chore: add benchmark suite with dashboard#244
Merged
Conversation
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add clone.sh and clone_with_hooks.sh that benchmark git-worktree-clone against manual git clone --bare + worktree add for small/medium/large repos. The hooks variant tests daft's automatic hook execution vs a competent manual user replicating hook work with parallelism. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add checkout.sh and checkout_with_hooks.sh that benchmark git-worktree-checkout and git-worktree-checkout-branch against manual git worktree add. Tests both existing branch checkout and new branch creation for small/medium/large repos. The hooks variant tests automatic hook execution vs parallel manual replication. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add four utility command benchmarks: - init.sh: git-worktree-init vs manual git init --bare + worktree add - prune.sh: git-worktree-prune vs git worktree prune with 10 stale worktrees - fetch.sh: git-worktree-fetch --all vs parallel xargs git fetch with 2 remotes - branch_delete.sh: git-worktree-branch-delete vs manual worktree remove + branch -D Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add workflow_full.sh that benchmarks an end-to-end workflow: clone a repo then create 3 feature branches. The git side uses maximum parallelism (& + wait) for worktree creation, giving it every advantage against daft's sequential execution model. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add vs_competition.sh for opt-in competitor benchmarking. Always compares daft against a shell alias pattern (the most common manual approach). Optionally benchmarks against git-town if installed. This scenario is excluded from run_all.sh and must be invoked explicitly. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add run_all.sh that runs all benchmark scenarios (except vs_competition) and aggregates hyperfine markdown results into results/summary.md. Supports --only <scenario> to run a single benchmark and --list to show available scenarios. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…workflow_full - Removed unused GIT_HOOK_WORK variable from clone_with_hooks.sh - workflow_full.sh now uses create_bare_repo_with_hooks and includes: - post-clone hook work on the git side - worktree-post-create hook work for each branch (parallelized) - pruning 2 worktrees at the end (parallelized) - Git side uses maximum parallelism at every stage Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
daft commands operate in the current directory rather than taking destination paths. Fixed all scenarios to cd into the target directory before running daft commands. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Remove benchmark results from docs/benchmarks/ and benches/history/ since results will now be pushed to the private daft-benchmarks repo. Update the CI workflow to package results and push them externally instead of committing directly to this repo. Clean up VitePress nav/sidebar entries for Benchmarks. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
test:perftask with a comprehensive benchmark suite comparing daft vs equivalent git scripting (with maximum parallelism)daft-benchmarksrepo with an Astro dashboard on GitHub PagesWhat's included
benches/— benchmark framework, fixtures, and 9+ scenario scriptsmise-tasks/bench/— 13 mise tasks (bench:init,bench:clone, etc.).github/workflows/bench.yml— CI workflow that runs benchmarks and pushes results to private repobenches/package_results.sh— packages hyperfine JSON into metadata envelopesManual setup required
After merging, add a
BENCH_REPO_TOKENrepository secret (Settings > Secrets > Actions) with a fine-grained PAT that has Contents write access toavihut/daft-benchmarks.Test plan
bench:initverified (three-way: daft vs daft-gitoxide vs git)bench:cloneverified (small/medium/large)bench:checkoutverified (existing + new branch, all sizes)🤖 Generated with Claude Code