feat: replace the MockolateUnionParameters opt-in with the MockolateDisableUnionParameters kill switch by vbreuss · Pull Request #929 · Testably/Mockolate

vbreuss · 2026-11-03T00:39:42Z

Now that C# 15 has shipped with .NET 11, union mode no longer needs an opt-in: it follows the project's language version on its own. The only switch left is MockolateDisableUnionParameters, which lets a project keep the classic setup and verify overloads with C# 15, e.g. to retarget to .NET 11 first and switch to the new methods later.

MockolateUnionParameters never shipped in a release, so removing it breaks no one.

The generator must still load in Roslyn 5.0, while the union tests need a Roslyn that knows C# 15. Mockolate.SourceGenerators.Tests therefore also runs on net10.0 against MinimumRoslynVersion (without the snapshot and union tests), so the classic output stays covered on the oldest supported compiler. Since net11.0 defaults to C# 15, the test projects no longer need <LangVersion>preview</LangVersion>.

Warning

Draft: the net11.0 generator tests need a Microsoft.CodeAnalysis.CSharp package with LanguageVersion.CSharp15, which is not on nuget.org yet. Until Directory.Packages.props references it, Mockolate.SourceGenerators.Tests does not compile for net11.0.

The union test job is renamed from "Union tests (net11.0, C# preview)" to "Union tests (net11.0, C# 15)".

github-actions · 2026-11-10T20:00:31Z

Test Results

0 files - 44 0 suites - 44 0s ⏱️ - 44m 2s
0 tests - 4 643 0 ✅ - 4 643 0 💤 ±0 0 ❌ ±0
0 runs - 45 750 0 ✅ - 45 750 0 💤 ±0 0 ❌ ±0

Results for commit 34d6fda. ± Comparison against base commit 130e803.

♻️ This comment has been updated with latest results.

github-actions · 2026-11-10T20:07:23Z

🚀 Benchmark Results

Compared to main @ 130e803f: memory to the latest run, time to the median of the last 10 runs, as the runners differ in speed. A relative change marks when Mockolate gets more than 15% faster or slower relative to Imposter and TUnitMocks: 🔴 / 🟢 against both of them, 🔸 / 🔹 against only one.

CallbackBenchmarks

✅ No changes in memory or relative performance (1 benchmarks)

All benchmarks compared to main

Benchmark	Time	Memory	vs Imposter	vs TUnitMocks
`Callback`	271.2 ns → 233.5 ns (-14%)	1,592 B → 1,592 B	0.69× → 0.66×	1.61× → 1.73×

Raw results

BenchmarkDotNet v0.15.8, Linux Ubuntu 26.04.1 LTS (Resolute Raccoon)
AMD EPYC 9V74 3.67GHz, 1 CPU, 4 logical and 2 physical cores
.NET SDK 11.0.100
  [Host] : .NET 11.0.0 (11.0.0, 11.0.26.53018), X64 RyuJIT x86-64-v4

Job=InProcess  Toolchain=InProcessEmitToolchain  IterationCount=15  
LaunchCount=1  WarmupCount=10

Callback	Mean	Error	StdDev	Ratio	Allocated	Alloc Ratio
Mockolate	233.5 ns	10.64 ns	9.95 ns	1.00	1592 B	1.00
Imposter	155.2 ns	2.51 ns	2.22 ns	0.67	616 B	0.39
TUnitMocks	404.7 ns	13.23 ns	12.38 ns	1.74	2056 B	1.29
Moq	60,020.0 ns	278.17 ns	232.29 ns	257.50	9063 B	5.69
NSubstitute	3,486.1 ns	95.85 ns	89.66 ns	14.96	7864 B	4.94
FakeItEasy	3,352.2 ns	115.29 ns	107.84 ns	14.38	6930 B	4.35

CompleteEventBenchmarks

✅ No changes in memory or relative performance (1 benchmarks)

All benchmarks compared to main

Benchmark	Time	Memory	vs Imposter	vs TUnitMocks
`Event`	201.4 ns → 209.7 ns (+4%)	1,728 B → 1,728 B	0.65× → 0.72×	0.60× → 0.57×

Raw results

BenchmarkDotNet v0.15.8, Linux Ubuntu 26.04.1 LTS (Resolute Raccoon)
AMD EPYC 9V74 3.69GHz, 1 CPU, 4 logical and 2 physical cores
.NET SDK 11.0.100
  [Host] : .NET 11.0.0 (11.0.0, 11.0.26.53018), X64 RyuJIT x86-64-v4

Job=InProcess  Toolchain=InProcessEmitToolchain  IterationCount=15  
LaunchCount=1  WarmupCount=10

Event	Mean	Error	StdDev	Ratio	Allocated	Alloc Ratio
Mockolate	209.7 ns	1.23 ns	1.09 ns	1.00	1728 B	1.00
Imposter	151.7 ns	0.93 ns	0.83 ns	0.72	512 B	0.30
TUnitMocks	119.7 ns	0.37 ns	0.29 ns	0.57	1392 B	0.81
Moq	10,407.8 ns	68.91 ns	57.54 ns	49.62	12761 B	7.38
NSubstitute	4,006.7 ns	14.41 ns	12.03 ns	19.10	9192 B	5.32
FakeItEasy	193,332.7 ns	460.43 ns	384.48 ns	921.81	15619 B	9.04

CompleteIndexerBenchmarks

✅ No changes in memory or relative performance (2 benchmarks)

All benchmarks compared to main

Benchmark	Time	Memory	vs Imposter
`Indexer (N=1)`	818.4 ns → 967.4 ns (+18%)	3,832 B → 3,832 B	0.35× → 0.32×
`Indexer (N=10)`	2.35 μs → 2.49 μs (+6%)	4,912 B → 4,912 B	0.30× → 0.30×

Raw results

BenchmarkDotNet v0.15.8, Linux Ubuntu 26.04.1 LTS (Resolute Raccoon)
AMD EPYC 7763 3.23GHz, 1 CPU, 4 logical and 2 physical cores
.NET SDK 11.0.100
  [Host] : .NET 11.0.0 (11.0.0, 11.0.26.53018), X64 RyuJIT x86-64-v3

Job=InProcess  Toolchain=InProcessEmitToolchain  IterationCount=15  
LaunchCount=1  WarmupCount=10

Indexer	N	Mean	Error	StdDev	Ratio	Allocated	Alloc Ratio
Mockolate	1	967.4 ns	24.27 ns	21.51 ns	1.00	3.74 KB	1.00
Imposter	1	311.1 ns	2.77 ns	2.45 ns	0.32	1.07 KB	0.29
Moq	1	230,947.6 ns	2,681.35 ns	2,239.05 ns	238.84	20.46 KB	5.47
NSubstitute	1	10,650.8 ns	48.39 ns	42.90 ns	11.01	12.78 KB	3.42
FakeItEasy	1	11,279.0 ns	71.01 ns	66.42 ns	11.66	13.27 KB	3.55

Mockolate	10	2,494.7 ns	52.43 ns	49.05 ns	1.00	4.8 KB	1.00
Imposter	10	753.3 ns	14.50 ns	13.56 ns	0.30	2.13 KB	0.44
Moq	10	241,357.3 ns	941.00 ns	880.21 ns	96.78	29.84 KB	6.22
NSubstitute	10	25,598.3 ns	241.61 ns	226.00 ns	10.26	25.58 KB	5.33
FakeItEasy	10	23,694.6 ns	338.35 ns	316.49 ns	9.50	32.47 KB	6.77

CompleteMethodBenchmarks

Relative performance (how many times as long the other library takes as Mockolate, main → PR)

Benchmark	vs Imposter	vs TUnitMocks
`Method (N=1)`	0.52× → 0.30× 🔴 +71%	1.30× → 0.77× 🔴 +70%

All benchmarks compared to main

Benchmark	Time	Memory	vs Imposter	vs TUnitMocks
`Method (N=1)`	290.1 ns → 451.8 ns (+56%)	1,936 B → 1,936 B	0.52× → 0.30× 🔴 +71%	1.30× → 0.77× 🔴 +70%
`Method (N=10)`	531.1 ns → 458.7 ns (-14%)	2,152 B → 2,152 B	0.56× → 0.60×	2.14× → 2.32×

Raw results

BenchmarkDotNet v0.15.8, Linux Ubuntu 26.04.1 LTS (Resolute Raccoon)
AMD EPYC 9V74 3.69GHz, 1 CPU, 4 logical and 2 physical cores
.NET SDK 11.0.100
  [Host] : .NET 11.0.0 (11.0.0, 11.0.26.53018), X64 RyuJIT x86-64-v4

Job=InProcess  Toolchain=InProcessEmitToolchain  IterationCount=15  
LaunchCount=1  WarmupCount=10

Method	N	Mean	Error	StdDev	Ratio	Allocated	Alloc Ratio
Mockolate	1	451.8 ns	5.09 ns	4.25 ns	1.00	1936 B	1.00
Imposter	1	137.6 ns	1.28 ns	1.13 ns	0.30	728 B	0.38
TUnitMocks	1	345.7 ns	1.82 ns	1.61 ns	0.77	2088 B	1.08
Moq	1	114,005.7 ns	582.13 ns	516.04 ns	252.37	15074 B	7.79
NSubstitute	1	4,350.0 ns	60.26 ns	53.42 ns	9.63	9320 B	4.81
FakeItEasy	1	4,018.9 ns	113.54 ns	106.21 ns	8.90	8160 B	4.21

Mockolate	10	458.7 ns	5.43 ns	4.54 ns	1.00	2152 B	1.00
Imposter	10	274.5 ns	2.17 ns	1.81 ns	0.60	1088 B	0.51
TUnitMocks	10	1,063.5 ns	12.27 ns	10.87 ns	2.32	3840 B	1.78
Moq	10	117,724.1 ns	629.61 ns	558.13 ns	256.68	19065 B	8.86
NSubstitute	10	6,777.7 ns	85.72 ns	75.99 ns	14.78	12344 B	5.74
FakeItEasy	10	6,806.4 ns	76.98 ns	64.28 ns	14.84	16393 B	7.62

CompletePropertyBenchmarks

Relative performance (how many times as long the other library takes as Mockolate, main → PR)

Benchmark	vs Imposter	vs TUnitMocks
`Property (N=1)`	0.33× → 0.16× 🔴 +102%	0.85× → 0.39× 🔴 +116%

All benchmarks compared to main

Benchmark	Time	Memory	vs Imposter	vs TUnitMocks
`Property (N=1)`	499.8 ns → 828.9 ns (+66%)	2,432 B → 2,432 B	0.33× → 0.16× 🔴 +102%	0.85× → 0.39× 🔴 +116%
`Property (N=10)`	999.3 ns → 779.6 ns (-22%)	2,936 B → 2,936 B	0.33× → 0.31×	1.56× → 1.52×

Raw results

BenchmarkDotNet v0.15.8, Linux Ubuntu 26.04.1 LTS (Resolute Raccoon)
AMD EPYC 9V74 3.69GHz, 1 CPU, 4 logical and 2 physical cores
.NET SDK 11.0.100
  [Host] : .NET 11.0.0 (11.0.0, 11.0.26.53018), X64 RyuJIT x86-64-v4

Job=InProcess  Toolchain=InProcessEmitToolchain  IterationCount=15  
LaunchCount=1  WarmupCount=10

Property	N	Mean	Error	StdDev	Ratio	Allocated	Alloc Ratio
Mockolate	1	828.9 ns	6.82 ns	6.38 ns	1.00	2432 B	1.00
Imposter	1	134.6 ns	1.18 ns	1.10 ns	0.16	672 B	0.28
TUnitMocks	1	326.9 ns	1.86 ns	1.45 ns	0.39	1696 B	0.70
Moq	1	7,500.2 ns	104.45 ns	97.70 ns	9.05	10584 B	4.35
NSubstitute	1	5,218.1 ns	29.54 ns	27.63 ns	6.30	11664 B	4.80
FakeItEasy	1	5,135.2 ns	15.61 ns	13.04 ns	6.20	11212 B	4.61

Mockolate	10	779.6 ns	7.59 ns	6.34 ns	1.00	2936 B	1.00
Imposter	10	245.5 ns	1.71 ns	1.60 ns	0.31	960 B	0.33
TUnitMocks	10	1,181.8 ns	5.63 ns	4.70 ns	1.52	4048 B	1.38
Moq	10	12,424.4 ns	79.27 ns	70.27 ns	15.94	18521 B	6.31
NSubstitute	10	12,046.6 ns	52.46 ns	46.51 ns	15.45	21529 B	7.33
FakeItEasy	10	12,527.8 ns	144.90 ns	120.99 ns	16.07	30801 B	10.49

MockCreationBenchmarks

✅ No changes in memory or relative performance (1 benchmarks)

All benchmarks compared to main

Benchmark	Time	Memory	vs Imposter	vs TUnitMocks
`CreateMock`	16.4 ns → 21.7 ns (+33%)	160 B → 160 B	5.55× → 5.10×	1.64× → 1.59×

Raw results

BenchmarkDotNet v0.15.8, Linux Ubuntu 26.04.1 LTS (Resolute Raccoon)
Intel Xeon Platinum 8370C CPU 2.80GHz (Max: 2.79GHz), 1 CPU, 4 logical and 2 physical cores
.NET SDK 11.0.100
  [Host] : .NET 11.0.0 (11.0.0, 11.0.26.53018), X64 RyuJIT x86-64-v4

Job=InProcess  Toolchain=InProcessEmitToolchain  IterationCount=15  
LaunchCount=1  WarmupCount=10

CreateMock	Mean	Error	StdDev	Ratio	Allocated	Alloc Ratio
Mockolate	21.69 ns	0.710 ns	0.664 ns	1.00	160 B	1.00
Imposter	110.68 ns	1.295 ns	1.212 ns	5.11	504 B	3.15
TUnitMocks	34.56 ns	0.701 ns	0.656 ns	1.59	216 B	1.35
Moq	1,345.17 ns	3.946 ns	3.498 ns	62.07	2088 B	13.05
NSubstitute	1,912.41 ns	11.392 ns	10.656 ns	88.24	5032 B	31.45
FakeItEasy	1,637.73 ns	8.903 ns	7.892 ns	75.57	2764 B	17.27

…DisableUnionParameters kill switch Once C# 15 ships, union mode follows the language version on its own; the remaining switch only lets projects keep the classic overloads. WIP: the generator tests need a Roslyn package with LanguageVersion.CSharp15, which is not on nuget.org yet, so this does not compile until that package is referenced. The v4 upgrade guide (Docs/changelogs/migration-to-v4.md) still describes the MockolateUnionParameters opt-in and has to move to the kill switch as well.

The generator must still load in Roslyn 5.0, so the generator tests also run against it on net10.0, while net11.0 needs a Roslyn that knows C# 15. Now that .NET 11 defaults to C# 15, the preview language version is no longer needed for the union-typed setup surface.

vbreuss self-assigned this Nov 3, 2026

Base automatically changed from feat/union-parameters to main November 3, 2026 02:02

vbreuss force-pushed the feat/union-parameters-disable-switch branch from 44eda65 to 934295d Compare November 10, 2026 19:44

vbreuss and others added 2 commits November 11, 2026 14:05

vbreuss force-pushed the feat/union-parameters-disable-switch branch from 43a29aa to 34d6fda Compare November 11, 2026 14:26

vbreuss changed the title ~~feat!: replace the MockolateUnionParameters opt-in with the MockolateDisableUnionParameters kill switch~~ feat: replace the MockolateUnionParameters opt-in with the MockolateDisableUnionParameters kill switch Nov 11, 2026

Navigation Menu

Repository navigation

feat: replace the MockolateUnionParameters opt-in with the MockolateDisableUnionParameters kill switch - #929

feat: replace the MockolateUnionParameters opt-in with the MockolateDisableUnionParameters kill switch#929
vbreuss wants to merge 2 commits into
mainfrom
feat/union-parameters-disable-switch

vbreuss commented Nov 3, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Nov 10, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Nov 10, 2026 •

edited

Loading

Uh oh!

This branch has not been deployed

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

vbreuss commented Nov 3, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Nov 10, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Test Results

Uh oh!

github-actions Bot commented Nov 10, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🚀 Benchmark Results

CallbackBenchmarks

CompleteEventBenchmarks

CompleteIndexerBenchmarks

CompleteMethodBenchmarks

CompletePropertyBenchmarks

MockCreationBenchmarks

Uh oh!

This branch has not been deployed

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

vbreuss commented Nov 3, 2026 •

edited

Loading

github-actions Bot commented Nov 10, 2026 •

edited

Loading

github-actions Bot commented Nov 10, 2026 •

edited

Loading