Related to #264.
We should update and fix E2E tests after all changes in 2.0.

Describe the bug
Currently, the scripts inside status-dashboard are using the following commands to work with the StatusDashboardStack:

  "scripts": {
    "deploy": "env-cmd -f ../../../.env.tools cdk deploy *StatusDashboardStack",
    "diff": "env-cmd -f ../../../.env.tools cdk diff *StatusDashboardStack"
  },

However, the .env.tools file is missing, which should contain the environment variables necessary for the deployment.

To Reproduce
Steps to reproduce the behavior:

1. Run nx run status-dashboard:diff
2. See error

> nx run status-dashboard:diff

> [email protected] diff /Users/przemyslawmichalek/PycharmProjects/saas-boilerplate/packages/internal/status-dashboard
> env-cmd -f ../../../.env.tools cdk diff *StatusDashboardStack
Error: Failed to find .env file at path: ../../../.env.tools
    at getEnvFile (/Users/przemyslawmichalek/PycharmProjects/saas-boilerplate/node_modules/.pnpm/[email protected]/node_modules/env-cmd/dist/get-env-vars.js:40:19)
 ELIFECYCLE  Command failed with exit code 1.

Expected behavior
Those env variables should be stored in AWS SSM Parameter Store and retrieved via chamber.

Describe the bug
On fresh db setup with stripe account already having subscription plans, duplicate ones are being created. That's because in run_local.sh script, init_subscriptions command runs without djstripe_sync_models. If djstripe models are synced, the duplicate plans are not created.

To Reproduce
Steps to reproduce the behavior:

1. Make sure your stripe models are not synced
2. Add STRIPE_TEST_SECRET_KEY inside packages/backend/.env which belongs to account with already created subscription plans
3. Simply run make up which triggers backend's run_local.sh script
4. See duplicated plans created in django admin/stripe dashboard

Expected behavior
Stripe's Product and Price models should be synced before running init_subscriptions command inside run_local.sh. Since there might already be plenty of data to sync, there's no point to sync all the data (the default behaviour of djstripe_sync_models command without any args). Running ./manage.py djstripe_sync_models Product Price should be sufficient to avoid creating duplicate subscription plans and to provide a quick backend startup time.

Is your feature request related to a problem? Please describe.
Some of the descriptions on feature pages are not clear enough to users who would like to check the demo.

Describe the solution you'd like
Please go through each description and update it so it's more clear. Describe what each feature does and what functionality for developer if ships.

When mobile sidebar is available the profile icon from the navbar is hidden. In this state in sidebar there is additional Logout option. I think we could add the Your profile option to the sidebar in this case and separate those two additional options with <Separator/>.

Related to #245

Is your feature request related to a problem? Please describe.
Currently docs contains old look of the application.

Describe the solution you'd like
We need to take a new screenshots and update them in docs

Is your feature request related to a problem? Please describe.

Current docs theme is a standard Docusaurus theme which is ok for now but we could have something much better looking.

Describe the solution you'd like

Update the styles so the docs pages look more attractive to the user.

Additional context

Example of the very nice looking docs: https://react.dev/reference/react, https://docs.pmnd.rs/react-three-fiber/getting-started/introduction

Describe the bug
HMR throws an error in console on component update.

To Reproduce
Steps to reproduce the behavior:

1. Run nx start:app
2. Open application in browser
3. Change component code
4. See error in console

Expected behavior
The component should be update in browser without page refresh.

Additional context
It might be connected with some dependencies changes. On master branch it's working, on radix branch it's broken.

Describe the bug
In services/backend/scripts/build.sh the docker pulling statement is before docker login

docker pull "$BACKEND_REPO_URI:latest" || true

# ...

if DOCKER_LOGIN=$(aws ecr get-login-password --region "$AWS_DEFAULT_REGION" 2> /dev/null); then
 echo "$DOCKER_LOGIN" | docker login --username AWS --password-stdin "$AWS_ACCOUNT_ID".dkr.ecr."$AWS_DEFAULT_REGION".amazonaws.com
else
 echo "get-login-password not supported by the AWS CLI, trying get-login instead..."
 # shellcheck disable=SC2091
 $(aws ecr get-login --no-include-email --region "$AWS_DEFAULT_REGION")
fi

This means it will always fail

Expected behavior
Backend build scripts pulls latest image to reuse existing layers during build.

Is your feature request related to a problem? Please describe.
Currently theme state is set based on the system preference on each page load. If user change it to the opposite it will remain until another refresh.

Describe the solution you'd like
State should be always saved in the LocalStorage after user changed it manually.

Additional context
Our current UI is quite simple but as an extension we might think about more advanced solution with the three possible states: light, dark and system which will be always based on the system state. It might be nice if someone have system theme changing during the day for example and would like the application to follow this behavior.

The next-themes package is very good example of this. Actually would be good to use this package instead of custom code but it has unnecessary next dependency.

Related to #174

Steps to reproduce

1. Run local environment for the first time:

   make shell
   docker-compose up -d

2. Restart backend container:

   make shell
   docker-compose restart backend

   Or by:

   make shell
   docker-compose stop backend
   docker-compose up -d --build backend

Encountered behavior

$ make shell
$ docker-compose logs backend
backend_1      | wait-for-it.sh: waiting 15 seconds for db:5432
backend_1      | wait-for-it.sh: db:5432 is available after 0 seconds
backend_1      | Install LocalStack fixture
backend_1      |
backend_1      | An error occurred (ResourceAlreadyExistsException) when calling the CreateEventBus operation: Event bus imx-local-workers already exists.

Expected behavior

backend container starts successfully.

Workaround fix

1. Rebuilt all the containers by forcing recreate:

   make shell
   docker-compose up -d --build --force-recreate

2. Disable event bus creation in services/backend/scripts/run_local.sh before the backend container restart:

   echo "Install LocalStack fixture"
   
   wait_for_s3
   # create_event_bus
   verify_email_in_ses

Some inspirations:
https://flowbite.com/docs/components/sidebar/
https://ui.shadcn.com/examples/forms

Describe the bug

pnpm lockfile of radix-ui branch not up to date with package.json

pnpm install --include-workspace-root --frozen-lockfile --filter=docs... --filter=backend...
  shell: /usr/bin/bash --noprofile --norc -e -o pipefail {0}
  env:
    PNPM_HOME: /home/runner/setup-pnpm/node_modules/.bin
Scope: 15 of 2[3](https://github.com/apptension/saas-boilerplate/actions/runs/5174312422/jobs/9320486808?pr=183#step:4:3) workspace projects
Lockfile is up to date, resolution step is skipped
 ERR_PNPM_OUTDATED_LOCKFILE  Cannot install with "frozen-lockfile" because pnpm-lock.yaml is not up to date with packages/webapp-libs/webapp-core/package.json

Note that in CI environments this setting is true by default. If you still need to run install in such cases, use "pnpm install --no-frozen-lockfile"
Error: Process completed with exit code 1.

To Reproduce
Steps to reproduce the behavior:

1. Checkout to branch feature/radix-ui
2. Use "pnpm install" in root folder

Is there interest in building out the web app?

Is your feature request related to a problem? Please describe.
There's a newer version of dj-stripe which has support for more payment methods.

There's also a newer version of stripe-mock docker image hosted in a different organisation. We should migrate.
I'm putting this information in the same ticket as probably new dj-stripe would need new API mock.

Describe the solution you'd like
A clear and concise description of what you want to happen.

Additional context
I tried updating myself but seems that a number of tests fail and it's not a quick drop-in change.

@all-contributors
please add @pziemkowski for being an original author.
please add @mkleszcz for being an original author.
please add @kailip for all contributions to backend code.
please add @kamilkedzierski for helping migrate from Relay to Apollo.
please add @sdrejkarz for helping migrate from Relay to Apollo.
please add @mikeqcp for implementing the first version of the web app service.
please add @Wojciech-Florczak for many bug fixes in the web app.

Is your feature request related to a problem? Please describe.
We do not validate automatically each pull request title semantics and sometimes it turns out that there are commits in master history that are not valid in terms of the message.

Describe the solution you'd like
We could add and configure GH action to automatically check PR titles: https://github.com/marketplace/actions/semantic-pull-request

Please move Whaat is SaaS Boilerplate up, right after Getting started. It should be second option

Popover is quite big and options are aligned in a wrong way

Is your feature request related to a problem? Please describe.
Currently running storybooks are very very slow on initial startup as well as on component update

Describe the solution you'd like
First thing is to update storybook package to latest version. If this won't help then we need to dig deeper why this is so slow.

Steps to reproduce

zsh
make shell
cd /services/workers
make test

Encountered behavior

imx_501_db_1 is up-to-date
bash ./scripts/run_tests.sh
All done! ✨ 🍰 ✨
25 files would be left unchanged.
./scripts/wait-for-it.sh: line 146: realpath: command not found
./scripts/wait-for-it.sh: line 58: timeout: command not found
wait-for-it.sh: timeout occurred after waiting 15 seconds for localhost:5432
make[1]: *** [test] Error 127

Expected behavior

make test run without error

Workaround fix

• Change shell to bash

make shell USER_SHELL=bash

or:

• Install coreutils before running tests

brew install coreutils

Please change old logo in the docs to the new one.

Describe the bug
On fresh SB installation when there is no Stripe configured the user register / login won't work because user profile is not created.

To Reproduce
Steps to reproduce the behavior:

1. Create new SB instance locally without Stripe env configuration1
2. Try to register and login

Expected behavior
The user should be able to register and login even if the app doesn't have Stripe integration enabled

Steps:

1. Launch the Saas app
2. Go to the "Payment demo" page
3. Choose the amount $10
4. Choose the amount $5 - the confirmation button changes the width.

Describe the bug
ssh-bastion image is common to all deployed environments. Since there's only one ECR repository (defined in global stack), and tagging docker image with ssh-bastion tag is part of building backend in Pipeline, this image always contains latest deployed code. It can cause problems, when entering with ssh-bastion to environment that wasn't deployed lately, and there were new migrations in meantime deployed to another environment. Trying to enter shell_plus results in exception caused by mismatch between the code and database (like new fields existing in models but not in the database).

Expected behavior
Have one ssh-bastion image per environment. Possible solutions may be having separate ECR repositories for each environment instead of one defined in global stack, or using different tag when tagging ssh-bastion image for each environment.

Is your feature request related to a problem? Please describe.
Ruff is already used inside the backend package so it should also be used inside workers.

Describe the solution you'd like
Replace flake8 with ruff inside workers package.

Additional context
Ruff github page

Describe the bug
Left menu in docs hides under header after scroll. Please check if there is any configuration that will prevent this.

To Reproduce
Steps to reproduce the behavior:

1. Go to https://docs.demo.saas.apptoku.com/
2. Scroll down
3. Spot the issue

Expected behavior
When all items in menu fit on screen, the menu should be always fully visible.

Screenshots

Describe the bug
I tried to setup a fresh clone with pnpm version 8.2.0 and pnpm install command failed with the error log pasted below. README.md states that pnpm version 8 or above should be sufficient, but it doesn't work with 8.2.0 . It worked for me with the latest version 8.6.1 .

 ~/PycharmProjects/saas-boilerplate git:(master) pnpm install
Scope: all 23 workspace projects
 WARN  Your pnpm-lock.yaml was generated by a newer version of pnpm. It is a compatible version but it might get downgraded to version 6.0
 ERROR  /[email protected] is an invalid relative dependency path

pnpm: /[email protected] is an invalid relative dependency path
    at Object.parse2 [as parse] (/usr/local/Cellar/pnpm/8.2.0/libexec/dist/pnpm.cjs:113619:15)
    at nameVerFromPkgSnapshot (/usr/local/Cellar/pnpm/8.2.0/libexec/dist/pnpm.cjs:113728:28)
    at addPreferredVersionsFromLockfile (/usr/local/Cellar/pnpm/8.2.0/libexec/dist/pnpm.cjs:186437:89)
    at getPreferredVersionsFromLockfileAndManifests (/usr/local/Cellar/pnpm/8.2.0/libexec/dist/pnpm.cjs:186431:7)
    at _installInContext (/usr/local/Cellar/pnpm/8.2.0/libexec/dist/pnpm.cjs:187332:142)
    at async installInContext (/usr/local/Cellar/pnpm/8.2.0/libexec/dist/pnpm.cjs:187663:16)
    at async _install (/usr/local/Cellar/pnpm/8.2.0/libexec/dist/pnpm.cjs:187202:25)
    at async mutateModules (/usr/local/Cellar/pnpm/8.2.0/libexec/dist/pnpm.cjs:186936:23)
    at async recursive (/usr/local/Cellar/pnpm/8.2.0/libexec/dist/pnpm.cjs:188424:29)
    at async installDeps (/usr/local/Cellar/pnpm/8.2.0/libexec/dist/pnpm.cjs:188720:11)

To Reproduce
Steps to reproduce the behavior:

1. Use pnpm version 8.2.0
2. Use a fresh clone of the repo
3. Run pnpm install
4. See error

Expected behavior
README.md should be more specific on pnpm version.

Describe the bug
Fresh deployment of WebAppStack fails because of the new S3 default policy.

To Reproduce
Steps to reproduce the behavior:

1. Install all global infra and push the code so new env should be deployed.
2. There will ba an error in WebAppStack deployment: API: s3:PutBucketPolicy Access Denied

Expected behavior
Initial stack deployment should be successful.

The user has to click the browser refresh button if he wants to see the newly added Contentful items in this list

Steps:

1. Launch the Saas
2. Go to the "Demo Contentful items" page
3. Go to ie. "Payment demo" page
4. Add a new item in using Contentful
5. Go back to the "Demo Contentful items" page - The user can't see newly added item

Is your feature request related to a problem? Please describe.

On top of the CodePipeline screen a message box with following copy is shown:

This pipeline has a source (saas-qa-source) that is configured for polling. Migrate (update) your pipeline to the recommended event-based mechanism for change detection. For details, see the migration guide.

This wasn't available at the time when we worked on this integration.

Describe the solution you'd like

Reaserch and modify CDK code to use event-based mechanism instead of polling.

Migrate polling pipelines (Amazon S3 source) (AWS CloudFormation template)

Describe the bug
The user does not receive a reset password email.

To Reproduce
Steps to reproduce the behavior:

1. Go to the log in page
2. Click on the "Forgot password" button
3. Povide previously registered mail address
4. Click on the "Send the link" button
5. Check the inbox of this mail address - no mail comes

Expected behavior
A password reset mail should be sent to the e-mail address provided earlier.

Screenshots

Please update logo in SB to the new one:

Steps to reproduce

1. Run Bitbucket Pipeline test async workers.

Encountered behavior

There is unhandled node: not found error during steps install-test and test-bb-ci. The steps continues and finishes successfully.

+ make -C services/workers install-test PROJECT_NAME=test ENV_STAGE=test
make: Entering directory '/opt/atlassian/pipelines/agent/build/services/workers'
pipenv install --dev
/bin/sh: 1: node: not found
Creating a virtualenv for this project…
Pipfile: /opt/atlassian/pipelines/agent/build/services/workers/Pipfile
Using /usr/local/bin/python3.8 (3.8.5) to create virtualenv…

+ make -C services/workers test-bb-ci PROJECT_NAME=test ENV_STAGE=test
make: Entering directory '/opt/atlassian/pipelines/agent/build/services/workers'
bash ./scripts/run_tests.sh
/bin/sh: 1: node: not found
All done! ✨ 🍰 ✨
26 files would be left unchanged.
wait-for-it.sh: waiting 15 seconds for localhost:5432
wait-for-it.sh: localhost:5432 is available after 0 seconds
running test

There is no such behavior in GitHub Actions defined for the project.

Expected behavior

Both steps (and the whole Pipeline in general) runs without any, especially unhandled, errors.

Unhandled errors that does not have actual influence on run of the Pipeline, unnecessary confuses the users.

Cause of the issue

Bitbucket Pipeline for test async workers uses image python:3.8, which does not have Node.js installed by default, while running the tests directly in this image:

$ docker run -it python:3.8 bash
root@420d6268cc7b:/# node
bash: node: command not found

GitHub Actions uses ubuntu-latest with all the dependencies pre-configured.

Also, the place where node is used is probably the base.mk:

define GetFromCfg
$(shell node -p "require('$(CONFIG_FILE)').$(1)")
endef

That would explain why the error does not influence the run of the Pipeline.

Possible solutions

Use image with all the dependencies installed

Similar to GitHub Actions.

For example, the python:3.8 with node.js installed, e.g. https://hub.docker.com/r/dstasczak/pythonwithnodejs.

Run tests of worker in docker

Similar to test backend.

It would probably require a lot of work (definition of docker-compose, Dockerfile, and adjustments in Makefile).

However, it could allow to also make this step more consistent between GH Actions, BB Pipelines and local environment.

Remove the reason of error

If node is not actually required during the test async workers then there should not be any error if it is not installed there.

Describe the solution you'd like
Please rename items in the menu to more descriptive names:

- Dashboard
- Payments
- Subscriptions
- OpenAI integration
- Content items
- CRUD
- Documents
- Static pages (subheader)
- Privacy policy
- Terms and conditions

Is your feature request related to a problem? Please describe.
Current frontend app UI code is not based on any well known library or design system. It's not very easy to remove currently used stylings and component primitives could be done better.

Describe the solution you'd like
We would like to use:

• tailwind for stylings and themes
• Radix for primitives
• shadcn/ui for initial component design of the app

Describe alternatives you've considered
We have researched other solutions like chakra or MUI but the proposed solution looks most elastic

Screens to cover

• Login screen
• Register screen
• Forgot password screen
• OTP screens
• Dashboard
• Profile form screen
• Demo items screen
• Subscription screens
• Payment screens
• Documents screen
• SaaS ideas screen
• Admin screen
• Privacy policy and T&C screens
• #243
• #230
• #244
• #246
• #245

Is your feature request related to a problem? Please describe.
After migrate styles to Radix and Tailwind there will be unused code with the old theme based on styled-components. We should decide either remove it entirely or clean it and prepare for future usage in case of if someone in the future would like to use styled-components instead of tailwind styles.

Additional context
Related to #174

Describe the solution you'd like
Add suggestion to the production guide to adjust the RDS DB instance size according to the actual needs and also suggest enabling Multi-AZ deployment for automatic failover to a standby instance.

Additional context
Amazon RDS Multi-AZ

Is your feature request related to a problem? Please describe.
CodePipeline PostDeploy step fails with:

#4 ERROR: failed to copy: httpReadSeeker: failed open: unexpected status code https://registry-1.docker.io/v2/cypress/included/manifests/sha256:86e3bff9634b85a5239a65974ac3c98f51277a8e1519fbd57b5e528f55ab8bf7: 429 Too Many Requests - Server message: toomanyrequests: You have reached your pull rate limit. You may increase the limit by authenticating and upgrading: https://www.docker.com/increase-rate-limit

Describe the solution you'd like
With no rate limit on ECR public it'd be possible to pull Docker official images as many times preferred. It should be combined with ECR Pull through cache mechanism.

Additional context
AWS docs on using pull through cache rules

Describe the bug
Running make up on a fresh repository throws an error because packages/e2e-tests/.env is not created automatically.

To Reproduce

1. run make up in repo root
2. See error

Expected behavior
docker-compose does not return any errors.

Additional context
Shell log

>  NX   Successfully ran target docker-login for project core
Couldn't find env file: /Users/<redacted>/Documents/repositories/saas-boilerplate/packages/e2e-tests/.env
 >  NX   ERROR: Something went wrong in run-commands - Command failed: docker-compose build workers
   Pass --verbose to see the stacktrace.
> nx run backend:compose-build-image
> nx run core:docker-login
 
 >  NX   Successfully ran target docker-login for project core
Couldn't find env file: /Users/<redacted>/Documents/repositories/saas-boilerplate/packages/e2e-tests/.env
 >  NX   ERROR: Something went wrong in run-commands - Command failed: docker-compose build backend
   Pass --verbose to see the stacktrace.
 
 >  NX   Running target compose-build-image for 2 projects and 2 tasks they depend on failed
   Failed tasks:
   
   - workers:compose-build-image
   - backend:compose-build-image
 >  NX   ERROR: Something went wrong in run-commands - Command failed: nx run-many --target=compose-build-image --projects=backend,workers
   Pass --verbose to see the stacktrace.

 ——————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————

 >  NX   Ran target docker-compose:up for project core and 2 task(s) they depend on (2s)
 
    ✖    1/3 failed
    ✔    2/3 succeeded [1 read from cache]

Describe the bug

I'm not 100% sure why but whenever you open a file that created TypedDocumentNode in one of webapp-libs you get TS2742 error:

TS2742: The inferred type of 'stripeCreatePaymentIntentMutation' cannot be named without a reference to 'packages/webapp-libs/webapp-api-client/node_modules/@graphql-typed-document-node/core/dist'. This is likely not portable. A type annotation is necessary.

To Reproduce
Steps to reproduce the behavior:

1. Go to any .graphql.ts file in Jetbrains IDE
2. See the error

Expected behavior
No error present.

Screenshots

Is your feature request related to a problem? Please describe.
Nx.js v16 has been released a month ago and we're still using v15.

Describe the solution you'd like
Upgrade the library to v16.

Describe the bug
Attempted to use the Starter kit CLI command "npm init saas-boilerplate" to start a new clone of the project, but it failed with the error message from my terminal log:

npm init saas-boilerplate
Performing system check:
√ Checking if docker is installed... it is!
√ Checking if pnpm is installed... it is!
√ Checking if nx is installed... it is!
Project will be initialized in directory: C:\saas-demo
Fetching latest release... done
Latest release: 2.0.0-alpha.1
Start cloning repository... !
Error: Cloning into 'C:\saas-demo'...
Host key verification failed.
fatal: Could not read from remote repository.

I believe this is because Github changed their public SSH key - see here
https://github.blog/2023-03-23-we-updated-our-rsa-ssh-host-key/

To Reproduce
run the start kit CLI command "npm init saas-boilerplate"

Expected behavior
we get a configured and working clone of the repo

Is your feature request related to a problem? Please describe.
E2E tests are outdated after all changes in the 2.0. We should update them but for now until it's not done they should be removed from the codebase.

Describe the solution you'd like
Just remove the code of E2E tests and we'll revert changes in the next release when they should be updated.

Steps:

1. Launch the Saas
2. Go to the CRUD example items page
3. Click on the "Add new item" button
4. Fill in the field
5. Click on the "Save changes" button - the item was saved successfully but there is displayed an item details page

Expected: User should be moved to the item list page

Is your feature request related to a problem? Please describe.
There's a new, 7.0 version, of CodeBuild image and we're currently using 6.0

Describe the solution you'd like
Change all build image of all CodeBuild projects to v7

Is your feature request related to a problem? Please describe.
Currently there are some random, Lorem ipsum items (one even with the car image...) which are not looking good.

Describe the solution you'd like
Please add some appropriate content to the items. Use images that we know that we can use.

After deleting an item there should be displayed an info popup as when adding or removing items.