Serialized Commit Graph sketch - for Git Rev News

for-edition-40.md

Rough timeline

• Re: What's cooking in git.git (Jan 2018, #03; Tue, 23) by Jeff Hostetler, 24 Jan 2018 -- announcement that Derrick Stolee is working on the commit-graph feature

• [PATCH 00/14] Serialized Commit Graph by Derrick Stolee, 25 Jan 2018 -- first version of first patch series (ds/commit-graph)

• Git Merge contributor summit notes by Alex Vandiver, 9 Mar 2018 -- in Git Rev News Edition 37

  Serialized commit graph (stolee)

  • current patch includes a way to write a file which is commit history
  • replaces object database lookups by separate file
  • stores object id, commit date, parent information
  • parent information includes reference id
  • git log --graph is 10% of the time it used to be
  • putting generation numbers in speeds things up further
  • queued up for the next series
  • generation is an integer, guarantees that everything that it can reach is lower
  • root is id 1, every child is max(parents)+1 [...]

• Making Git for Windows - Git Merge 2018 Derrick Stolee part starts at 14:50 information that videos of the 2018 Git Merge talks have been released in Git Rev News Edition 38

• [PATCH v2 00/12] Integrate commit-graph into fsck and gc by Derrick Stolee, 11 May 2018 -- currently most recent series

• Re: [RFC] Other chunks for commit-graph, part 1 - Bloom filters, topo order, etc. by Derrick Stolee, 14 May 2018 -- milestones

  I suppose that is what I intended in the "Future Work" section of Documentation/technical/commit-graph.txt. It gives a set of things that need to be done in order to make this a default feature, not just an expert-level feature. When I wrote the section, I was focused entirely on "commit-graph version 1.0" and I didn't know how long that would take. The series have been getting interest and review (in great part to your interest, Jakub) so they have been moving to 'next' faster than I anticipated.

  I'll plan on writing a more detailed list of future directions, but for now I'll list the things I know about and how I see them in priority order:

  Commit-graph v1.0:

  • ds/generation-numbers
  • 'verify' and fsck/gc integration
  • correct behavior with shallow clones, replace-objects, and grafts

  Commit-graph v1.1:

  • Place commit-graph storage in the_repository
  • 'git tag --merged' use generation numbers
  • 'git log --graph' use generation numbers

  Commit-graph v1.X:

  • Protocol v2 verb for sending commit-graph
  • Bloom filters for changed paths