper annoying bug where when queries are too deeply nested, it will return a class instead of the dictionary that we had previously been receiving. This caused a lot of pain when trying to figure out why nested queries would stop rendering, and an odd solution of checking the instance of any query result before returning.
For most team members, they were learning something completely new. Most of us had never written in the language our whole job was in
Accomplishments that we're proud of 🍀🍾:
The majority of our team would consider themselves beginner coders. Despite the steep challenges and high goals, our team of inexperinced coders were able to overcome many difficult roadblocks.
Thanks to Hargun, we were able to set up a completely custom cidc pipline through github actions digitalocean ssh.
Being able to take real word data from the weather API and integrate it into our app to provide useful information for our user.
One of our biggest roadblocks was not being able to do more than one nested query within graphql. Being able to overcome this problem after many hours was a great accomplishment.
What we learned 🧠👍:
- For many team members, this was their first hackathon, or first big project, so there was a lot of learning.
- Learned how to access and integrate data from an APL
- Learned how to combine navigations on mobile.
- We all learned new technologies
- Python Flask
- React Native
What's next for Activity 🥚🐣🐥:
- Adding more paramaters to validity
- Getting a user base
- Adding better filtering for distances with coordinates, better mapping location system
- Improving UI
- adding validity score for events
- Proper ways to add friends, with qr codes etc.
- React Native with TypeScript
- GraphQL API in python with flask
Log in or sign up for Devpost to join the conversation.