Right now we have no easy way to install node dependencies before an app starts which is problematic since they are often critical for starting up the server.

I think its probably pretty safe to assume we can just run npm install if a node service has a defined start command

We can take a good deal of work from our internal platform.sh "integration"

I think we ideally want:

1. The sharing or webuser to be available in every container
2. For it to have the host uid/guid if needed (eg on linux)

This will need to have a section for globally installed node deps a la grunt

In recipes in the docs: https://docs.lndo.io/recipes/backdrop.html

  # If ommitted this will use the stable backdrush
  backdrush: stable

It would be nice to be able to put any tag there or a commit hash to get specific versions of the tools.

• LAMP
• LEMP
• LARAVEL 5
• DRUPAL 6
• DRUPAL 7
• DRUPAL 8
• BACKDROP 1
• WORDPRESS 4

might be nice to give users some URLs and proxy things so they can hit up the solr admin in their browser.

might also be nice to add config file locations

• Better handling of webroot assignment for nginx/apache in context of sharing plugin
• Make sure node does not expose webroots if its running in "cli" mode
• Redo D7/D8 examples

Please complete the following steps. The more steps you complete the more
likely we are to address your issue!

1. Please review the current list of issues to make sure you are not submitting a duplicate.
2. Please tag your issue appropriately. Please review the tagging criteria.
3. Based on the issue type tag selected above please fill in the relevant section below.

Bug Report

Tell us about your setup

What is your lando version and operating system?

Example: v3.0.0-alpha1 on Windows 10 Professional

Have you previously installed lando? If so, what version?

Example: Yes, I've installed lando v3.0.0-alpha4

Tell us about your bug

Give a brief description of what happened, including any error messages that appeared.

Example: I tried to create a pantheon site but got a LowerCase not defined exception

Please detail the steps we can take to replicate this bug. Be as detailed as possible.

Example:

1. Started up GUI
2. Clicked on my Pantheon account in the sidebar
3. Selected a site and an environment
4. Kept the default create options and pressed submit

Tell us more

Could better documentation have prevented this issue? if so what should we change?

Does this bug prevent you from using lando?

Feature/Enhancement Request

Please explain your request and its importance in the form of a user story.

Example: As a lando user who works with Drupal sites that have large databases, I want the option to exclude unnecessary database tables from being pulled when I update my site because I want faster "site pull" actions.

Please provide some detail on how lando can help solve this problem

Example: Place an option on the "site pull" screen to exclude cache tables.

These things might not be needed:

• Handle SOLR over HTTPS
• Handle persistent index data

Feature/Enhancement Request

As a Lando user who is developing for desktop and mobile devices, I want to be able to access my Lando URLs from mobile phones on my same network so I can speed my development and QA process.

This user story has come up several times on the Kalabox support channel.

Please provide some detail on how lando can help solve this problem

There are different apps that can solve this problem (see this article), wondering if we can package a built-in tool with Lando, or if we just need to have docs around covering this use case.

Should be able to replicate by going to company/936/edit

ErrorException in 4108468fa409a02f91195fdf326571f0cd622455.php line 60:
Trying to get property of non-object (View: /app/resources/views/company/edit.blade.php)

Guessing this might be because the company doesn't have any revenue/employee eg "numbers" data so an existence check before we do the failing thing is probably all that is required.

There are times when file sharing appears to hang (eg the container can no longer shut down). This seems to happen when transferring in a large amount of files eg on a laravel app that is set up to use one level above the webroot and you need to sync node_modules and vendor.

• Document the user to set up their app so that they are only sharing the webroot and the webroot only contains relevant prod dependencies
• Allow filesharing to select/exclude specific directories
• Try out native osxfs and see where that is at
• Wait for docker to fix its shit

Let's move examples like Drupal and Backdrop (and all others) to use package managers like composer or npm if applicable. This should cut down on repo size. Once we do this we should also remove the relevant repo history to further trim the repo size.

Let's move over our issue and pr templates from github.

see:

• docker/for-win#445
• docker-library/postgres#213

AND MANY OTHERS ON THE GOOGLES

Set up a basic example backdrop app.

• We will also want to have some sort of "passthrough" option a la kbox . <cmd>
• and some sort of ssh option.
• Handle app spin up if needed

We've got a few places where we do logging plus there are docker logs. We should consider the best and easiest way for users to access these logs

1. Some sort of add-on service like pimpmylogs
2. Some sort of docker logs wrapper

For our php and node services we should allow the user to specify global dependencies eg drush, terminus or gulp grunt in their .lando.yml file so that these things get installed in a post build event.

• We also will want to add some common SSH deps to our php services.
• We also will want to add a "build extras" section to all services that allow arbitrary post-build commands to run.
• Add in php 7.1 option