docs: versioning model and proofs (draft)

[connortsui20]

Summary

Preserves the versioning model and proofs for separate review, split out of #9947. This is a draft and is not ready to merge.

Changes

Adds the Typst source and its existing PDF without changes. The proof depends on assumptions about reader and writer behavior, including planned compressor support.

[codspeed]

Merging this PR will degrade performance by 1.57%

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 2 improved benchmarks
❌ 3 regressed benchmarks
✅ 2248 untouched benchmarks
⏩ 218 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	WallTime	filtered_owned_i64_avx2[OneNullInEight]	22.3 µs	25.9 µs	-13.8%
❌	WallTime	dbp_split_kernel_mixed_null_neon[(I256, 1024)]	1.9 µs	2.2 µs	-12.01%
❌	WallTime	mul_i32_nonnull_avx512	7.1 µs	7.9 µs	-10.39%
⚡	Simulation	decompress[u64, (4000, 1024)]	86.6 µs	71.1 µs	+21.88%
⚡	WallTime	mul_u32_nonnull_avx512	6.3 µs	5.6 µs	+11.57%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing ct/versioning-proof-draft (4595fc3) with develop (c6678f3)}

Footnotes

1. 218 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩