Check for 'deps' attr in scala_doc rule #1740

vinnybod · 2025-05-22T17:25:03Z

Description

Add a guard for deps existing in ctx.rules.attr when generating scala_docs.

As a side note: Would the maintainers be interested in a scala_export rule in rules_scala? I implemented this in our rules_jvm_external fork but it won't be upstreamed.

Motivation

It's possible to for ctx.rules.attr to not have deps. Specifically, I ran into a failure generating scala_docs when a maven_project_jar is in the deps of the scala_library.

scala/private/rules/scala_doc.bzl

vinnybod · 2025-05-22T18:03:46Z

test/shell/test_macros.sh

@@ -5,7 +5,7 @@ dir=$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )
 runner=$(get_test_runner "${1:-local}")

 incorrect_macro_user_does_not_build() {
-  (! bazel build //test/macros:incorrect-macro-user) |&
+  (! bazel build //test/macros:incorrect-macro-user) 2>&1 |


This was invalid on my mac GNU bash, version 3.2.57(1)-release (arm64-apple-darwin24)

Heh, yeah, I forgot to mention that. I just fixed that locally, too. Wouldn't hurt to put a #!/usr/bin/env bash at the top, either.

@jayconrod

Adds a `Label()` wrapper to `setup_scala_toolchains()` and `scala_cross_version_select() to fix `@rules_scala_config` invisibilty build breakages under Bzlmod. Enables Bash strict mode for all test scripts via `set -euo pipefail` and fixes the resulting breakages. Removes `use_repo(scala_config, "rules_scala_config")` from `dt_patches/test_dt_patches{,_user_srcjar}/MODULE.bazel. This caused `dt_patches/dt_patch_test.sh` to break before adding the `Label()` wrapper to `setup_scala_toolchains()`. Also updates `dt_patches/dt_patch_test.sh` to use `test/shell/test_runner.sh` and `test/shell/test_helper.sh`. Replaces previous instances of its own `run_test_local` with the result of `get_test_runner`. Updates `run_in_test_repo` argument parsing, updates `run_in_test_repo` calls to use proper string quoting, and ensures any failures exit the script. Other general test updates include: - Adds a new `dt_patches_linux` job to `.bazelci/presubmit.yml` to run `dt_patches/dt_patch_test.sh` as part of the CI build. - Sets `set -euo pipefail` in `test_runner.sh` and `test_helper.sh` and fixes all unbound variables and pipe failures. Updates test helpers in those files to perform better argument quoting, argument pass through, command execution, and output printing. - Updates test helpers in `test_runner.sh` and `test_helper.sh` to show more information when the new `verbose_test_output` helper returns success, even when a test passes. Removes `exit 0` everywhere so successful tests return control to `run_test_local` when `verbose_test_output` is in effect. Adds an exit in `run_test_local` when `RULES_SCALA_TEST_ONLY` is nonempty. - Adds `#!/usr/bin/env bash` to a number of test files. This avoids having some test helper `echo` statements from printing the `-e` argument when running the test file directly. (As opposed to running them via `test_all.sh` or `test_rules_scala.sh`.) - Adds `test/shell/test_macros.sh` to `test_rules_scala.sh`. Updates that fixed previously incorrect behavior of specific tests include: - Fixes `test_scala_import_source_jar.sh` after Bash strict mode fixes revealed underlying breakages. Specifically, `if [[ $1 ]]` in `assert_file_exists()` always returned true, which meant the function always returned true, because the src jar never existed. Also, the expected repo path was never valid under Bzlmod, since it relied on the `WORKSPACE` specific repo path naming convention. Changing the condition to `if [[ -n "$1" ]]` fulfilled the actual intention, and revealed the broken repo path and `fetch_sources` never being true. Ensures `fetch_sources` is `True` for `@com_google_guava_guava_21_0` from `//scala/private/extensions/dev_deps.bzl%dev_deps`. Also makes the directory path to the repo Bzlmod compatible. (Fortunately, the underlying behavior the test intended to validate was already correct.) - Fixes `test_scala_classpath.sh` to actually grep for the `expected` string. Previously is was grepping for the empty string, thanks to the unbound `$method` variable, and always passed. Fixing this revealed that the test would fail when run if the test target artifacts already existed. Updated it to remove the test target artifacts so that the test could run correctly every time. (Again, the underlying behavior was already correct.) Test updates resulting from Bash strict mode that didn't fundamentally change test behavior include: - Fixed the unbound `$target_bes` variable in `test_build_event_protocol.sh` by replacing it with `${bes_file}`. The original intention appeared to be `${target}_bes.txt`, and the test previously just happened to work by using a file named `.txt`. - Refactored `test_invalid_scalacopts.sh` to use `action_should_fail_with{,out}_message`. This resolves the previous pipe failure after turning on Bash strict mode. - Updates `test_macros.sh` to use Bash stderr redirection syntax. Overlaps with a change in bazel-contrib#1740, but shouldn't conflict. - All other test updates defined previously unbound variables and fixed improper argument quotation. --- This contains an easy fix for a basic Bzlmod breakage. In looking to provide a test for it, I went down the path of improving _all_ the shell tests by applying Bash strict mode. This turned into a larger pull request than I intended; as always, I'm happy to split it up if so desired. Vince Rose reported the `setup_scala_toolchains` breakage under Bzlmod in the #scala channel of the Bazel Slack Workspace on 2025-05-21: - https://bazelbuild.slack.com/archives/CDCKJ2KFZ/p1747870770592779 `dt_patches/dt_patch_test.sh` would've caught it had I not added `use_repo(scala_config, "rules_scala_config")` to the `MODULE.bazel` files. Removing these `use_repo` calls reproduced the error, and they passed again once I reintroduced the proper `Label` instances. Rather than add another test, I thought it was time to ensure `dt_patch_test.sh` behaved properly and could run in CI. Also, @jayconrod recently reminded me of Bash strict mode during the course of an internal code review. So I added Bash strict mode to `dt_patch_test.sh` and refactored it to ensure that a test failure properly exited the script. That went well, so I decided to set Bash strict mode in `test_runner.sh` and `test_helper.sh`. Hilarity ensued. Having Bash strict mode in place should greatly reduce the chances of adding tests that pass without actually doing what we think they're doing. Several of the updates show more info when `RULES_SCALA_TEST_ONLY` and `RULES_SCALA_TEST_VERBOSE` are in effect. This helped me immensely while getting to the root of these problems, especially when defining a previously unbound variable and having a test immediately pass. These `RULES_SCALA_TEST_*` variables helped me verify that they were _really_ passing, or accidentally getting by. There's still opportunities to refactor some of the tests to use more of the existing test helpers. I could do more test script cleanup passes in one or more future pull requests.

@jayconrod

Adds a `Label()` wrapper to `setup_scala_toolchains()` and `scala_cross_version_select()` to fix `@rules_scala_config` invisibilty build breakages under Bzlmod. Enables Bash strict mode for all test scripts via `set -euo pipefail` and fixes the resulting breakages. Removes `use_repo(scala_config, "rules_scala_config")` from `dt_patches/test_dt_patches{,_user_srcjar}/MODULE.bazel. This caused `dt_patches/dt_patch_test.sh` to break before adding the `Label()` wrapper to `setup_scala_toolchains()`. Also updates `dt_patches/dt_patch_test.sh` to use `test/shell/test_runner.sh` and `test/shell/test_helper.sh`. Replaces previous instances of its own `run_test_local` with the result of `get_test_runner`. Updates `run_in_test_repo` argument parsing, updates `run_in_test_repo` calls to use proper string quoting, and ensures any failures exit the script. Other general test updates include: - Adds a new `dt_patches_linux` job to `.bazelci/presubmit.yml` to run `dt_patches/dt_patch_test.sh` as part of the CI build. - Sets `set -euo pipefail` in `test_runner.sh` and `test_helper.sh` and fixes all unbound variables and pipe failures. Updates test helpers in those files to perform better argument quoting, argument pass through, command execution, and output printing. - Updates test helpers in `test_runner.sh` and `test_helper.sh` to show more information when the new `verbose_test_output` helper returns success, even when a test passes. Removes `exit 0` everywhere so successful tests return control to `run_test_local` when `verbose_test_output` is in effect. Adds an exit in `run_test_local` when `RULES_SCALA_TEST_ONLY` is nonempty. - Adds `#!/usr/bin/env bash` to a number of test files. This avoids having some test helper `echo` statements from printing the `-e` argument when running the test file directly. (As opposed to running them via `test_all.sh` or `test_rules_scala.sh`.) - Adds `test/shell/test_macros.sh` to `test_rules_scala.sh`. Updates that fixed previously incorrect behavior of specific tests include: - Fixes `test_scala_import_source_jar.sh` after Bash strict mode fixes revealed underlying breakages. Specifically, `if [[ $1 ]]` in `assert_file_exists()` always returned true, which meant the function always returned true, because the src jar never existed. Also, the expected repo path was never valid under Bzlmod, since it relied on the `WORKSPACE` specific repo path naming convention. Changing the condition to `if [[ -n "$1" ]]` fulfilled the actual intention, and revealed the broken repo path and `fetch_sources` never being true. Ensures `fetch_sources` is `True` for `@com_google_guava_guava_21_0` from `//scala/private/extensions/dev_deps.bzl%dev_deps`. Also makes the directory path to the repo Bzlmod compatible. (Fortunately, the underlying behavior the test intended to validate was already correct.) - Fixes `test_scala_classpath.sh` to actually grep for the `expected` string. Previously is was grepping for the empty string, thanks to the unbound `$method` variable, and always passed. Fixing this revealed that the test would fail when run if the test target artifacts already existed. Updated it to remove the test target artifacts so that the test could run correctly every time. (Again, the underlying behavior was already correct.) Test updates resulting from Bash strict mode that didn't fundamentally change test behavior include: - Fixed the unbound `$target_bes` variable in `test_build_event_protocol.sh` by replacing it with `${bes_file}`. The original intention appeared to be `${target}_bes.txt`, and the test previously just happened to work by using a file named `.txt`. - Refactored `test_invalid_scalacopts.sh` to use `action_should_fail_with{,out}_message`. This resolves the previous pipe failure after turning on Bash strict mode. - Updates `test_macros.sh` to use Bash stderr redirection syntax. Overlaps with a change in bazel-contrib#1740, but shouldn't conflict. - All other test updates defined previously unbound variables and fixed improper argument quotation. --- This contains an easy fix for a basic Bzlmod breakage. In looking to provide a test for it, I went down the path of improving _all_ the shell tests by applying Bash strict mode. This turned into a larger pull request than I intended; as always, I'm happy to split it up if so desired. Vince Rose reported the `setup_scala_toolchains` breakage under Bzlmod in the #scala channel of the Bazel Slack Workspace on 2025-05-21: - https://bazelbuild.slack.com/archives/CDCKJ2KFZ/p1747870770592779 `dt_patches/dt_patch_test.sh` would've caught it had I not added `use_repo(scala_config, "rules_scala_config")` to the `MODULE.bazel` files. Removing these `use_repo` calls reproduced the error, and they passed again once I reintroduced the proper `Label` instances. Rather than add another test, I thought it was time to ensure `dt_patch_test.sh` behaved properly and could run in CI. Also, @jayconrod recently reminded me of Bash strict mode during the course of an internal code review. So I added Bash strict mode to `dt_patch_test.sh` and refactored it to ensure that a test failure properly exited the script. That went well, so I decided to set Bash strict mode in `test_runner.sh` and `test_helper.sh`. Hilarity ensued. Having Bash strict mode in place should greatly reduce the chances of adding tests that pass without actually doing what we think they're doing. Several of the updates show more info when `RULES_SCALA_TEST_ONLY` and `RULES_SCALA_TEST_VERBOSE` are in effect. This helped me immensely while getting to the root of these problems, especially when defining a previously unbound variable and having a test immediately pass. These `RULES_SCALA_TEST_*` variables helped me verify that they were _really_ passing, or accidentally getting by. There's still opportunities to refactor some of the tests to use more of the existing test helpers. I could do more test script cleanup passes in one or more future pull requests.

@jayconrod

Adds a `Label()` wrapper to `setup_scala_toolchains()` and `scala_cross_version_select()` to fix `@rules_scala_config` invisibilty build breakages under Bzlmod. Enables Bash strict mode for all test scripts via `set -euo pipefail` and fixes the resulting breakages. Removes `use_repo(scala_config, "rules_scala_config")` from `dt_patches/test_dt_patches{,_user_srcjar}/MODULE.bazel`. This caused `dt_patches/dt_patch_test.sh` to break before adding the `Label()` wrapper to `setup_scala_toolchains()`. Also updates `dt_patches/dt_patch_test.sh` to use `test/shell/test_runner.sh` and `test/shell/test_helper.sh`. Replaces previous instances of its own `run_test_local` with the result of `get_test_runner`. Updates `run_in_test_repo` argument parsing, updates `run_in_test_repo` calls to use proper string quoting, and ensures any failures exit the script. Other general test updates include: - Adds a new `dt_patches_linux` job to `.bazelci/presubmit.yml` to run `dt_patches/dt_patch_test.sh` as part of the CI build. - Sets `set -euo pipefail` in `test_runner.sh` and `test_helper.sh` and fixes all unbound variables and pipe failures. Updates test helpers in those files to perform better argument quoting, argument pass through, command execution, and output printing. - Updates test helpers in `test_runner.sh` and `test_helper.sh` to show more information when the new `verbose_test_output` helper returns success, even when a test passes. Removes `exit 0` everywhere so successful tests return control to `run_test_local` when `verbose_test_output` is in effect. Adds an exit in `run_test_local` when `RULES_SCALA_TEST_ONLY` is nonempty. - Adds `#!/usr/bin/env bash` to a number of test files. This avoids having some test helper `echo` statements from printing the `-e` argument when running the test file directly. (As opposed to running them via `test_all.sh` or `test_rules_scala.sh`.) - Adds `test/shell/test_macros.sh` to `test_rules_scala.sh`. Updates that fixed previously incorrect behavior of specific tests include: - Fixes `test_scala_import_source_jar.sh` after Bash strict mode fixes revealed underlying breakages. Specifically, `if [[ $1 ]]` in `assert_file_exists()` always returned true, which meant the function always returned true, because the src jar never existed. Also, the expected repo path was never valid under Bzlmod, since it relied on the `WORKSPACE` specific repo path naming convention. Changing the condition to `if [[ -n "$1" ]]` fulfilled the actual intention, and revealed the broken repo path and `fetch_sources` never being true. Ensures `fetch_sources` is `True` for `@com_google_guava_guava_21_0` from `//scala/private/extensions/dev_deps.bzl%dev_deps`. Also makes the directory path to the repo Bzlmod compatible. (Fortunately, the underlying behavior the test intended to validate was already correct.) - Fixes `test_scala_classpath.sh` to actually grep for the `expected` string. Previously is was grepping for the empty string, thanks to the unbound `$method` variable, and always passed. Fixing this revealed that the test would fail when run if the test target artifacts already existed. Updated it to remove the test target artifacts so that the test could run correctly every time. (Again, the underlying behavior was already correct.) Test updates resulting from Bash strict mode that didn't fundamentally change test behavior include: - Fixed the unbound `$target_bes` variable in `test_build_event_protocol.sh` by replacing it with `${bes_file}`. The original intention appeared to be `${target}_bes.txt`, and the test previously just happened to work by using a file named `.txt`. - Refactored `test_invalid_scalacopts.sh` to use `action_should_fail_with{,out}_message`. This resolves the previous pipe failure after turning on Bash strict mode. - Updates `test_macros.sh` to use Bash stderr redirection syntax. Overlaps with a change in bazel-contrib#1740, but shouldn't conflict. - All other test updates defined previously unbound variables and fixed improper argument quotation. --- This contains an easy fix for a basic Bzlmod breakage. In looking to provide a test for it, I went down the path of improving _all_ the shell tests by applying Bash strict mode. This turned into a larger pull request than I intended; as always, I'm happy to split it up if so desired. Vince Rose reported the `setup_scala_toolchains` breakage under Bzlmod in the #scala channel of the Bazel Slack Workspace on 2025-05-21: - https://bazelbuild.slack.com/archives/CDCKJ2KFZ/p1747870770592779 `dt_patches/dt_patch_test.sh` would've caught it had I not added `use_repo(scala_config, "rules_scala_config")` to the `MODULE.bazel` files. Removing these `use_repo` calls reproduced the error, and they passed again once I reintroduced the proper `Label` instances. Rather than add another test, I thought it was time to ensure `dt_patch_test.sh` behaved properly and could run in CI. Also, @jayconrod recently reminded me of Bash strict mode during the course of an internal code review. So I added Bash strict mode to `dt_patch_test.sh` and refactored it to ensure that a test failure properly exited the script. That went well, so I decided to set Bash strict mode in `test_runner.sh` and `test_helper.sh`. Hilarity ensued. Having Bash strict mode in place should greatly reduce the chances of adding tests that pass without actually doing what we think they're doing. Several of the updates show more info when `RULES_SCALA_TEST_ONLY` and `RULES_SCALA_TEST_VERBOSE` are in effect. This helped me immensely while getting to the root of these problems, especially when defining a previously unbound variable and having a test immediately pass. These `RULES_SCALA_TEST_*` variables helped me verify that they were _really_ passing, or accidentally getting by. There's still opportunities to refactor some of the tests to use more of the existing test helpers. I could do more test script cleanup passes in one or more future pull requests.

@jayconrod

* Fix setup_scala_toolchains, clean up test scripts Adds a `Label()` wrapper to `setup_scala_toolchains()` and `scala_cross_version_select()` to fix `@rules_scala_config` invisibilty build breakages under Bzlmod. Enables Bash strict mode for all test scripts via `set -euo pipefail` and fixes the resulting breakages. Removes `use_repo(scala_config, "rules_scala_config")` from `dt_patches/test_dt_patches{,_user_srcjar}/MODULE.bazel`. This caused `dt_patches/dt_patch_test.sh` to break before adding the `Label()` wrapper to `setup_scala_toolchains()`. Also updates `dt_patches/dt_patch_test.sh` to use `test/shell/test_runner.sh` and `test/shell/test_helper.sh`. Replaces previous instances of its own `run_test_local` with the result of `get_test_runner`. Updates `run_in_test_repo` argument parsing, updates `run_in_test_repo` calls to use proper string quoting, and ensures any failures exit the script. Other general test updates include: - Adds a new `dt_patches_linux` job to `.bazelci/presubmit.yml` to run `dt_patches/dt_patch_test.sh` as part of the CI build. - Sets `set -euo pipefail` in `test_runner.sh` and `test_helper.sh` and fixes all unbound variables and pipe failures. Updates test helpers in those files to perform better argument quoting, argument pass through, command execution, and output printing. - Updates test helpers in `test_runner.sh` and `test_helper.sh` to show more information when the new `verbose_test_output` helper returns success, even when a test passes. Removes `exit 0` everywhere so successful tests return control to `run_test_local` when `verbose_test_output` is in effect. Adds an exit in `run_test_local` when `RULES_SCALA_TEST_ONLY` is nonempty. - Adds `#!/usr/bin/env bash` to a number of test files. This avoids having some test helper `echo` statements from printing the `-e` argument when running the test file directly. (As opposed to running them via `test_all.sh` or `test_rules_scala.sh`.) - Adds `test/shell/test_macros.sh` to `test_rules_scala.sh`. Updates that fixed previously incorrect behavior of specific tests include: - Fixes `test_scala_import_source_jar.sh` after Bash strict mode fixes revealed underlying breakages. Specifically, `if [[ $1 ]]` in `assert_file_exists()` always returned true, which meant the function always returned true, because the src jar never existed. Also, the expected repo path was never valid under Bzlmod, since it relied on the `WORKSPACE` specific repo path naming convention. Changing the condition to `if [[ -n "$1" ]]` fulfilled the actual intention, and revealed the broken repo path and `fetch_sources` never being true. Ensures `fetch_sources` is `True` for `@com_google_guava_guava_21_0` from `//scala/private/extensions/dev_deps.bzl%dev_deps`. Also makes the directory path to the repo Bzlmod compatible. (Fortunately, the underlying behavior the test intended to validate was already correct.) - Fixes `test_scala_classpath.sh` to actually grep for the `expected` string. Previously is was grepping for the empty string, thanks to the unbound `$method` variable, and always passed. Fixing this revealed that the test would fail when run if the test target artifacts already existed. Updated it to remove the test target artifacts so that the test could run correctly every time. (Again, the underlying behavior was already correct.) Test updates resulting from Bash strict mode that didn't fundamentally change test behavior include: - Fixed the unbound `$target_bes` variable in `test_build_event_protocol.sh` by replacing it with `${bes_file}`. The original intention appeared to be `${target}_bes.txt`, and the test previously just happened to work by using a file named `.txt`. - Refactored `test_invalid_scalacopts.sh` to use `action_should_fail_with{,out}_message`. This resolves the previous pipe failure after turning on Bash strict mode. - Updates `test_macros.sh` to use Bash stderr redirection syntax. Overlaps with a change in #1740, but shouldn't conflict. - All other test updates defined previously unbound variables and fixed improper argument quotation. --- This contains an easy fix for a basic Bzlmod breakage. In looking to provide a test for it, I went down the path of improving _all_ the shell tests by applying Bash strict mode. This turned into a larger pull request than I intended; as always, I'm happy to split it up if so desired. Vince Rose reported the `setup_scala_toolchains` breakage under Bzlmod in the #scala channel of the Bazel Slack Workspace on 2025-05-21: - https://bazelbuild.slack.com/archives/CDCKJ2KFZ/p1747870770592779 `dt_patches/dt_patch_test.sh` would've caught it had I not added `use_repo(scala_config, "rules_scala_config")` to the `MODULE.bazel` files. Removing these `use_repo` calls reproduced the error, and they passed again once I reintroduced the proper `Label` instances. Rather than add another test, I thought it was time to ensure `dt_patch_test.sh` behaved properly and could run in CI. Also, @jayconrod recently reminded me of Bash strict mode during the course of an internal code review. So I added Bash strict mode to `dt_patch_test.sh` and refactored it to ensure that a test failure properly exited the script. That went well, so I decided to set Bash strict mode in `test_runner.sh` and `test_helper.sh`. Hilarity ensued. Having Bash strict mode in place should greatly reduce the chances of adding tests that pass without actually doing what we think they're doing. Several of the updates show more info when `RULES_SCALA_TEST_ONLY` and `RULES_SCALA_TEST_VERBOSE` are in effect. This helped me immensely while getting to the root of these problems, especially when defining a previously unbound variable and having a test immediately pass. These `RULES_SCALA_TEST_*` variables helped me verify that they were _really_ passing, or accidentally getting by. There's still opportunities to refactor some of the tests to use more of the existing test helpers. I could do more test script cleanup passes in one or more future pull requests. * Refactor test/shell/test_persistent_worker.sh `test_persistent_worker_handles_exception_in_macro_invocation` failed on macOS for #1742: - https://buildkite.com/bazel/rules-scala-scala/builds/5620#01970821-a4f1-4333-9fe6-521679bf0a98 Not sure why it failed exactly, but I took the opportunity to refactor it for `verbose_test_output` visibility. Seems to pass fine locally on my macOS machine. Updated `assert_matches` to take an optional `msg` argument to use it in `test_persistent_worker.sh` test cases. There's more opportunity to refactor here, but holding off for now.

mbland · 2025-05-26T16:22:28Z

@vinnybod Looks like there's a small conflict after #1742 landed.

@simuons I think this should go in before the next patch release as well.

…ala-export

vinnybod · 2025-05-27T15:45:20Z

I resolved the merge conflict

Check for 'deps' attr in scala_doc

2abe7cb

vinnybod requested review from liucijus and simuons as code owners May 22, 2025 17:25

mbland reviewed May 22, 2025

View reviewed changes

scala/private/rules/scala_doc.bzl Outdated Show resolved Hide resolved

code review suggestions

7955b8c

vinnybod commented May 22, 2025

View reviewed changes

mbland mentioned this pull request May 25, 2025

Fix setup_scala_toolchains, clean up test scripts #1742

Merged

Merge remote-tracking branch 'origin/master' into vinnybod/fix-for-sc…

989ff80

…ala-export

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

Check for 'deps' attr in scala_doc rule #1740

Check for 'deps' attr in scala_doc rule #1740

Uh oh!

vinnybod commented May 22, 2025

Uh oh!

Uh oh!

vinnybod May 22, 2025

Uh oh!

mbland May 22, 2025 •

edited

Loading

Uh oh!

mbland commented May 26, 2025

Uh oh!

vinnybod commented May 27, 2025

Uh oh!

Uh oh!

Uh oh!

Check for 'deps' attr in scala_doc rule #1740

Are you sure you want to change the base?

Check for 'deps' attr in scala_doc rule #1740

Uh oh!

Conversation

vinnybod commented May 22, 2025

Description

Motivation

Uh oh!

Uh oh!

vinnybod May 22, 2025

Choose a reason for hiding this comment

Uh oh!

mbland May 22, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mbland commented May 26, 2025

Uh oh!

vinnybod commented May 27, 2025

Uh oh!

Uh oh!

mbland May 22, 2025 •

edited

Loading