feat: support "1" for environment variables by deviantintegral · Pull Request #8246 · ddev/ddev

deviantintegral · 2026-03-21T18:40:43Z

The Issue

I keep trying to use DDEV_VERBOSE=1 instead of true as I use it just rarely enough to remember. It's pretty common for variables in shells use 0 / 1, so let's support that here! Feel free to close this if you think it's overcomplicated.

How This PR Solves The Issue

Updates all user-facing environment variables to support 1 as well as true.

AI generated, but human reviewed and tested.

Manual Testing Instructions

Try DDEV_VERBOSE=1 ddev start ...

Automated Testing Overview

Adds an automated test to cover the new behaviour.

github-actions · 2026-03-21T18:47:23Z

Download the artifacts for this pull request:

See Testing a PR.

deviantintegral · 2026-03-22T00:11:44Z

I'm not sure why the quickstart test failed - I don't see any errors in the logs, it just seems to have died.

rfay · 2026-03-22T04:39:18Z

I didn't understand either (on main). So

chore(quickstart): Make quickstart drupal.bats drush launch usage less fragile #8247

But couldn't figure out what happened to make it appear.

rfay · 2026-03-22T15:50:11Z

Rebased

rfay

I didn't study carefully, but it sure looks fine.

What I have really been hoping to do over the years is to add global flags for these.

#7414

deviantintegral · 2026-03-22T20:21:03Z

I agree! Would you like me instead to work on supporting those?

rfay · 2026-03-22T22:25:38Z

I'd loe to have you do those, but finish this first, it's a good thing, and I think env-based control will still be valuable.

rfay

I manually tested and did fine, and the code looks fine. But why isEnvTrue() and IsEnvTrue()?

rfay · 2026-03-28T21:35:59Z

+// has a value accepted by strconv.ParseBool.
+// This duplicates nodeps.IsEnvTrue() because pkg/nodeps imports pkg/output
+// (via wsl.go), creating an import cycle if we import nodeps here.
+func isEnvTrue(envVar string) bool {


Why do we need this duplicate of IsEnvTrue()?

There's the import cycle mentioned in the comment. Is there some other way I could fix this?

rfay · 2026-03-31T18:15:11Z

Hate those import cycles!!! I guess that's the only thing I hate about go!

rfay · 2026-03-31T18:24:36Z

This comment is from Claude Code (AI assistant), at the request of @rfay.

Looking at the import cycle that forced the duplication of isEnvTrue in pkg/output, the root cause is in pkg/nodeps/wsl.go.

pkg/nodeps is meant to be a low-level package with no dependencies on other DDEV packages (the name says so). But IsWSL2MirroredMode() imports pkg/output just to emit a warning on an error path:

output.UserErr.Warnf("Unable to get WSL2 networking mode: %v", err)

This creates the cycle: pkg/nodeps → pkg/output, which then prevents pkg/output from importing pkg/nodeps to use IsEnvTrue.

The fix is to have IsWSL2MirroredMode return the error instead of swallowing it with a warning:

func IsWSL2MirroredMode() (bool, error) {
    if !IsWSL2() {
        return false, nil
    }
    mode, err := GetWSL2NetworkingMode()
    if err != nil {
        return false, err
    }
    return mode == "mirrored", nil
}

The caller (in a higher-level package that is allowed to import pkg/output) handles the warning. That restores proper layering, breaks the cycle, and lets pkg/output use nodeps.IsEnvTrue directly — eliminating the duplicate.

rfay · 2026-03-31T18:26:04Z

I imagine that conflicts with #8262 but it can be fixed there. OK to solve this problem in this PR. It's poor existing usage I think.

Add nodeps.IsEnvTrue() helper that treats both "true" and "1" as truthy, matching common Unix conventions. Applied to DDEV_DEBUG, DDEV_VERBOSE, DDEV_NO_INSTRUMENTATION, and DDEV_NONINTERACTIVE. Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>

Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>

stasadev · 2026-04-09T17:07:40Z

Changes:

Behavior improvements
- DDEV_NO_TUI=1 now correctly disables TUI (old code only responded to empty/non-empty)
- NO_COLOR now follows no-color.org spec: any non-empty value disables color (old code accidentally treated NO_COLOR=0 as "colors enabled")
- DDEV_GOROUTINES changed from != "" to IsEnvTrue - only truthy values enable goroutine reporting
Required fix: import cycle
pkg/nodeps/wsl.go removed output.UserErr.Warnf(...) and its import of pkg/output. This was necessary because pkg/output/output_setup.go now imports pkg/nodeps (for IsEnvTrue), so the reverse import would cause a cycle.
Test cleanup
Manual save/restore of DDEV_DEBUG replaced with t.Setenv("DDEV_DEBUG", "") throughout - cleaner and race-safe.
Correctly left unchanged (external env vars, always set to "true" by infrastructure)
- os.Getenv("GITHUB_ACTIONS") == "true" - set by GitHub Actions
- os.Getenv("CI") == "true" - set by CI systems (GitHub Actions, GitLab, etc.)
- os.Getenv("CODESPACES") == "true" - set by GitHub Codespaces
- os.Getenv("IN_DEVCONTAINER") == "true" - set by VS Code devcontainers

rfay

Yay, thank you!

deviantintegral · 2026-04-09T23:11:07Z

Thank you both for picking this up!

stasadev

Looks good to me.

deviantintegral requested a review from a team as a code owner March 21, 2026 18:40

github-actions Bot added the enhancement label Mar 21, 2026

rfay force-pushed the 20260321_deviantintegral_env_true_or_1 branch from 55a8b1e to 649c283 Compare March 22, 2026 15:50

rfay reviewed Mar 22, 2026

View reviewed changes

stasadev requested changes Mar 23, 2026

View reviewed changes

Comment thread cmd/ddev-hostname/ddev-hostname.go

Comment thread pkg/output/output_setup.go Outdated

rfay requested changes Mar 28, 2026

View reviewed changes

deviantintegral force-pushed the 20260321_deviantintegral_env_true_or_1 branch from 5802773 to 2998ed6 Compare March 31, 2026 18:14

deviantintegral and others added 4 commits April 9, 2026 17:56

style: clarify why isEnvTrue is duplicated in pkg/output

f4331e5

Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>

refactor: use strconv.ParseBool instead of custom conditional

028bbf6

remove extra isEnvTrue

3851b5a

stasadev force-pushed the 20260321_deviantintegral_env_true_or_1 branch from 2998ed6 to 3851b5a Compare April 9, 2026 15:11

stasadev added 3 commits April 9, 2026 19:33

use IsEnvTrue/IsEnvFalse almost everywhere

9a73cac

fix NO_COLOR, spec doesn't use '0'

63b3ee8

Keep CI=true check

25f30a6

rfay approved these changes Apr 9, 2026

View reviewed changes

stasadev approved these changes Apr 10, 2026

View reviewed changes

rfay merged commit f01e6db into ddev:main Apr 10, 2026
40 of 43 checks passed

chenrui333 mentioned this pull request Apr 22, 2026

ddev 1.25.2 chenrui333/homebrew-tap#6275

Closed

Uh oh!

Conversation

deviantintegral commented Mar 21, 2026

The Issue

How This PR Solves The Issue

Manual Testing Instructions

Automated Testing Overview

Uh oh!

github-actions Bot commented Mar 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

deviantintegral commented Mar 22, 2026

Uh oh!

rfay commented Mar 22, 2026

Uh oh!

rfay commented Mar 22, 2026

Uh oh!

rfay left a comment

Choose a reason for hiding this comment

Uh oh!

deviantintegral commented Mar 22, 2026

Uh oh!

rfay commented Mar 22, 2026

Uh oh!

Uh oh!

Uh oh!

rfay left a comment

Choose a reason for hiding this comment

Uh oh!

rfay Mar 28, 2026

Choose a reason for hiding this comment

Uh oh!

deviantintegral Mar 31, 2026

Choose a reason for hiding this comment

Uh oh!

rfay commented Mar 31, 2026

Uh oh!

rfay commented Mar 31, 2026

Uh oh!

rfay commented Mar 31, 2026

Uh oh!

stasadev commented Apr 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rfay left a comment

Choose a reason for hiding this comment

Uh oh!

deviantintegral commented Apr 9, 2026

Uh oh!

stasadev left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

github-actions Bot commented Mar 21, 2026 •

edited

Loading

stasadev commented Apr 9, 2026 •

edited

Loading