Introduction to Git-SVN

Total Page:16

File Type:pdf, Size:1020Kb

Introduction to Git-SVN Introduction to Git-SVN Patrick Burma, Solutions Architect, WANdisco Agenda • Migration of SVN and Git • Tools and Techniques • Git-SVN • Sub-Git • Use Cases • References Integrating Subversion and Git Migration of SVN and Git Overview • Lot’s of existing production Subversion repositories, Git is newer and not as mature but growing rapidly – Need a way to mitigate risk, and keep “golden processes” in place while Git infrastructure matures – Others? • Types of migration – One time convert from one system to another – Continuously push data from one system to another – Sync data back and forth between the two Migration of SVN and Git Challenges • Some good ways to push data from one system to another • No great ways to bidirectional sync • Why? – System architectures too different • What’s the answer – Let’s look at our options and then decide…. • Maybe its best to keep one, the other or both as separate systems Migration of SVN and Git High Level Overview • Subversion system – Revs are based on Changesets – Merges are tracked using metadata (svn:mergeinfo) – Data stored in the FSFS • Git – Based on SHA’s – Different merge algorithms (eg; up-to-date, fast-forward) – Data stored in .git repository format • Terminologies and actions differ – Actions in one that don’t exist in the other • Consider clone, push, pull, propset • Some actions do overlap, here is where integration occurs – Commit, rename, move, branch Migration of SVN and Git Use Cases • Best – One time import, for example importing a Subversion repository into Git • Ok – Single direction, for example pushing Git changes into Subversion • Worst – Continuous bi-directional integration between Subversion and Git Migration of SVN and Git Summary • Git and Subversion don’t directly translate – There is no equivalent git rebase command in Subversion • Differences in system architectures and data storage without a common language to interface between them (eg; SQL) • Troubles if you want to do bi-directional integration because: – Branching and merging won’t translate, without merge tracking you’ll be stuck doing 2-way diffs – At the very least you have to minimize some capabilities of either tools and go to a simple and very linear mainline development approach. – Other issues, such as externals, which there are no solutions for? – Any Gothcha’s? • For imports historical accuracy. Especially around moves/renames/ deletes. Tools and Techniques Tools • Git-SVN – Comes with Git • Sub-Git – Free to download and try, has a cost to use in production • Custom scripts and services • Others – Clearvision – Tailor (dead project) – SSCM (also dead) – More? Tools and Techniques Techniques • Consider a few Git best practices. – Tendency is to have many more Git repositories then Subversion • Scope out mapping options – Repo-to-repo, this would mean one Git repository to one Subversion repository – Project based? Many Git repositories map to a single Subversion repository as projects – Branches, Tags and Trunks. • Common layouts or unique layouts? Tools and Techniques Techniques • Automate with post-commit hooks • Develop new repo generation scripts – Sets config options and points to proper places • Author mapping • Understanding Limitations – Git Rebase – Fast-Forward Merges – Squashed History – Weird Stuff Git-SVN About • Comes with Git • Perl script • Several iterations and improvements Git-SVN Commands • Main git svn commands – Clone – Fetch – Log Git-SVN Examples • Create repositories from test Subversion repository for Git and workspace for SVN git svn clone https://pat-wandisco.googlecode.com/svn/trunk/ pat- wandisco.git --username [email protected] svn checkout https://pat-wandisco.googlecode.com/svn/trunk/ pat- wandisco --username [email protected] --password hH3YV5jc6CP4 • Modify and commit files from Git cd pat-wandisco.git/trunk echo test git commit >> file1.txt git commit –m “Adding my first change to git, my name is” git svn dcommit Git-SVN Examples • Modify files in Subversion cd ../../pat-wandisco/trunk svn up svn log echo test svn commit >> file1.txt svn ci –m “Adding my commit from Subversion” • Add another change from Git cd ../../pat-wandisco.git/trunk git svn fetch git rebase echo second change from git >> file1.txt” it svn dcommit Sub-Git Overview • Builds a Git repo from an SVN repo • Remote Installation – If the repositories are not on the same system • Local Installation – They are on the same box • Config file stores items such as mappings for projects, directories/branches, and authors – Allows for easy re-use/re-run of the syncs • Includes ignores, symlinks, merge tracking, renames, eol settings • Fairly simple to use • Not free beyond trial use Sub-Git Overview Sub-Git Examples • Setup service accounts – Git/Apache/SVN, run them as same user (easiest) • Run subgit configure – subgit configure reponame • Modify configuration file • reponame/conf/Subgit.conf • Run subgit install – subgit install reponame • Modify .git/config settings • Write and implement hooks Use Cases From the Audience • Git is new and growing, but all our infrastructure is built around Subversion • Don’t want to maintain two distinct ALM stacks • Can’t enforce security on Git, but can on Subversion • No Git infrastructure to speak of (some devs are just using it) • Need to collaborate between two or more BU’s, vendors, customers using either Git or Subversion References External Links • http://git-scm.com/book/en/Git-and-Other-Systems-Migrating-to-Git • http://www.drdobbs.com/architecture-and-design/migrating-from-subversion-to-git-and-the/240009175 • http://code.google.com/p/support/wiki/ConvertingSvnToGit • http://thomasrast.ch/git/git-svn-conversion.html • http://subgit.com • http://jaibeermalik.wordpress.com/2013/10/23/svn2git-migrating-repository-from-subversion-to-git/ • http://git-scm.com/book/ch8-1.html • http://maymay.net/blog/2009/02/24/how-to-use-git-svn-as-the-only-subversion-client-youll-need/ • https://git.wiki.kernel.org/index.php/Git-svn Thank You Pat Burma .
Recommended publications
  • Generating Commit Messages from Git Diffs
    Generating Commit Messages from Git Diffs Sven van Hal Mathieu Post Kasper Wendel Delft University of Technology Delft University of Technology Delft University of Technology [email protected] [email protected] [email protected] ABSTRACT be exploited by machine learning. The hypothesis is that methods Commit messages aid developers in their understanding of a con- based on machine learning, given enough training data, are able tinuously evolving codebase. However, developers not always doc- to extract more contextual information and latent factors about ument code changes properly. Automatically generating commit the why of a change. Furthermore, Allamanis et al. [1] state that messages would relieve this burden on developers. source code is “a form of human communication [and] has similar Recently, a number of different works have demonstrated the statistical properties to natural language corpora”. Following the feasibility of using methods from neural machine translation to success of (deep) machine learning in the field of natural language generate commit messages. This work aims to reproduce a promi- processing, neural networks seem promising for automated commit nent research paper in this field, as well as attempt to improve upon message generation as well. their results by proposing a novel preprocessing technique. Jiang et al. [12] have demonstrated that generating commit mes- A reproduction of the reference neural machine translation sages with neural networks is feasible. This work aims to reproduce model was able to achieve slightly better results on the same dataset. the results from [12] on the same and a different dataset. Addition- When applying more rigorous preprocessing, however, the per- ally, efforts are made to improve upon these results by applying a formance dropped significantly.
    [Show full text]
  • Introduction to Version Control with Git
    Warwick Research Software Engineering Introduction to Version Control with Git H. Ratcliffe and C.S. Brady Senior Research Software Engineers \The Angry Penguin", used under creative commons licence from Swantje Hess and Jannis Pohlmann. March 12, 2018 Contents 1 About these Notes1 2 Introduction to Version Control2 3 Basic Version Control with Git4 4 Releases and Versioning 11 Glossary 14 1 About these Notes These notes were written by H Ratcliffe and C S Brady, both Senior Research Software Engineers in the Scientific Computing Research Technology Platform at the University of Warwick for a series of Workshops first run in December 2017 at the University of Warwick. This document contains notes for a half-day session on version control, an essential part of the life of a software developer. This work, except where otherwise noted, is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International Li- cense. To view a copy of this license, visit http://creativecommons.org/ licenses/by-nc-nd/4.0/. The notes may redistributed freely with attribution, but may not be used for commercial purposes nor altered or modified. The Angry Penguin and other reproduced material, is clearly marked in the text and is not included in this declaration. The notes were typeset in LATEXby H Ratcliffe. Errors can be reported to [email protected] 1.1 Other Useful Information Throughout these notes, we present snippets of code and pseudocode, in particular snippets of commands for shell, make, or git. These often contain parts which you should substitute with the relevant text you want to use.
    [Show full text]
  • Higher Inductive Types (Hits) Are a New Type Former!
    Git as a HIT Dan Licata Wesleyan University 1 1 Darcs Git as a HIT Dan Licata Wesleyan University 1 1 HITs 2 Generator for 2 equality of equality HITs Homotopy Type Theory is an extension of Agda/Coq based on connections with homotopy theory [Hofmann&Streicher,Awodey&Warren,Voevodsky,Lumsdaine,Garner&van den Berg] 2 Generator for 2 equality of equality HITs Homotopy Type Theory is an extension of Agda/Coq based on connections with homotopy theory [Hofmann&Streicher,Awodey&Warren,Voevodsky,Lumsdaine,Garner&van den Berg] Higher inductive types (HITs) are a new type former! 2 Generator for 2 equality of equality HITs Homotopy Type Theory is an extension of Agda/Coq based on connections with homotopy theory [Hofmann&Streicher,Awodey&Warren,Voevodsky,Lumsdaine,Garner&van den Berg] Higher inductive types (HITs) are a new type former! They were originally invented[Lumsdaine,Shulman,…] to model basic spaces (circle, spheres, the torus, …) and constructions in homotopy theory 2 Generator for 2 equality of equality HITs Homotopy Type Theory is an extension of Agda/Coq based on connections with homotopy theory [Hofmann&Streicher,Awodey&Warren,Voevodsky,Lumsdaine,Garner&van den Berg] Higher inductive types (HITs) are a new type former! They were originally invented[Lumsdaine,Shulman,…] to model basic spaces (circle, spheres, the torus, …) and constructions in homotopy theory But they have many other applications, including some programming ones! 2 Generator for 2 equality of equality Patches Patch a a 2c2 diff b d = < b c c --- > d 3 3 id a a b b
    [Show full text]
  • Homework 0: Account Setup for Course and Cloud FPGA Intro Questions
    Cloud FPGA Homework 0 Fall 2019 Homework 0 Jakub Szefer 2019/10/20 Please follow the three setup sections to create BitBucket git repository, install LATEX tools or setup Overleaf account, and get access to the course's git repository. Once you have these done, answer the questions that follow. Submit your solutions as a single PDF file generated from a template; more information is at end in the Submission Instructions section. Setup BitBucket git Repository This course will use git repositories for code development. Each student should setup a free BitBucket (https://bitbucket.org) account and create a git repository for the course. Please make the repository private and give WRITE access to your instructor ([email protected]). Please send the URL address of the repository to the instructor by e-mail. Make sure there is a README:md file in the repository (access to the repository will be tested by a script that tries to download the README:md from the repository address you share). Also, if you are using a Apple computer, please add :gitignore file which contains one line: :DS Store (to prevent the hidden :DS Store files from accidentally being added to the repository). If you have problems accessing BitBucket git from the command line, please see the Appendix. Setup LATEX and Overleaf Any written work (including this homework's solutions) will be submitted as PDF files generated using LATEX [1] from provided templates. Students can setup a free Overleaf (https://www. overleaf.com) account to edit LATEX files and generate PDFs online; or students can install LATEX tools on their computer.
    [Show full text]
  • Version Control – Agile Workflow with Git/Github
    Version Control – Agile Workflow with Git/GitHub 19/20 November 2019 | Guido Trensch (JSC, SimLab Neuroscience) Content Motivation Version Control Systems (VCS) Understanding Git GitHub (Agile Workflow) References Forschungszentrum Jülich, JSC:SimLab Neuroscience 2 Content Motivation Version Control Systems (VCS) Understanding Git GitHub (Agile Workflow) References Forschungszentrum Jülich, JSC:SimLab Neuroscience 3 Motivation • Version control is one aspect of configuration management (CM). The main CM processes are concerned with: • System building • Preparing software for releases and keeping track of system versions. • Change management • Keeping track of requests for changes, working out the costs and impact. • Release management • Preparing software for releases and keeping track of system versions. • Version control • Keep track of different versions of software components and allow independent development. [Ian Sommerville,“Software Engineering”] Forschungszentrum Jülich, JSC:SimLab Neuroscience 4 Motivation • Keep track of different versions of software components • Identify, store, organize and control revisions and access to it • Essential for the organization of multi-developer projects is independent development • Ensure that changes made by different developers do not interfere with each other • Provide strategies to solve conflicts CONFLICT Alice Bob Forschungszentrum Jülich, JSC:SimLab Neuroscience 5 Content Motivation Version Control Systems (VCS) Understanding Git GitHub (Agile Workflow) References Forschungszentrum Jülich,
    [Show full text]
  • Scaling Git with Bitbucket Data Center
    Scaling Git with Bitbucket Data Center Considerations for large teams switching to Git Contents What is Git, why do I want it, and why is it hard to scale? 01 Scaling Git with Bitbucket Data Center 05 What about compliance? 11 Why choose Bitbucket Data Center? 13 01 What is Git, why do I want it, and why is it hard to scale? So. Your software team is expanding and taking on more high-value projects. That’s great news! The bad news, however, is that your centralized version control system isn’t really cutting it anymore. For growing IT organizations, Some of the key benefits Codebase safety moving to a distributed version control system is now of adopting Git are: Git is designed with maintaining the integrity considered an inevitable shift. This paper outlines some of managed source code as a top priority, using secure algorithms to preserve your code, change of the benefits of Git as a distributed version control system history, and traceability against both accidental and how Bitbucket Data Center can help your company scale and malicious change. Distributed development its Git-powered operations smoothly. Community Distributed development gives each developer a working copy of the full repository history, Git has become the expected version control making development faster by speeding up systems in many circles, and is very popular As software development increases in complexity, and the commit process and reducing developers’ among open source projects. This means its easy development teams become more globalized, centralized interdependence, as well as their dependence to take advantage of third party libraries and on a network connection.
    [Show full text]
  • Git Basics Git Expertise
    Overview Git basics Git expertise Git A GNU Alternative to Bitkeeper Mohamed Barakat University of Kaiserslautern ITWM Kaiserslautern, January 2010 Mohamed Barakat Git Overview Git basics Git expertise 1 Git basics The Git configuration file Create a Git-repository Using a Git-repository 2 Git expertise Branching Pulling Merging and Cherry-Picking Mohamed Barakat Git Overview The Git configuration file Git basics Create a Git-repository Git expertise Using a Git-repository Overview 1 Git basics The Git configuration file Create a Git-repository Using a Git-repository 2 Git expertise Branching Pulling Merging and Cherry-Picking Mohamed Barakat Git Overview The Git configuration file Git basics Create a Git-repository Git expertise Using a Git-repository Git working copy and Git-repsoitory Git is a distributed SCM Git is a distributed Source Code Management system (SCM), i.e. each Git working copy sits on top of its own local Git repository located in a single top-level subdirectory .git. SVN is in contrast to Git a centralized SCM system, i.e. the SVN-repository is on exactly one server. Git allows you to commit, checkout, reset, etc. without contacting a server! Mohamed Barakat Git Overview The Git configuration file Git basics Create a Git-repository Git expertise Using a Git-repository The configuration file ˜/.gitconfig Create a Git configuration file vi ˜/.gitconfig [svn] authorsfile = .git/info/svn-authors [color] diff = auto status = auto branch = auto [user] name = Firstname Lastname email = [email protected] Mohamed Barakat Git Overview The Git configuration file Git basics Create a Git-repository Git expertise Using a Git-repository Create a Git-repository (quick version) init, add, commit Create a Git-repository in the directory XY: 1 cd XY 2 git init 3 git add .
    [Show full text]
  • INF5750/9750 - Lecture 1 (Part III) Problem Area
    Revision control INF5750/9750 - Lecture 1 (Part III) Problem area ● Software projects with multiple developers need to coordinate and synchronize the source code Approaches to version control ● Work on same computer and take turns coding ○ Nah... ● Send files by e-mail or put them online ○ Lots of manual work ● Put files on a shared disk ○ Files get overwritten or deleted and work is lost, lots of direct coordination ● In short: Error prone and inefficient The preferred solution ● Use a revision control system. RCS - software that allows for multiple developers to work on the same codebase in a coordinated fashion ● History of Revision Control Systems: ○ File versioning tools, e.g. SCCS, RCS ○ Central Style - tree versioning tools. e.g. CVS ○ Central Style 2 - tree versioning tools e.g. SVN ○ Distributed style - tree versioning tools e.g. Bazaar ● Modern DVCS include Git, Mercurial, Bazaar Which system in this course? ● In this course we will be using GIT as the version control system ● We will use the UIO git system, but you can also make git accounts on github.com or bitbucket for your own projects ● DHIS2 uses a different system: Launchpad/Bazaar How it works Working tree: Local copy of the source code Repository: residing on the Central storage of developer’s the source code at computer (a client) a server synchronize synchronize Commit Commit locally Centralized De-centralized The repository Central ● Remembers every change ever written to it (called commits) ● You can have a central or local repository. ○ Central = big server in
    [Show full text]
  • Hg Mercurial Cheat Sheet Serge Y
    Hg Mercurial Cheat Sheet Serge Y. Stroobandt Copyright 2013–2020, licensed under Creative Commons BY-NC-SA #This page is work in progress! Much of the explanatory text still needs to be written. Nonetheless, the basic outline of this page may already be useful and this is why I am sharing it. In the mean time, please, bare with me and check back for updates. Distributed revision control Why I went with Mercurial • Python, Mozilla, Java, Vim • Mercurial has been better supported under Windows. • Mercurial also offers named branches Emil Sit: • August 2008: Mercurial offers a comfortable command-line experience, learning Git can be a bit daunting • December 2011: Git has three “philosophical” distinctions in its favour, as well as more attention to detail Lowest common denominator It is more important that people start using dis- tributed revision control instead of nothing at all. The Pro Git book is available online. Collaboration styles • Mercurial working practices • Collaborating with other people Use SSH shorthand 1 Installation $ sudo apt-get update $ sudo apt-get install mercurial mercurial-git meld Configuration Local system-wide configuration $ nano .bashrc export NAME="John Doe" export EMAIL="[email protected]" $ source .bashrc ~/.hgrc on a client user@client $ nano ~/.hgrc [ui] username = user@client editor = nano merge = meld ssh = ssh -C [extensions] convert = graphlog = mq = progress = strip = 2 ~/.hgrc on the server user@server $ nano ~/.hgrc [ui] username = user@server editor = nano merge = meld ssh = ssh -C [extensions] convert = graphlog = mq = progress = strip = [hooks] changegroup = hg update >&2 Initiating One starts with initiate a new repository.
    [Show full text]
  • Git/Hg Rosetta Stone
    Git/Hg Rosetta Stone hg git hg cat -r rev some_file git show rev:some_file hg clone http://hg.sympy.org/sympy-git.hg git clone git://git.sympy.org/sympy.git hg clone -U http://hg.sympy.org/sympy-git.hg git clone --bare git://git.sympy.org/sympy.git hg diff git diff HEAD hg status git status hg status -c git ls-files -t | grep '^H' hg manifest git ls-tree -r --name-only --full-tree HEAD hg parents git show --pretty=format:'%P' -s hg commit git commit -a hg record git add -p; git commit # or, for a more detailed interface: git add -i; git commit hg email -r tip git send-email HEAD^ # or: git format-patch HEAD^ ; git send-email 0001-Add-whitespace.patch hg view gitk, git gui hg help command git help command ~/.hgrc ~/.gitconfig .hg/hgrc .git/config hg paths git remote -v git remote add name url # see "git help remote" for more info how to edit paths; alternatively, you can edit them by hand in editing paths in .hg/hgrc .git/config too .hgignore .gitignore .hg/hgrc [ui] ignore .git/info/exclude hg add git add (note, it adds _content_ to index; can work on a hunk-by-hunk basis with -p!) hg rm git rm hg push git push hg pull git fetch hg pull -u git pull hg addremove git add -A (or: git add .; git ls-files --deleted xargs git rm) hg revert -a git reset --hard hg revert some_file git checkout some_file hg purge git clean -fd hg purge --all git clean -fdx git reset --hard 2fccd4c^ (on a normal repository) hg strip 2fccd4c git reset --soft 2fccd4c^ (on a bare repository) hg export git format-patch hg import --no-commit some.patch git apply some.patch hg import some.patch git am some.patch hg out git fetch && git log origin/master.
    [Show full text]
  • Git Cheat Sheet
    Git Cheat Sheet GIT BASICS REWRITING GIT HISTORY Create empty Git repo in specified directory. Run with no Replace the last commit with the staged changes and last commit git init git commit arguments to initialize the current directory as a git repository. combined. Use with nothing staged to edit the last commit’s message. <directory> --amend Clone repo located at onto local machine. Original repo can be <repo> Rebase the current branch onto <base>. <base> can be a commit ID, git clone <repo> located on the local filesystem or on a remote machine via or . git rebase <base> HTTP SSH branch name, a tag, or a relative reference to HEAD. Define author name to be used for all commits in current repo. Devs git config Show a log of changes to the local repository’s HEAD. commonly use flag to set config options for current user. git reflog Add flag to show date info or to show all refs. user.name <name> --global --relative-date --all git add Stage all changes in <directory> for the next commit. GIT BRANCHES <directory> Replace <directory> with a <file> to change a specific file. Commit the staged snapshot, but instead of launching git commit -m List all of the branches in your repo. Add a <branch> argument to a text editor, use as the commit message. git branch "<message>" <message> create a new branch with the name <branch>. git checkout -b Create and check out a new branch named <branch>. git status List which files are staged, unstaged, and untracked. <branch> Drop the -b flag to checkout an existing branch.
    [Show full text]
  • Everything You Need to Know About Openjdk's Move to Git and Github
    Menu Topics Archives Downloads Subscribe Everything you need to know JAVA 17 about OpenJDK’s move to Git and GitHub Everything you need to know Blame or thank BitKeeper about OpenJDK’s move to Git Why not Mercurial? and GitHub Why Git? Why GitHub? Why the move, and why now? The move from Mercurial to Git Getting the source code and provided an opportunity to consolidate building the OpenJDK the source code repositories. Conclusion by Ian Darwin Dig deeper May 14, 2021 Download a PDF of this article Have you ever built your own Java Development Kit from source? Most end users of the JDK will not need to build their own JDK from the Oracle source code. I’ve needed to do that only a few times when I was running on the OpenBSD UNIX-like system, which is not one of the three supported platforms. Sure, you might want to build your own JDK to try out a new feature that you think should be added to Java. You might choose to build from source to be sure you are running a more trustworthy binary. Having the complete source code readily available, and now in a more commonly used download format, means it is easier than ever to build your own JDK. Yes, it’s a better-documented, easily configured process than in the past. But it’s still a bit confusing. The source code for the OpenJDK recently moved from the Mercurial version control system (VCS) to the Git VCS and the GitHub repository system, and that’s probably a good thing.
    [Show full text]