git graphs, hashes, and compression, oh my

at with Matthew McCullough

GitGraphs, Hashes, and Compression, Oh My!

@matthewmccull

[email protected]

github.com/training

GIT - the stupid content tracker

"git" can mean anything, depending on

your mood.

* random three-letter combination that

is pronounceable, and not actually

used by any

common UNIX command. The fact that it

is a mispronunciation of "get" may or

may not be

relevant.* stupid. contempti

ble and despicable. simple. Take your

pick from the dictionary of

slang.* "global informati

on tracker": you're in a good mood, an

d it actually works for you.

Angels sing, and a light suddenly fill

s the room.

* "goddamn idiotic truckload of sh*t":

when it breaks

Git is a fast, scalable, distributed r

evision control system with an unusual

ly rich

command set that provides both high-le

vel operations and full access to inte

rnals.

Git is an Open Source project covered

by the GNU General Public License. It

was originally

written by Linus Torvalds with help of

a group of hackers around the net. It

is currently

maintained by Junio C Hamano.

2005git

ArchitectureHashes

centralized VCSs use sequential revision numbers

Git uses a SHA-1 hash

40 hex characters (20 bytes)

9AB223D28B1AA46EF1780B22F304982E39872C34

</body></html>

</body></html>

9AB223D28B1AA46EF1780B22F304982E39872C34

Hashed ContentGit commits without Git

# Green field project$ git init newproject$ cd newproject# ...start coding

.git!"" COMMIT_EDITMSG!"" HEAD!"" MERGE_RR!"" config!"" description!"" hooks# !"" pre-commit.sample# $"" update.sample!"" index!"" info# $"" exclude!"" logs# !"" HEAD# $"" refs# $"" heads# $"" master!"" objects# !"" 54# # $"" 3b9bebdc6bd5c4b22136034a95dd097a57d3dd# !"" info# $"" pack!"" refs !"" heads # $"" master $"" tags

$ printf "blob 12\000Hello World\n" | shasum

ArchitectureHash Shortcuts

use as little of it as is unique

9AB223D28B1AA46EF1780B22F304982E39

872C34a certain commit

9AB22Fa certain commit

git rev-parse 9AB22Fexpand the commit ref

ArchitectureStorage

Typical SCMs use delta storage

CVS / Subversion / darcs / Mercurial

v1 v2 v3 v4

File A

File B

File C

File A

File B File B

File C

File A

File BFile B

v1 v2 v3 v4

File A

File B

File C

File A

File B File B

File C

File A

File BFile B

ΔΔ ΔΔΔ

Checkin

Checkin Checkin Che

Checkin

Delta storage gets slower as the history of a file gets longer

Directed Acyclic Graph

Copy of the entire tree per checkin

cp -r srcfolder srcfolder.prev

v1 v2 v3 v4

File A

File B

File C

File A

File B File B

File C

File A

File BFile B

File A File A

File C File C File C

v1 v2 v3 v4

File A

File B

File C

File A

File B File B

File C

File A

File BFile B

File A File A

hard link to existing identical blobs

v1 v2 v3 v4

File A

File B

File C

File A

File B File B

File C

File A

File BFile B

File A File A

v1 v2 v3 v4

File A

File B

File C

File A

File B File B

File C

File A

File BFile B

File A File A

File C File C File Cß

zlib deflates each blob at commit

v1 v2 v3 v4

File A

File B

File C

File A

File B File B

File C

File A

File BFile B

File A File A

v1 v2 v3 v4

File A

File B

File C

File A

File B File B

File C

File A

File BFile B

File A File A

zlib deflates the entire repo

v1 v2 v3 v4

File A

File B

File C

File A

File B File B

File C

File A

File BFile B

File A File A

v1 v2 v3 v4

File A

File B

File C

File A

File B File B

File C

File A

File BFile B

File A File A

2100 MB became 205 MB

ArchitectureHash relationships

‣Blob‣Tree‣Commit‣Tag

treetree: 7e8b1 webblob: 9ab16 index.html

a10b3 treeblob: 8d162 logo.jpgblob: 51d22 draw.js

7e8b1commit

tree: a10b3parent: nilauthor: Firdcommitter: Matthewmessage:

Major refactoringof the web content.

blob<html>

blob//Some more javascriptvar renderSize

blob7D 8D B3 7F BD 12 9F E97B 78 9D 3F 5C A6 72 CB

7e8b1commit

blob<html>

7e8b1commit

blob<html>

7e8b1commit

blob<html>

7e8b1commit

blob<html>

7e8b1commit

blob<html>

7e8b1commit

blob<html>

7e8b1commit

blob<html>

7e8b1commit

blob<html>

v1 v2 v3

committree: 9a87bparent: nilauthor: Firdcommitter: Matthewmessage:

Major refactoringof the Javascript renderingengine.

committree: b22c1parent: c67dbauthor: Timcommitter: Firdmessage:

Minor update to HTML

committree: b22c1parent: 9bd21author: Johnnycommitter: Joemessage:

New language transations

v1 v2 v3

RELEASE_1.0 HEAD

bug979branch

commitc67db

commit9bd21

commit1c2d7

commit8c2d1

commit1bdcd

commit2daa1

RELEASE_1.0 HEAD

bug979branch

commitc67db

commit9bd21

commit1c2d7

commit8c2d1

commit1bdcd

commit2daa1

RELEASE_1.0 HEAD

bug979branch

commitc67db

commit9bd21

commit1c2d7

commit8c2d1

commit1bdcd

commit2daa1

RELEASE_1.0 HEAD

bug979branch

commitc67db

commit9bd21

commit1c2d7

commit8c2d1

commit1bdcd

commit2daa1

RELEASE_1.0 HEAD

bug979branch

commitc67db

commit9bd21

commit1c2d7

commit8c2d1

commit1bdcd

commit2daa1

RELEASE_1.0 HEAD

bug979branch

commitc67db

commit9bd21

commit1c2d7

commit8c2d1

commit1bdcd

commit2daa1

ArchitectureHash shortcuts

commitish & treeish

commitish = shorthand for commit hashes

treeish = shorthand for tree hashes

9AB22Fa certain commit

9AB22F^one commit before a certain commit

9AB22F^^two commits before a certain commit

9AB22F~5five commits before a certain commit

9AB223..56CD77between these two commits

HEADthe most recent commit on this branch

HEAD^one commit before the most recent commit

HEAD~2two commits before the most recent commit

HEAD..HEAD^^^between the given recent commits

masterthe most recent commit on this branch

master^^two commits before the most recent commit on this branch

master~5five commits before the most recent commit on this branch

remotes/origin/masterthe most recent commit on this remote tracking branch

origin/masterthe most recent commit on this remote tracking branch

The GraphVerification

git fsck

git verify-pack -v .git/objects/pack/FILENAME

The GraphMore graph navigation

master^{tree}find this commit’s tree

git describe HASHfind nearest tag

:/some wordscommit message that starts with

REF:FILEblob spec

:0:FILEindex

:1:FILEcommon ancestor

:2:FILEtarget

:3:FILEmerging in

@matthewmccull

[email protected]

github.com/training

at with Matthew McCullough

GitGraphs, Hashes, and Compression, Oh My!

git graphs, hashes, and compression, oh my

Technology