Modify global analyzer config precedence: #45871

chsienki · 2020-07-10T17:19:09Z

Pre-parse the global options into their own collection
Don't put global diagnostics on the tree level options
Add global diagnostics onto the tree options provider
Pass through the global diagnostics during compilation
Update the analyzer supression code
Add tests

Precedence level is now:

Warning level
Syntax tree level
Compilation level
Global analyzer level
Global warning level

chsienki · 2020-07-10T17:27:25Z

src/Compilers/Core/Portable/Compilation/SyntaxTreeOptionsProvider.cs

+        /// <summary>
+        /// Get diagnostic severity set globally for a given diagnostic identifier
+        /// </summary>
+        public abstract bool TryGetGlobalDiagnosticValue(string diagnosticId, out ReportDiagnostic severity);
    }

    internal sealed class CompilerSyntaxTreeOptionsProvider : SyntaxTreeOptionsProvider


❓ This type is a new public api. Should we rename it to better reflect that it's not really about TreeOptions?

Perhaps DiagnosticSeverityProvider? Although its based on top of analyzer options it really only returns ReportDiagnostics #Closed

I'm not sure if there is time to rename APIs. Isn't the API locked with .NET 5 RC1 shipping?

In reply to: 452976584 [](ancestors = 452976584)

This better not be locked given we just had our review of APIs today and we have other things to change here too. 😄

@chsienki Just to make sure we understood, this needs to be a different API than TryGetGlobalDiagnosticValue because other options have to be slotted in the middle in terms of the precedence? And that one is just the compilation level diagnostic settings?

TryGetGlobalDiagnosticValue has to exist as the user has to have a way of asking for diagnostics that still apply when not in the context of a syntax tree.

My comment is that I think we should rename CompilerSyntaxTreeOptionsProvider as we only use it to get diagnostics. Though i'm now wondering if keeping as TreeOptions is better as as it allows us to add things onto it in the future

In reply to: 473402105 [](ancestors = 473402105)

These are not locked yet, we can still change them if needed.

mavasani · 2020-07-10T17:33:30Z

src/Compilers/CSharp/Portable/Compilation/CSharpDiagnosticFilter.cs

                        }
                        break;
                }
            }

+            // 5. Global analyzer options


I believe this should go immediately after // 3. Compilation level check. Any rule ID specific configuration should still trump over generic diagnostic option handled in 4. Global options and be done before pragma related checks. Basically, if global config options suppress a specific diagnostic ID, then it would need a rule ID specific configuration from user to override it. #Resolved

Ah ok, that makes sense I think. I wasn't sure how we wanted to handle the precedence here.

So, if a NuGet package ups a specific diagnostic to a warning, but the user specifies the global -warn:0 then we would still expect to warn about that particular diagnostic?
#Resolved

Ah right yes thinking about it, the user would say -warn:0, but then still have to explicity add a -nowarn for the one the global config upped. That makes sense I think. #Resolved

Yep, any diagnostic ID specific setting should override diagnostic ID agnostic setting. That has been the common design for diagnostic ID configuration from all sources. #Resolved

mavasani · 2020-07-10T17:46:22Z

src/Compilers/Core/Portable/DiagnosticAnalyzer/AnalyzerManager.cs

@@ -335,6 +335,12 @@ internal bool IsSupportedDiagnostic(DiagnosticAnalyzer analyzer, Diagnostic diag
                    isSuppressed = true;
                }

+                // Global editorconfig settings only apply if it wasn't supressed via commandline already
+                if (!isSuppressed && options.SyntaxTreeOptionsProvider is object && options.SyntaxTreeOptionsProvider.TryGetGlobalDiagnosticValue(diag.Id, out severity))


I think it is better to instead move this check upfront based on precedence order:

// Is this diagnostic suppressed by default (as written by the rule author) var isSuppressed = !diag.IsEnabledByDefault; // Global editorconfig settings overrides the analyzer author // Compilation wide user settings (diagnosticOptions) from ruleset/nowarn/warnaserror overrides the analyzer author and global editorconfig settings. if (diagnosticOptions.TryGetValue(diag.Id, out var severity) || options.SyntaxTreeOptionsProvider is object && options.SyntaxTreeOptionsProvider.TryGetGlobalDiagnosticValue(diag.Id, out severity)) { isSuppressed = severity == ReportDiagnostic.Suppress; } else { severity = isSuppressed ? ReportDiagnostic.Suppress : DiagnosticDescriptor.MapSeverityToReport(diag.DefaultSeverity); } ``` #Resolved

mavasani · 2020-07-10T17:48:21Z

src/Compilers/VisualBasic/Portable/Compilation/VisualBasicDiagnosticFilter.vb

@@ -176,6 +176,11 @@ Namespace Microsoft.CodeAnalysis.VisualBasic
                End If
            End If

+            Dim globalReport As ReportDiagnostic
+            If (Not isSpecified) AndAlso syntaxTreeOptions IsNot Nothing AndAlso syntaxTreeOptions.TryGetGlobalDiagnosticValue(id, globalReport) Then


Same suggestion here. Rule ID specific global config options should override Rule ID agnostic options. #Resolved

chsienki · 2020-08-11T18:01:59Z

@dotnet/roslyn-compiler for review please.

//cc @mavasani @jasonmalinowski

jcouv · 2020-08-11T22:51:48Z

~~Looking~~
Sorry for the delay (got distracted), will get back to this PR in the morning. #Closed

jaredpar · 2020-08-12T13:40:30Z

Could we add a quick summary of what the precedence is now to the description of the PR? #Resolved

chsienki · 2020-08-17T17:03:55Z

Ping @dotnet/roslyn-compiler for reviews please :)

jcouv · 2020-08-17T18:27:52Z

src/Compilers/Core/Portable/CommandLine/AnalyzerConfigSet.cs

@@ -145,8 +145,35 @@ private AnalyzerConfigSet(ImmutableArray<AnalyzerConfig> analyzerConfigs, Global
            Debug.Assert(allMatchers.Count == _analyzerConfigs.Length);

            _analyzerMatchers = allMatchers.ToImmutableAndFree();
+
+            // parse the global options upfront


what's the motivation to move this parsing of global options up? (why not keep it in GetOptionsForSourcePath?) #Closed

Because we're returning them both as a property and inline in GetOptionsForSourcePath we should probably just make it lazy though, rather than doing it unconditionally.

In reply to: 471683120 [](ancestors = 471683120)

jcouv · 2020-08-17T18:50:11Z

src/Compilers/Core/Portable/CommandLine/CommonCompiler.cs

@@ -87,7 +87,8 @@ internal abstract partial class CommonCompiler
            TextWriter consoleOutput,
            TouchedFileLogger touchedFilesLogger,
            ErrorLogger errorLoggerOpt,
-            ImmutableArray<AnalyzerConfigOptionsResult> analyzerConfigOptions);
+            ImmutableArray<AnalyzerConfigOptionsResult> analyzerConfigOptions,
+            AnalyzerConfigOptionsResult globalConfigOptions);


nit: since global options now take precedence, consider moving them up in the parameter list (before analyzerConfigOptions). #Closed

Global options are lower precedence than analyzer options.

In reply to: 471708791 [](ancestors = 471708791)

jcouv · 2020-08-17T18:52:52Z

src/Compilers/Core/Portable/Compilation/SyntaxTreeOptionsProvider.cs

+        /// <summary>
+        /// Get diagnostic severity set globally for a given diagnostic identifier
+        /// </summary>
+        public abstract bool TryGetGlobalDiagnosticValue(string diagnosticId, out ReportDiagnostic severity);


TryGetGlobalDiagnosticValue [](start = 29, length = 27)

Sorry if this was already discussed with feature crew offline, but what consumer of our public API needs this new API?
I would have thought the existing TryGetDiagnosticValue would be sufficient. #Closed

Because global options can exist 'ambiently' without applying to any specific tree. One option would be to call TryGetDiagnosticValue with a null tree but that seems icky.

In reply to: 471710188 [](ancestors = 471710188)

jcouv · 2020-08-17T19:47:40Z

    public void GlobalConfigInvalidSeverity()

Do we have a test for GlobalConfigOptions when there is no global config? #Resolved

Refers to: src/Compilers/Core/CodeAnalysisTest/Analyzers/AnalyzerConfigTests.cs:1920 in 89ba3bf. [](commit_id = 89ba3bf2955022f11a1d0ab2b1c3b8b96df22f8e, deletion_comment = False)

jcouv · 2020-08-17T19:49:30Z

src/Compilers/VisualBasic/Test/CommandLine/CommandLineTests.vb

@@ -10020,6 +10020,58 @@ End Class")
            End If
        End Sub

+        <Fact>
+        <WorkItem(44087, "https://github.com/dotnet/roslyn/issues/44804")>
+        Public Sub GlobalAnalyzerConfigDiagnosticOptionsCanBeOverridenByCommandLine()


GlobalAnalyzerConfigDiagnosticOptionsCanBeOverridenByCommandLine [](start = 19, length = 64)

nit: thanks for naming tests to make intent clear. Makes the review easier :-)

jcouv

LGTM Thanks (iteration 10)

chsienki · 2020-08-17T20:13:42Z

    public void GlobalConfigInvalidSeverity()

Added

In reply to: 675076806 [](ancestors = 675076806)

Refers to: src/Compilers/Core/CodeAnalysisTest/Analyzers/AnalyzerConfigTests.cs:1920 in 89ba3bf. [](commit_id = 89ba3bf2955022f11a1d0ab2b1c3b8b96df22f8e, deletion_comment = False)

chsienki · 2020-08-19T19:14:39Z

Ping @dotnet/roslyn-compiler for a second review please :)

jasonmalinowski · 2020-08-19T22:29:30Z

src/Compilers/CSharp/Portable/Compilation/CSharpDiagnosticFilter.cs

@@ -119,7 +119,8 @@ internal static class CSharpDiagnosticFilter
        ///     1. Warning level
        ///     2. Syntax tree level
        ///     3. Compilation level
-        ///     4. Global warning level
+        ///     4. Global analyzer config
+        ///     5. Global warning level


How is 'global warning level' different than 'warning level'? I'm trying to understand this list and keep coming up confused.

We could probably make the wording clearer. Warning level is the /warn:3 that enables/disables big swathes of warnings.

Global warning level is /warnaserror:+/- that bumps all warnings up to errors

In reply to: 473388737 [](ancestors = 473388737)

jasonmalinowski · 2020-08-19T22:38:08Z

src/Workspaces/CoreTest/SolutionTests/SolutionTests.cs

+        [CombinatorialData]
+        public void TestAddingAndRemovingGlobalEditorConfigFileWithDiagnosticSeverity([CombinatorialValues(LanguageNames.CSharp, LanguageNames.VisualBasic)] string languageName, bool useRecoverableTrees)
+        {
+            using var workspace = useRecoverableTrees ? CreateWorkspaceWithRecoverableSyntaxTrees() : CreateWorkspace();


You can remove the 'useRecoverableTrees' and make this just a regular [Fact]. The distinction mattered when diagnostic values were held on the trees since we could have bugs where we didn't handle the trees correctly in both cases. It's now moot. I'm guessing this test predates merging the other PR so you were following the correct pattern, at the time. #Resolved

jasonmalinowski · 2020-08-19T22:41:04Z

src/Test/Utilities/Portable/Compilation/TestSyntaxTreeOptionsProvider.cs

@@ -26,17 +27,27 @@ public sealed class TestSyntaxTreeOptionsProvider : SyntaxTreeOptionsProvider
                    x => x.Item2,
                    comparer)
            );
+            if (globalOption.Item1 is object)


Can we instead make globalOption a (string, ReportDiagnostic) to make the null handling clearer? This file is #nullable enabled so presumably you shouldn't be able to get a null first parameter; the fact you can pass default in line 50 is a compiler bug: #38957 #Resolved

jasonmalinowski · 2020-08-19T22:47:37Z

src/Compilers/Core/Portable/Compilation/SyntaxTreeOptionsProvider.cs

+        /// <summary>
+        /// Get diagnostic severity set globally for a given diagnostic identifier
+        /// </summary>
+        public abstract bool TryGetGlobalDiagnosticValue(string diagnosticId, out ReportDiagnostic severity);
    }

    internal sealed class CompilerSyntaxTreeOptionsProvider : SyntaxTreeOptionsProvider


This better not be locked given we just had our review of APIs today and we have other things to change here too. 😄

@chsienki Just to make sure we understood, this needs to be a different API than TryGetGlobalDiagnosticValue because other options have to be slotted in the middle in terms of the precedence? And that one is just the compilation level diagnostic settings?

jaredpar · 2020-08-24T01:02:25Z

Do we have a document describing the precedence levels? If not we should create one and if we do we should update it with this PR.

Assuming that "Syntax Tree Level" includes non-global analyzer configurations correct? #Resolved

jaredpar · 2020-08-24T01:07:32Z

src/Compilers/Core/Portable/CommandLine/AnalyzerConfigSet.cs

+            {
+                if (_lazyConfigOptions is null)
+                {
+                    Interlocked.CompareExchange(


What is the value in using this lazy pattern given that AnalyzerConfigOptionsResult is a struct? This property will always return a copy of the value, the ParseGlobalConfigOptions can run multiple times even in this config. Can't see the value StrongBox<AnalyzerConfigOptionResult> provides over AnalyzerConfigOptionResult? (nullable value).

What am I missing here? #Resolved

Because the struct is bigger than an IntPtr, assignment isn't guaranteed to be atomic (AFAIK). Means one thread could read from it and get a half initialized struct if we don't box + interlock exchange.

In reply to: 475294138 [](ancestors = 475294138)

jaredpar · 2020-08-24T01:09:31Z

src/Compilers/Core/Portable/CommandLine/AnalyzerConfigSet.cs

+        private static void ParseSectionOptions(Section section, TreeOptions.Builder treeBuilder, AnalyzerOptions.Builder analyzerBuilder, ArrayBuilder<Diagnostic> diagnosticBuilder, string analyzerConfigPath, ConcurrentDictionary<ReadOnlyMemory<char>, string> diagIdCache)
+        {
+            const string DiagnosticOptionPrefix = "dotnet_diagnostic.";
+            const string DiagnosticOptionSuffix = ".severity";


Believe const locals should be camel cased #Resolved

jaredpar · 2020-08-24T01:11:12Z

src/Compilers/Core/Portable/Compilation/SyntaxTreeOptionsProvider.cs

+        /// <summary>
+        /// Get diagnostic severity set globally for a given diagnostic identifier
+        /// </summary>
+        public abstract bool TryGetGlobalDiagnosticValue(string diagnosticId, out ReportDiagnostic severity);
    }

    internal sealed class CompilerSyntaxTreeOptionsProvider : SyntaxTreeOptionsProvider


These are not locked yet, we can still change them if needed.

jaredpar · 2020-08-24T01:12:43Z

src/Compilers/Core/Portable/DiagnosticAnalyzer/AnalyzerDriver.cs

@@ -1255,8 +1255,13 @@ private ImmutableHashSet<DiagnosticAnalyzer> ComputeSuppressedAnalyzersForTree(S
                var hasUnsuppressedDiagnostic = false;
                foreach (var descriptor in descriptors)
                {
-                    if (!options.TryGetDiagnosticValue(tree, descriptor.Id, out var configuredSeverity) ||
-                        configuredSeverity != ReportDiagnostic.Suppress)
+                    options.TryGetGlobalDiagnosticValue(descriptor.Id, out var configuredSeverity);


Consider explicitly discarding to acknowledge you're deliberately ignoring this result rather than accidentally doing so.

Suggested change

options.TryGetGlobalDiagnosticValue(descriptor.Id, out var configuredSeverity);

_ = options.TryGetGlobalDiagnosticValue(descriptor.Id, out var configuredSeverity);

``` #Resolved

The way you are using configuredSeverity here implies that on a false return the value is guaranteed to be default. What happens if we pass a non-defaulted value into the API. Does it default the value?

It explicitly sets it to Default (see https://github.com/dotnet/roslyn/pull/45871/files/55ac7e5acdd2a1f91625ce31caa99af5cc43b5b9#diff-ffb0437c61e5f1029ae0c11cb59507d9R89). It's an out param though, so I don't understand how we'd pass in an explicit value?

In reply to: 475295255 [](ancestors = 475295255)

jaredpar · 2020-08-24T01:14:54Z

src/Test/Utilities/Portable/Compilation/TestSyntaxTreeOptionsProvider.cs

@@ -26,17 +27,27 @@ public sealed class TestSyntaxTreeOptionsProvider : SyntaxTreeOptionsProvider
                    x => x.Item2,
                    comparer)
            );
+            if (globalOption.key is object)
+            {
+                _globalOptions = new Dictionary<string, ReportDiagnostic>() { { globalOption.key, globalOption.diagnostic } };


Please use an explicit comparer. #Resolved

jaredpar · 2020-08-24T01:17:50Z

src/Compilers/CSharp/Portable/Compilation/CSharpDiagnosticFilter.cs

@@ -186,6 +187,11 @@ internal static class CSharpDiagnosticFilter
                // 3. Compilation level
                isSpecified = true;
            }
+            else if (syntaxTreeOptions is object && syntaxTreeOptions.TryGetGlobalDiagnosticValue(id, out report))
+            {
+                // 4. Global analyzer config options


Crazy specific NIT: the rest of the comments match up with the comment verbiage at the top of the method.

Suggested change

// 4. Global analyzer config options

// 4. Global analyzer config level

``` #Resolved

chsienki · 2020-08-24T16:30:27Z

Hmm, I'm not sure we do? We should write one, unless @mavasani knows of such a thing already?

In reply to: 678853235 [](ancestors = 678853235)

mavasani · 2020-08-24T16:34:28Z

Hmm, I'm not sure we do? We should write one, unless @mavasani knows of such a thing already?

In reply to: 678853235 [](ancestors = 678853235)

I don't believe have such a document, might be good to add something here: https://github.com/dotnet/roslyn/tree/master/docs/analyzers

chsienki · 2020-08-26T16:58:31Z

Ping @dotnet/roslyn-compiler for second review please :)

chsienki · 2020-08-26T17:03:28Z

Filed #47154 to track creating a doc

In reply to: 679233670 [](ancestors = 679233670,678853235)

- Pre-parse the global options into their own collection - Don't put global diagnostics on the tree level options - Add global diagnostics onto the tree options provider - Pass through the global diagnostics during compilation - Update the analyzer supression code - Add tests

Add tests to verify

add a test for no global config

* upstream/master: (43 commits) Fix typos (dotnet#47264) Disable CS0660 where Full Solution Analysis produces a different result Pass in correct arguments to TypeScript handler Add skipped failing test for IDE0044 (dotnet#45288) Avoid first chance exception Code review feedback part 3 Update stale URLs in the readme Update IntegrationTestSourceGenerator.cs Fix comment Update IntegrationTestSourceGenerator.cs Remove unused reference to obsolete class Update nullable annotations in Classification folder Update generator public API names: (dotnet#47222) NullableContextAttribute for property parameters is from accessor (dotnet#47223) Modify global analyzer config precedence: (dotnet#45871) Skip the C# source generator integration test running for now Additional XAML LSP handlers (dotnet#47217) Remap diagnostics using IWorkspaceVenusSpanMappingService in-proc Warn on type named "record" (dotnet#47094) Bail out early before getting SyntaxTree ...

chsienki added Area-Analyzers Area-Compilers and removed Area-Analyzers labels Jul 10, 2020

chsienki force-pushed the analyer_config_global_precedence2 branch from 890c36e to 47f9622 Compare July 10, 2020 17:22

chsienki commented Jul 10, 2020

View reviewed changes

mavasani reviewed Jul 10, 2020

View reviewed changes

chsienki force-pushed the analyer_config_global_precedence2 branch 2 times, most recently from 10ea50a to 26bf2ee Compare August 10, 2020 21:58

chsienki marked this pull request as ready for review August 11, 2020 18:01

chsienki requested review from a team as code owners August 11, 2020 18:01

jcouv reviewed Aug 17, 2020

View reviewed changes

jcouv approved these changes Aug 17, 2020

View reviewed changes

jcouv self-assigned this Aug 17, 2020

jasonmalinowski reviewed Aug 19, 2020

View reviewed changes

mavasani mentioned this pull request Aug 20, 2020

Precedence of diagnostic severity settings from compilation options, editorconfig and global analyzerconfig #44804

Closed

jaredpar reviewed Aug 24, 2020

View reviewed changes

chsienki mentioned this pull request Aug 26, 2020

Add documentation explaining warning suppression precendence #47154

Closed

jaredpar approved these changes Aug 28, 2020

View reviewed changes

chsienki added 13 commits August 28, 2020 12:01

Make specific diagnostics higher precedence than general diagnostics:

942681f

Add tests to verify

Fix tests

d9523c9

Fix nullability

cd8c17d

Fix vb tests (again :/)

ff5a735

Narf.

b934742

Fix merge

4331c74

Fixup workspace test

5fa6fd3

Lazily initialize global options

f16cb38

add a test for no global config

Fix caching analyzer config set

be98a85

PR Feedback

d601b36

PR feedback

8dfe71d

Add cancellation token to global version to match tree based one

daadd45

chsienki force-pushed the analyer_config_global_precedence2 branch from 7413077 to daadd45 Compare August 28, 2020 19:15

chsienki merged commit c75bc71 into dotnet:master Aug 29, 2020

ghost added this to the Next milestone Aug 29, 2020

allisonchou modified the milestones: Next, 16.8.P3 Aug 31, 2020

mavasani mentioned this pull request Sep 3, 2020

.editorconfig diagnostic suppressions don't apply to source-generated files #47384

Open

JasonBock mentioned this pull request Oct 20, 2023

Allow Types With Obsolete Members to Be Generated if ROCK10 is Suppressed JasonBock/Rocks#267

Closed

	options.TryGetGlobalDiagnosticValue(descriptor.Id, out var configuredSeverity);
	_ = options.TryGetGlobalDiagnosticValue(descriptor.Id, out var configuredSeverity);
	``` #Resolved

	// 4. Global analyzer config options
	// 4. Global analyzer config level
	``` #Resolved

Modify global analyzer config precedence: #45871

Modify global analyzer config precedence: #45871

Conversation

chsienki commented Jul 10, 2020 • edited Loading

chsienki Jul 10, 2020 • edited by jcouv Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

mavasani Jul 10, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

chsienki Jul 10, 2020 • edited Loading

Choose a reason for hiding this comment

chsienki Jul 10, 2020 • edited Loading

Choose a reason for hiding this comment

mavasani Jul 10, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

mavasani Jul 10, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

mavasani Jul 10, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

chsienki commented Aug 11, 2020

jcouv commented Aug 11, 2020 • edited Loading

jaredpar commented Aug 12, 2020 • edited by chsienki Loading

chsienki commented Aug 17, 2020

jcouv Aug 17, 2020 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jcouv Aug 17, 2020 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jcouv Aug 17, 2020 • edited Loading

Choose a reason for hiding this comment

chsienki Aug 17, 2020 • edited Loading

Choose a reason for hiding this comment

jcouv commented Aug 17, 2020 • edited by chsienki Loading

jcouv Aug 17, 2020 • edited Loading

Choose a reason for hiding this comment

jcouv left a comment

Choose a reason for hiding this comment

chsienki commented Aug 17, 2020

chsienki commented Aug 19, 2020

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jasonmalinowski Aug 19, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

jasonmalinowski Aug 19, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jaredpar commented Aug 24, 2020 • edited by chsienki Loading

jaredpar Aug 24, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jaredpar Aug 24, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jaredpar Aug 24, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jaredpar Aug 24, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

jaredpar Aug 24, 2020 • edited by chsienki Loading

Choose a reason for hiding this comment

chsienki commented Aug 24, 2020

mavasani commented Aug 24, 2020

chsienki commented Aug 26, 2020

chsienki commented Aug 26, 2020

chsienki commented Jul 10, 2020 •

edited

Loading

chsienki Jul 10, 2020 •

edited by jcouv

Loading

mavasani Jul 10, 2020 •

edited by chsienki

Loading

chsienki Jul 10, 2020 •

edited

Loading

chsienki Jul 10, 2020 •

edited

Loading

mavasani Jul 10, 2020 •

edited by chsienki

Loading

mavasani Jul 10, 2020 •

edited by chsienki

Loading

mavasani Jul 10, 2020 •

edited by chsienki

Loading

jcouv commented Aug 11, 2020 •

edited

Loading

jaredpar commented Aug 12, 2020 •

edited by chsienki

Loading

jcouv Aug 17, 2020 •

edited

Loading

jcouv Aug 17, 2020 •

edited

Loading

jcouv Aug 17, 2020 •

edited

Loading

chsienki Aug 17, 2020 •

edited

Loading

jcouv commented Aug 17, 2020 •

edited by chsienki

Loading

jcouv Aug 17, 2020 •

edited

Loading

jasonmalinowski Aug 19, 2020 •

edited by chsienki

Loading

jasonmalinowski Aug 19, 2020 •

edited by chsienki

Loading

jaredpar commented Aug 24, 2020 •

edited by chsienki

Loading

jaredpar Aug 24, 2020 •

edited by chsienki

Loading

jaredpar Aug 24, 2020 •

edited by chsienki

Loading

jaredpar Aug 24, 2020 •

edited by chsienki

Loading

jaredpar Aug 24, 2020 •

edited by chsienki

Loading

jaredpar Aug 24, 2020 •

edited by chsienki

Loading