Currently, there is no easy way to match all that are in need with those that are providing help, at the same time is difficult to find all the information at the local and global level of the different initiatives, projects, researchers, donors, investors, FREE or discounted tools for builders, the grants available and other relevant information. That's why Fight Pandemics was born.
Please check our notion to get a better understanding of the problem that we are trying to solve: https://www.notion.so/fightpandemics/FightPandemics-Overview-cd01dcfc05f24312ac454ac94a37eb5e
Before getting started please be familiar with the stack below. Additionally it is important to understand basic Git and Github for smooth collaboration as a team.
- Frontend
- React Hooks, Ant Design, Ant Design Mobile, styled-components
- Backend
- Fastify, MongoDB, Mongoose, Redis, Socket.io
- Auth0, Google Maps API
- DevOps
- Docker, Terraform, Nginx, AWS
- Fork or Clone this repository
- There are two ways to set up the app: Docker or Local Setup. It is strongly recommended to use the Docker Setup, but if you are unable to install Docker, you have the option of setting up the app locally.
If, at any point, you are having trouble setting up the app, please use Slack's search bar to search for exisiting questions. Also do not hesitate to ask us for assistance in the #engineering
Slack channel! In order to speed up assistance, please mention the method that you are using to set up the app, as well as the operating system that you are using.
- Follow the steps for your OS to install Docker and Docker Compose. Note that depending on the Docker install, Docker Compose may already be included, so it would not be necessary to install Docker Compose separately.
- Copy
backend/.env.example_docker
tobackend/.env
- Replace
TODO
entries in the.env
file with correct values (this is not needed to run the project in a responding but non-functional state).- For
AIRTABLE
andAUTH
variable values, either use Slack's search bar to search for exisiting requests, ask in the #engineering Slack channel, or consult the non-engineer guide.
- For
- Run
docker-compose up
this will take quite some time. (Starting the development server...
is not the final line). - Finally, navigate to localhost:3000 in your browser - the page title should be "Fight Pandemics" and you should see a styled page.
- To import posts data from Airtable run
docker-compose exec backend-service npm run import-posts
. By default 100 records are returned. To get a specific number of records pass a numeric argument, e.g.docker-compose exec backend-service npm run import-posts -- 10
. Use-1
to get all records (~2500 as of this writing).
If, for some reason, you are unable to use Docker, you can still set up the app locally. Note that you can set up each of these three services separately, but some functionality may not work. For example, if only the client is running, none of the calls to the backend will work.
Also note that features that require the Localstack Docker container to be running will not work. The following features will not work with local setup:
- Avatar upload
In order to test these features, please deploy to the review environment; see the Review branches section for more details.
Follow the MongoDB installation instructions for your operating system.
Follow the Redis installation instructions for your operating system.
- ubuntu: Follow this guide to install redis server on ubuntu.
- other: Follow these installation instructions.
Follow the installation instructions for windows.
Follow this guide to install Redis on Mac OS via Homebrew.
- Install nvm (for Mac/Linux), or nvm-windows (for Windows).
- Using nvm, install Node 12.16.2:
nvm install 12.16.2
. - Enter the
client
directory and runnpm install
. - Start the client app by running
npm start
, and wait for the app to start up. (Starting the development server...
is not the final line). - Finally, navigate to localhost:3000 in your browser - the page title should be "Fight Pandemics" and you should see a styled page.
- If not already installed, install MongoDB.
- If not already installed, install Redis.
- If not already installed, install nvm and Node 12.16.2 (see steps 1 and 2 in the Client section above).
- Enter the
backend
directory and runnpm install
. - Copy the
.env.example_local
in thebackend
directory to.env
. - Replace
TODO
entries inbackend/.env
with correct values (this is not needed to run the project in a responding but non-functional state).- For
AIRTABLE
andAUTH
variable values, either use Slack's search bar to search for exisiting requests, ask in the #engineering Slack channel, or consult the non-engineer guide.
- For
- Start the backend server by running
npm start
. NOTE: If you want the server to automatically restart on code changes, runnpm run dev
instead. - The backend can be accessed at
localhost:8000
using cURL, Postman, or a similar API testing tool. - To import posts data from Airtable, from the
backend
directory, runnpm run import-posts
. By default 100 records are returned. To get a specific number of records pass a numeric argument, e.g.npm run import-posts -- 10
. Use-1
to get all records (~2500 as of this writing).
We are using the Google Maps API via our backend for geolocation requirements. It is not required for most frontend and backend development and is mocked with a limited set of fixed results when no key is present. This is to reduce friction for developer onboarding. If you need to work on this feature ask in Slack engineering channel about how to get your own development key.
Check the issues and projects tab of the original repository for available tickets. Please follow our instructions on how to properly open an issue and make pull requests to our repo.
Additionally please review the "Important Notes and Considerations" section in this README.
Due to the importance of what we are trying to accomplish and the fact that some features are critical for certain releases, sometimes issues need to be completed within a specfic timeframe. In order to ensure that this can happen, it is very important that you communicate your progress with the team in either the engineering or general standups. Please also try and push your code to GitHub every day that you work on an issue.
In the event that there is an upcoming deadline for a particular feature, someone from the team may try and contact you to see what the current status is and if you possibly need help with the task. If we don't hear anything from you after 48 hours, another engineer might have to take over the issue and it will be re-assigned on GitHub. Therefore it is CRITICAL that you have your current code pushed to GitHub so that others can possibly use your work as a reference to ensure the issue gets completed as quickly as possible.
To check the API documentation which is automatically generated using fastify-oas , go to http://localhost:8000/api/documentation
Note that whenever you add a new NPM dependency, you must run npm install
from within the container. This is because
NPM dependency installs may not necessarily be cross-platform. There are two ways that you can install the dependencies
within the container:
Run docker-compose run backend-service npm install
, or cd
into the backend
directory and run npm run install-docker
.
Run docker-compose run client npm install
, or cd
into the client
directory and run npm run install-docker
.
Be sure to also commit any changes to the package-lock.json
so that dependencies used by third-parties are also locked to specific versions.
We are noting any special considerations and handling done in the code so that developers are aware of these caveats while developing. This is a living document, so feel free to add any notes that you feel are worth mentioning here.
We collaborate closely with the design and product team. The design team provides us with designs of the application on Figma. While the product team provides us with MVP requirements on Notion.
- Please be sure to often merge or rebase the latest from the staging branch into your feature/working branches to minimize merge conflicts and so that it doesn't fall too far behind staging.
- If possible, try to keep changes in pull requests small so that it'll be faster for reviewers to review and easier for contributors to make revisions. If you foresee there will be many changes, make sure to commit often. This will help break down a pull request more easily.
- API requests are validated using Ajv and JSON schema's are defined with FluentSchema
- By default, the schema's allow additional properties unless explicitly set that additional properties are not allowed.
There's a utility function
strictSchema()
inlib/endpoints/schemas/utils.js
that can be used to initialize all schema's. - Rather than silently suppress additional properties, as is the default behavior for Fastify's ajv configuration, we are instead returning a 400 bad request error if additional properties are passed in. This makes it easier to debug issues due to a misspelled property name.
- Read more about validation in Fastify in the docs https://www.fastify.io/docs/v2.2.x/Validation-and-Serialization/#validation
- By default, the schema's allow additional properties unless explicitly set that additional properties are not allowed.
There's a utility function
- Endpoint handlers should use async functions, and return values or throw httpErrors.
- This is using some sensible defaults provided by https://github.com/fastify/fastify-sensible
- If an error is thrown, a generic error handler returns a 500 Internal Server Error response and logs the error details, without returning all error details in the response.
- Some more useful tools are provided by fastify-sensible such as built-in
fastify.httpErrors
andfastify.to()
to wrap async-await - Read more in the Fastify documentation about using asyc/await with fastify https://www.fastify.io/docs/latest/Routes/#async-await
- We have the beginnings of a theme. The file can be found here:
src/constants/theme.js
which has sections for typography, colors, buttons styles and media queries.- Limit the use of inline styling.
- Limit the use of
px
values, currently the root font size is 62.5%,1rem = 10px
- No use of
.css
or.scss
files, please style with styled-components instead.
- Use assets such as icons from the designs on Figma
- Select the screen that you are going to design
- On the left side of Figma, unlock the screen (Only the one that you are going to use)
- Select an icon or component -> right click -> Copy/Paste -> Copy as CSS/SVG
- You can also get exact CSS values on the top right of Figma
- After you are done with your screen, lock it again!!
- Use absolute imports. We have a resolved path of
client/src
. For example, import like this:import Button from "components/Button"
, rather thanimport Button from "../../components/Button
. - For most components, we can use components from Ant Design and override styling where needed. Please refrain from reinventing the wheel.
import styled from "styled-components"; import { Button } from "antd"; const StyledButton = styled(Button)` // target antd selectors and override styles here `;
- For new pages, create a new route in
client/src/routes.js
and import a page component fromclient/src/pages
- Refrain from making the Redux state too big as it will affect speed performance. React’s Context API is ONLY helpful for avoiding nested prop threading so stick with Redux where Context fails.
-
We are using i18next to manage localization on our platform. If you need to add text or change existing text, please do so in the
src/locales/translations/en_US
file. Lokalise is the tool we use to manage our translations. It is configured to automatically pull changes inen_US
fromstaging
branch and edit or add new keys on the Lokalise platform so our volunteers can go and translate the new changes. -
In order to not bottleneck development, keys that have not been translated in a language will be filled with the English translation. Once translations are completed on Lokalise, it will then initiate a PR to merge the translations into
staging
-
When a Github issue is started, create a branch with either the
feature/
orservice/
prefix. Most branches will usefeature/
prefix, but if you are developing a microservice, you must use theservice/
prefix. -
Every push to branches with the
feature/
prefix will automatically deploy to the review AWS environment. See the Review branches section for details on how to access this build. -
Every push to branches with the
service/
prefix will deploy the appropriate microservice to the review environment, depending on where the code change is located. Be careful not to have multiple branches open for the same microservice, as deployments can clobber each other (there is only one instance of the microservice in the review environment). All microservice code is located in theservices
folder in this repo. -
After a pull request for a
feature/
orservice/
branch is approved, it is merged into thestaging
branch. This will trigger an automatic deployment to the staging AWS environment. Note that changes may take several minutes to take effect. The staging app is accessible at https://staging.fightpandemics.work. -
When we are ready to release to production, a pull request is opened to merge the
staging
branch toproduction
. After this pull request is approved and merged, it will trigger an automatic deployment to the production AWS environment. Note that changes may take several minutes to take effect. The production app is accessible at https://fightpandemics.com -
A
hotfix/
branch is created if there is a critical fix that needs to be deployed to production as soon as possible, but staging has untested code that is not safe to deploy to production.- This branch must be created off the
production
branch. - Every push to
hotfix/
branches will automatically deploy to the review AWS environment. See the Review branches section for details on how to access this build. - The pull request for a hotfix must be opened to merge into the
production
branch rather than thestaging
branch. - Hotfixes must be tested in the staging AWS environment before being merged to the
production
branch. In order to deploy only the hotfix to the staging enviroment, create and push a git tag with ahotfix-
prefix. This will trigger a deployment to the staging environment:git tag -a hotfix-<GITHUB_ISSUE_NUMBER> && git push origin hotfix-<GITHUB_ISSUE_NUMBER>
. Note that if there are changes in thestaging
branch that are not yet in theproduction
branch, this deployment may impact testing done by QA in the staging environment, so it is best practice to coordinate with the QA team before deploying a hotfix to the staging environment. - After QA has tested the hotfix in the staging environment, the hotfix PR is merged to the
production
branch, where it will automatically deploy to the production AWS environment. - After merging the hotfix into the
production
branch, open another pull request to backmerge the hotfix into thestaging
environment (i.e. mergeproduction
intostaging
). TheCreate a merge commit
merge method should be used for this pull request.
- This branch must be created off the
-
All pull request merges to the
staging
branch, and allhotfix/
merges to theproduction
branch are done using theSquash and Merge
strategy. This allows for a cleaner commit history by combining all commits in a branch into a single commit. The message for this squashed commit must use the following convention:
<GITHUB_TICKET_NUMBER(S)> - <BRIEF_TICKET_DESCRIPTION>
- Additional comments if necessary, for more context
Example:
#1127/#1130 - Update Git Branching Model
- Update Readme
- Update Github workflows
- Update PR template
If there are too many Github ticket numbers to include in the title of the commit, make a list of ticket numbers in the body. Example:
Adding dummy feature
Resolves:
#1234
#1235
#1236
#1237
- When merging
staging
intoproduction
, and vice versa, theCreate a merge commit
merge method should be used to preserve commit history.
Every time you push code up to this repository on a branch with the feature/
or hotfix/
prefix, a review build based off of your feature branch will be deployed to AWS. For the build to deploy successfully you must be a member of this organization (ask in Slack) and push to this repo. Pull requests are still welcome from forked repos, just omit the feature/
or hotfix/
prefix to skip this build step.
You can view the build logs in GitHub Actions. After the build successfully completes, you can view the URL to which your app was deployed by clicking on the "Deployment URL" step in the Deploy Review Build
job in the workflow run for your build:
Note that it may take a few minutes for the app to be accessible, or for changes to be reflected, since it takes time for AWS to spin up the Docker containers.
Please note that review builds are shut down if no new commits are pushed to the branch in more than a day. This is controlled by an AWS Lambda function that runs every day at midnight PST. In order to revive the review build, simply push a new commit to the branch, merge or rebase from staging, or click the "Rerun all jobs" button on the review build page in Github Actions.
When a pull request is merged to staging, it will automatically be deployed to the staging environment. You can view the build logs in GitHub Actions. After the build successfully completes, wait a few minutes for the changes to be reflected, and then access the staging app at http://staging.fightpandemics.work.
When a pull request is merged to production, it will automatically be deployed to the production environment. You can view the build logs in GitHub Actions. After the build successfully completes, wait a few minutes for the changes to be reflected, and then access the production app at https://fightpandemics.com.