threads / discuss / 24133

SVN migration

Subject: SVN migration

## tl;dr

10 messages between Jun 16, 2010 and Jul 4, 2010.

replies: 9people: 4as markdown or json

William Hall· Jun 16, 2010, 23:02 UTC · lore
Hi gitters,

Background - I'm trying to convince my company to ditch SVN for git - the usual story. So for the duration of a project I'll be running git and SVN in parallel - the idea being is that we will still commit to SVN (and update), but the development work internal to my team will be using git.

An absolute *must* is for the SVN repo to continue as the SCM authority 
- at least until I can persuade the company to switch to git permanently.
Here's some crap ascii art to show the situation,
--------------
| non-git dev |
--------------
      |
      |
   -------         ----------------
  |  SVN  |-------| git/SVN bridge |
   -------         ----------------
                         |
                   ---------------
                  | bare git Repo |
                   ---------------
                         |
          ------------------------------
          |              |              |
        dev_1          dev_2          dev 3
- the git/SVN bridge is a git repo created with git-svn-clone.
- the 'bare git' repo is a typical standard git repo and I'm keen for 
the developers to experience a 'normal' git environment and not have to 
worry about SVN interactions.

Am sure this problem has been considered many times before, but I cannot seem to find an effective solution.

The issue is the dcommit operation from the bridge. The rebase part of this re-writes the commit messages to include the SVN commit-ids which is nice, but screws up the push/pulls between the bridge and the bare repo.

One solution I've tried is to create a branch in the bridge that tracks the bare repo, and another branch to track the SVN server. If the branches are kept separate then I can git-cherry-pick to replay changes from one side to the other (or at least merge one-way). his is not ideal as I really should use git's merge facility. I'd like to guarantee that the sides are not diverging over time.

Actually I've tried all permutations of merges/rebases/update-ref, I always fall into the same trap that befits a rebase in conjunction with remote repositories.

I can live without tags and branches for the time being - I just want to get a robust workflow defined in the bridge for the SVN 'trunk' - ie read/writes in both directions.

If anyone can offer any advice then it would be hugely appreciated. Perhaps you'll say that it cannot be done, which would make the git sell much harder.

Hopefully by the end of this exercise git will have 800 more fans.
Many thanks
Will
Steven Michalske· Jun 17, 2010, 00:41 UTC · re: William Hall · lore

Re: SVN migration

On Jun 16, 2010, at 4:02 PM, William Hall wrote:
> 
> The issue is the dcommit operation from the bridge. The rebase part of this re-writes the commit messages to include the SVN commit-ids which is nice, but screws up the push/pulls between the bridge and the bare repo.
Look into svn.noMetadata configuration option.  It will prevent you from rebuilding the svn to git bridge if something seriously goes wrong, but it prevents the messages from changing.

svn-remote.<name>.noMetadata This gets rid of the git-svn-id: lines at the end of every commit. If you lose your .git/svn/git-svn/.rev_db file, git svn will not be able to rebuild it and you won't be able to fetch again, either. This is fine for one-shot imports. The git svn log command will not work on repositories using this, either. Using this conflicts with the useSvmProps option for (hopefully) obvious reasons

William Hall· Jun 17, 2010, 10:33 UTC · re: Steven Michalske · lore

Re: SVN migration

Thanks Steven,

The noMetadata option will prevent me from doing anything other than a one-shot import, which is not what I want. I need to somehow devise a workflow that allows me bidirectional push/pull between an svn repo and a remote git repo.

Steven Michalske wrote:
Show 10 quoted lines
> On Jun 16, 2010, at 4:02 PM, William Hall wrote:
> 
>> The issue is the dcommit operation from the bridge. The rebase part of this re-writes the commit messages to include the SVN commit-ids which is nice, but screws up the push/pulls between the bridge and the bare repo.
> 
> Look into svn.noMetadata configuration option.  It will prevent you from rebuilding the svn to git bridge if something seriously goes wrong, but it prevents the messages from changing.
> 
> svn-remote.<name>.noMetadata
> This gets rid of the git-svn-id: lines at the end of every commit.
> If you lose your .git/svn/git-svn/.rev_db file, git svn will not be able to rebuild it and you won't be able to fetch again, either. This is fine for one-shot imports.
> The git svn log command will not work on repositories using this, either. Using this conflicts with the useSvmProps option for (hopefully) obvious reasons
William Hall· Jun 17, 2010, 16:27 UTC · re: William Hall · lore

Re: SVN migration

I'm going to answer my own post, I *think* I have something that works - please check if I'm doing anything idiotic.

Just to recap: trying to convince my company to move from svn to git, and they have agreed to try one project using git as long as all commits find their way to the svn repo as well.

So I have a standard bare git repo serving the developers, a git/svn "bridge" repo that performs bi-directional updates to and from svn and the bare git repo.

Ok, here's what I've done

Create bridge ------------- $ git svn init -s file:///path_to_svn /path_to_git_svn_bridge/ $ cd /path_to_git_svn_bridge $ git svn fetch --authors-file=/tmp/authors.map

Configure bare repo ------------------------------- create bare repo that developers will use $ git init --shared=all --bare /path_to_git_repo.git

configure bridge ---------------- $ git remote add -f -m master origin /path_to_git_repo.git $ git push origin master $ git branch --set-upstream master origin/master

this branch will be used to perform svn rebases and fetches $ git checkout -t -b svn svn/trunk

Workflow -------- Developer A clones from /path_to_git_repo.git, does some work, commits and pushes back to origin

Now, in the bridge repo, fetch changes from origin (where developer A pushed) $ git checkout master $ git pull

Replay all changes manually, in order, onto svn branch
$ git checkout svn
$ git rev-list --reverse heads/master@{1}..heads/master | while read rev; do
         git cherry-pick -n $rev
   done

Create one commit for all changes and synchronise with svn $ git commit -am "cherry pick merge" $ git svn rebase $ git svn dcommit

Now merge in anything picked up from svn, plus the rebased final commit $ git checkout master $ git merge svn

Send back to bare repo (at least the final merge commit) $ git push

It seems to handle changes and preserves linear history on both sides ok. Can anyone see anything obviously wrong with this approach?

thanks,
Will
William Hall wrote:
 > Thanks Steven,
 >
 > The noMetadata option will prevent me from doing anything other than 
a one-shot import, which is not what I want. I need to somehow devise a 
workflow that allows me bidirectional push/pull between an svn repo and 
a remote git repo.
 >
 >
 >
 > Steven Michalske wrote:
 >> On Jun 16, 2010, at 4:02 PM, William Hall wrote:
 >>
 >>> The issue is the dcommit operation from the bridge. The rebase part 
of this re-writes the commit messages to include the SVN commit-ids 
which is nice, but screws up the push/pulls between the bridge and the 
bare repo.
 >>
 >> Look into svn.noMetadata configuration option.  It will prevent you 
from rebuilding the svn to git bridge if something seriously goes wrong, 
but it prevents the messages from changing.
 >>
 >> svn-remote.<name>.noMetadata
 >> This gets rid of the git-svn-id: lines at the end of every commit.
 >> If you lose your .git/svn/git-svn/.rev_db file, git svn will not be 
able to rebuild it and you won't be able to fetch again, either. This is 
fine for one-shot imports.
 >> The git svn log command will not work on repositories using this, 
either. Using this conflicts with the useSvmProps option for (hopefully) 
obvious reasons
 > --
 > To unsubscribe from this list: send the line "unsubscribe git" in
 > the body of a message to majordomo@vger.kernel.org
 > More majordomo info at  http://vger.kernel.org/majordomo-info.html
William Hall wrote:
Show 33 quoted lines
> Thanks Steven,
> 
> The noMetadata option will prevent me from doing anything other than a 
> one-shot import, which is not what I want. I need to somehow devise a 
> workflow that allows me bidirectional push/pull between an svn repo and 
> a remote git repo.
> 
> 
> 
> Steven Michalske wrote:
>> On Jun 16, 2010, at 4:02 PM, William Hall wrote:
>>
>>> The issue is the dcommit operation from the bridge. The rebase part 
>>> of this re-writes the commit messages to include the SVN commit-ids 
>>> which is nice, but screws up the push/pulls between the bridge and 
>>> the bare repo.
>>
>> Look into svn.noMetadata configuration option.  It will prevent you 
>> from rebuilding the svn to git bridge if something seriously goes 
>> wrong, but it prevents the messages from changing.
>>
>> svn-remote.<name>.noMetadata
>> This gets rid of the git-svn-id: lines at the end of every commit.
>> If you lose your .git/svn/git-svn/.rev_db file, git svn will not be 
>> able to rebuild it and you won't be able to fetch again, either. This 
>> is fine for one-shot imports.
>> The git svn log command will not work on repositories using this, 
>> either. Using this conflicts with the useSvmProps option for 
>> (hopefully) obvious reasons
> -- 
> To unsubscribe from this list: send the line "unsubscribe git" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
Joshua Shrader· Jun 21, 2010, 21:12 UTC · re: William Hall · lore

Re: SVN migration

In Jon Loeliger's "Version Control with Git", Chapter 16, he describes a similar situation in which there is a Subversion repository, and at least a couple users that want to be using Git. He proposes a single "gatekeeper" git repository, what you refer to as a bridge, which is the only interface to subversion. After git svn cloneing the subversion repo (with --prefix=svn/), all the branches are then pushed to a bare repository (git push ../svn-bare.git 'refs/remotes/svn/*:refs/heads/svn/*', and other git users are told to clone this repo, which now contains local branches of all the svn remotes. Then, to merge back to subversion, in the gatekeeper repo, you do

git checkout svn/trunk (or other branch - this is checking out a detached head as svn/trunk is a remote) git merge --no-ff new-feature git svn dcommit

This results in a merge commit on a detached head, and then the modified commit (after the git-svn-id line is added) is put on the real svn/trunk branch. The commit on the detached head is "worse than redundant. Using it for anything else eventually results in conflicts. So, just forget about that commit. If you haven't put it on a branch in the first place, it's that much easier to forget" (Jon Loeliger).

I haven't tried this yet, but I'm in a similar situation in that I'm trying to convince my project to convert to using Git. We're going to use this gatekeeper approach for a while, and allow users to migrate over at their own discretion. Then hopefully, if there's not too much resistance, we'll get rid of the subversion repo entirely.

If there's a problem with this workflow, I'd love to hear about it. I'm only in the middle of setting this up, but hopefully I should know if it works by the end of the week.

On Thu, Jun 17, 2010 at 12:27 PM, William Hall <will@gnatter.net> wrote:
Show 151 quoted lines
> I'm going to answer my own post, I *think* I have something that works -
> please check if I'm doing anything idiotic.
>
> Just to recap: trying to convince my company to move from svn to git, and
> they have agreed to try one project using git as long as all commits find
> their way to the svn repo as well.
>
> So I have a standard bare git repo serving the developers, a git/svn
> "bridge" repo that performs bi-directional updates to and from svn and the
> bare git repo.
>
> Ok, here's what I've done
>
> Create bridge
> -------------
> $ git svn init -s file:///path_to_svn /path_to_git_svn_bridge/
> $ cd /path_to_git_svn_bridge
> $ git svn fetch --authors-file=/tmp/authors.map
>
> Configure bare repo
> -------------------------------
> create bare repo that developers will use
> $ git init --shared=all --bare /path_to_git_repo.git
>
> configure bridge
> ----------------
> $ git remote add -f -m master origin /path_to_git_repo.git
> $ git push origin master
> $ git branch --set-upstream master origin/master
>
> this branch will be used to perform svn rebases and fetches
> $ git checkout -t -b svn svn/trunk
>
>
> Workflow
> --------
> Developer A clones from /path_to_git_repo.git, does some work, commits and
> pushes back to origin
>
> Now, in the bridge repo, fetch changes from origin (where developer A
> pushed)
> $ git checkout master
> $ git pull
>
> Replay all changes manually, in order, onto svn branch
> $ git checkout svn
> $ git rev-list --reverse heads/master@{1}..heads/master | while read rev; do
>        git cherry-pick -n $rev
>  done
>
> Create one commit for all changes and synchronise with svn
> $ git commit -am "cherry pick merge"
> $ git svn rebase
> $ git svn dcommit
>
> Now merge in anything picked up from svn, plus the rebased final commit
> $ git checkout master
> $ git merge svn
>
> Send back to bare repo (at least the final merge commit)
> $ git push
>
> It seems to handle changes and preserves linear history on both sides ok.
> Can anyone see anything obviously wrong with this approach?
>
> thanks,
>
> Will
>
>
>
>
>
>
> William Hall wrote:
>> Thanks Steven,
>>
>> The noMetadata option will prevent me from doing anything other than a
>> one-shot import, which is not what I want. I need to somehow devise a
>> workflow that allows me bidirectional push/pull between an svn repo and a
>> remote git repo.
>>
>>
>>
>> Steven Michalske wrote:
>>> On Jun 16, 2010, at 4:02 PM, William Hall wrote:
>>>
>>>> The issue is the dcommit operation from the bridge. The rebase part of
>>>> this re-writes the commit messages to include the SVN commit-ids which is
>>>> nice, but screws up the push/pulls between the bridge and the bare repo.
>>>
>>> Look into svn.noMetadata configuration option.  It will prevent you from
>>> rebuilding the svn to git bridge if something seriously goes wrong, but it
>>> prevents the messages from changing.
>>>
>>> svn-remote.<name>.noMetadata
>>> This gets rid of the git-svn-id: lines at the end of every commit.
>>> If you lose your .git/svn/git-svn/.rev_db file, git svn will not be able
>>> to rebuild it and you won't be able to fetch again, either. This is fine for
>>> one-shot imports.
>>> The git svn log command will not work on repositories using this, either.
>>> Using this conflicts with the useSvmProps option for (hopefully) obvious
>>> reasons
>> --
>> To unsubscribe from this list: send the line "unsubscribe git" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
>
> William Hall wrote:
>>
>> Thanks Steven,
>>
>> The noMetadata option will prevent me from doing anything other than a
>> one-shot import, which is not what I want. I need to somehow devise a
>> workflow that allows me bidirectional push/pull between an svn repo and a
>> remote git repo.
>>
>>
>>
>> Steven Michalske wrote:
>>>
>>> On Jun 16, 2010, at 4:02 PM, William Hall wrote:
>>>
>>>> The issue is the dcommit operation from the bridge. The rebase part of
>>>> this re-writes the commit messages to include the SVN commit-ids which is
>>>> nice, but screws up the push/pulls between the bridge and the bare repo.
>>>
>>> Look into svn.noMetadata configuration option.  It will prevent you from
>>> rebuilding the svn to git bridge if something seriously goes wrong, but it
>>> prevents the messages from changing.
>>>
>>> svn-remote.<name>.noMetadata
>>> This gets rid of the git-svn-id: lines at the end of every commit.
>>> If you lose your .git/svn/git-svn/.rev_db file, git svn will not be able
>>> to rebuild it and you won't be able to fetch again, either. This is fine for
>>> one-shot imports.
>>> The git svn log command will not work on repositories using this, either.
>>> Using this conflicts with the useSvmProps option for (hopefully) obvious
>>> reasons
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe git" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
> --
> To unsubscribe from this list: send the line "unsubscribe git" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
William Hall· Jun 21, 2010, 22:26 UTC · re: Joshua Shrader · lore

Re: SVN migration

Hi Joshua - glad I'm not the only one facing this conundrum.

Yes, I've followed the Jon Loeliger's guide and am using something very similar.

I've come up with a couple of solutions, but this is my latest effort. The only real difference in our situations is that I wanted the developers to get a feel of a bog-standard bare git repo to which they pull/push as usual - the gateway (...much better name) will also push and pull to this.

I've created a post-update hook in the bare git repo that does the following on the gateway repo

( there's a branch "svn" that tracks refs/remotes/svn/trunk)
- pulls commits from bare repo into master (all current dev work)
- on svn branch, run "git merge --squash master"
- commit changes
- git svn rebase (pick up changes from other depts)
- git svn dcommit
- switch back to master and merge in svn changes (with --no-ff)
- git push, back to bare repo

All git commits between pushes are squashed which probably represents the frequency of usual svn commits (ie, only when something is finished) so it's probably ok. Also, it prevents the developers from seeing duplicate commit messages - one from master and a duplicate (with svn commit-id) which may get annoying or confusing.

I've also tried doing a rebase with --onto to replay all changes to master onto the svn branch, which is ok, but all commits are duplicated and propagated.

I'm keen to merge the branches at some stage otherwise I'll end up with two branches on the gateway that never converge - which is probably not a good idea.

Have tried (briefly) your solution which also seems to work, but without any merge commits - will try more.

Naturally, I'd like a completely linear history, but the does not seem viable. I don't think git will be hard sell (most of our developers seem to agree that git is the future), but I need to get this right from the outset or confidence will be lost.

will
On 21/06/10 22:12, Joshua Shrader wrote:
Show 187 quoted lines
> In Jon Loeliger's "Version Control with Git", Chapter 16, he describes
> a similar situation in which there is a Subversion repository, and at
> least a couple users that want to be using Git.  He proposes a single
> "gatekeeper" git repository, what you refer to as a bridge, which is
> the only interface to subversion.  After git svn cloneing the
> subversion repo (with --prefix=svn/), all the branches are then pushed
> to a bare repository (git push ../svn-bare.git
> 'refs/remotes/svn/*:refs/heads/svn/*', and other git users are told to
> clone this repo, which now contains local branches of all the svn
> remotes.  Then, to merge back to subversion, in the gatekeeper repo,
> you do
>
> git checkout svn/trunk (or other branch - this is checking out a
> detached head as svn/trunk is a remote)
> git merge --no-ff new-feature
> git svn dcommit
>
> This results in a merge commit on a detached head, and then the
> modified commit (after the git-svn-id line is added) is put on the
> real svn/trunk branch.  The commit on the detached head is "worse than
> redundant.  Using it for anything else eventually results in
> conflicts.  So, just forget about that commit.  If you haven't put it
> on a branch in the first place, it's that much easier to forget" (Jon
> Loeliger).
>
> I haven't tried this yet, but I'm in a similar situation in that I'm
> trying to convince my project to convert to using Git.  We're going to
> use this gatekeeper approach for a while, and allow users to migrate
> over at their own discretion.  Then hopefully, if there's not too much
> resistance, we'll get rid of the subversion repo entirely.
>
> If there's a problem with this workflow, I'd love to hear about it.
> I'm only in the middle of setting this up, but hopefully I should know
> if it works by the end of the week.
>
> On Thu, Jun 17, 2010 at 12:27 PM, William Hall<will@gnatter.net>  wrote:
>> I'm going to answer my own post, I *think* I have something that works -
>> please check if I'm doing anything idiotic.
>>
>> Just to recap: trying to convince my company to move from svn to git, and
>> they have agreed to try one project using git as long as all commits find
>> their way to the svn repo as well.
>>
>> So I have a standard bare git repo serving the developers, a git/svn
>> "bridge" repo that performs bi-directional updates to and from svn and the
>> bare git repo.
>>
>> Ok, here's what I've done
>>
>> Create bridge
>> -------------
>> $ git svn init -s file:///path_to_svn /path_to_git_svn_bridge/
>> $ cd /path_to_git_svn_bridge
>> $ git svn fetch --authors-file=/tmp/authors.map
>>
>> Configure bare repo
>> -------------------------------
>> create bare repo that developers will use
>> $ git init --shared=all --bare /path_to_git_repo.git
>>
>> configure bridge
>> ----------------
>> $ git remote add -f -m master origin /path_to_git_repo.git
>> $ git push origin master
>> $ git branch --set-upstream master origin/master
>>
>> this branch will be used to perform svn rebases and fetches
>> $ git checkout -t -b svn svn/trunk
>>
>>
>> Workflow
>> --------
>> Developer A clones from /path_to_git_repo.git, does some work, commits and
>> pushes back to origin
>>
>> Now, in the bridge repo, fetch changes from origin (where developer A
>> pushed)
>> $ git checkout master
>> $ git pull
>>
>> Replay all changes manually, in order, onto svn branch
>> $ git checkout svn
>> $ git rev-list --reverse heads/master@{1}..heads/master | while read rev; do
>>         git cherry-pick -n $rev
>>   done
>>
>> Create one commit for all changes and synchronise with svn
>> $ git commit -am "cherry pick merge"
>> $ git svn rebase
>> $ git svn dcommit
>>
>> Now merge in anything picked up from svn, plus the rebased final commit
>> $ git checkout master
>> $ git merge svn
>>
>> Send back to bare repo (at least the final merge commit)
>> $ git push
>>
>> It seems to handle changes and preserves linear history on both sides ok.
>> Can anyone see anything obviously wrong with this approach?
>>
>> thanks,
>>
>> Will
>>
>>
>>
>>
>>
>>
>> William Hall wrote:
>>> Thanks Steven,
>>>
>>> The noMetadata option will prevent me from doing anything other than a
>>> one-shot import, which is not what I want. I need to somehow devise a
>>> workflow that allows me bidirectional push/pull between an svn repo and a
>>> remote git repo.
>>>
>>>
>>>
>>> Steven Michalske wrote:
>>>> On Jun 16, 2010, at 4:02 PM, William Hall wrote:
>>>>
>>>>> The issue is the dcommit operation from the bridge. The rebase part of
>>>>> this re-writes the commit messages to include the SVN commit-ids which is
>>>>> nice, but screws up the push/pulls between the bridge and the bare repo.
>>>>
>>>> Look into svn.noMetadata configuration option.  It will prevent you from
>>>> rebuilding the svn to git bridge if something seriously goes wrong, but it
>>>> prevents the messages from changing.
>>>>
>>>> svn-remote.<name>.noMetadata
>>>> This gets rid of the git-svn-id: lines at the end of every commit.
>>>> If you lose your .git/svn/git-svn/.rev_db file, git svn will not be able
>>>> to rebuild it and you won't be able to fetch again, either. This is fine for
>>>> one-shot imports.
>>>> The git svn log command will not work on repositories using this, either.
>>>> Using this conflicts with the useSvmProps option for (hopefully) obvious
>>>> reasons
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe git" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
>>
>> William Hall wrote:
>>>
>>> Thanks Steven,
>>>
>>> The noMetadata option will prevent me from doing anything other than a
>>> one-shot import, which is not what I want. I need to somehow devise a
>>> workflow that allows me bidirectional push/pull between an svn repo and a
>>> remote git repo.
>>>
>>>
>>>
>>> Steven Michalske wrote:
>>>>
>>>> On Jun 16, 2010, at 4:02 PM, William Hall wrote:
>>>>
>>>>> The issue is the dcommit operation from the bridge. The rebase part of
>>>>> this re-writes the commit messages to include the SVN commit-ids which is
>>>>> nice, but screws up the push/pulls between the bridge and the bare repo.
>>>>
>>>> Look into svn.noMetadata configuration option.  It will prevent you from
>>>> rebuilding the svn to git bridge if something seriously goes wrong, but it
>>>> prevents the messages from changing.
>>>>
>>>> svn-remote.<name>.noMetadata
>>>> This gets rid of the git-svn-id: lines at the end of every commit.
>>>> If you lose your .git/svn/git-svn/.rev_db file, git svn will not be able
>>>> to rebuild it and you won't be able to fetch again, either. This is fine for
>>>> one-shot imports.
>>>> The git svn log command will not work on repositories using this, either.
>>>> Using this conflicts with the useSvmProps option for (hopefully) obvious
>>>> reasons
>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe git" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe git" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
William Hall· Jun 26, 2010, 10:33 UTC · re: William Hall · lore

Re: SVN migration

http://github.com/innerhippy/svnAndGit

I've created a new project on github called svnAndGit that attempts to create a bi-directional workflow to enable git and SVN to live side-by-side.

It's not perfect by any means, so all comments most welcome!
David Bainbridge· Jul 3, 2010, 11:37 UTC · re: William Hall · lore

Re: SVN migration

Hi William,

I have been following this thread with interest so I thought that I would just throw in my thoughts!

While maintaining synchronization with Git is part of what is needed I suspect that this will not entirely convince the management of your company that Git is the way forward.

They probably see Svn as a safe repository ... The company's assets (intellectual property) are on a central server that is backed up, and the contents of that repository can be audited and so on. They may be thinking about things like SOX compliance too.

So if you want them to accept Git as a replacement for svn then you need to understand and address these concerns. This means that you will have to have a conversation with them. To a large extent this a people thing ... technical solutions won't necessary convince them. They are running a company based on the knowledge and information they own - and they want to make sure that it doesn't get lost, stolen, corrupted, or whatever. And they are accountable to the shareholders for this.

Also, you say that they have been using Svn for donkey's years, so from a corporate perspective it probably does what they want and need. Otherwise THEY would have decided to change it.

I am in a similar situation and while developers clearly want to use gIt, the motivation from a corporate perspective is less clear and can be perceived as introducing risk. So we are looking at the wya in which repositories are set up, the topology of git repository networks, use of Gitosis. Gitolite and Gitorious, and so on, to provide some security in the corporate environment.

Every company will have a different view of this so there is no 'right' answer. A lot depends on the type of product you produce and how long it will need to be supported. If you have products that need to be supported for 10 years or more then promoting a tool that is 5 years old may also raise some eyebrows! You need to have the answers ready :-)

Get it right and you will be seen as a hero who understands the business. Get it wrong and you will consigned to the religious nerd category who just wants to promote his favourite tool ... which I would hope is not the case :-)

Good luck with this ... you are not alone!
Dave Bainbridge
William Hall· Jul 4, 2010, 17:55 UTC · re: David Bainbridge · lore

Re: SVN migration

Hi David, Thanks for your thoughts!

I agree with your points. To an extent, management don't really care how we implement SCM - as long as it's effective and secure they will trust the "tech-wranglers" to do the right thing and not impede upon the company's workflow. Fortunately the industry in which I work is VFX, so "cutting edge" software is at the core of what we do. I am not imposing Git because of my own personal preference, I honestly believe that SVN is simply not the right tool for the job - which will increasingly involve multi-site collaboration that spans departments as well as timezones. The ability for two disparate teams of developers to collaborate effectively without polluting the global codebase is essential.

The limitations with SVN are becoming more and more apparent - especially now that we have now embarked upon a fairly radical shake-up of our existing software stack.

I have explained all this with senior management, some who have heard of Git (and its reputation) and they pretty much say "about time too".

The hard part is that we have two tiers of developers - core software techies (C++, python) and scripters (python, MEL - these are the people who make VFX movies, for example, happen). The former will have no problem with Git, the latter probably just don't care - they just want to check stuff in and out.)

What I need to do is create this hybrid system that enables the scripters to pretty much carry on as usual, and to provide the necessary tools to do SCM more effectively - ie without the overhead of a brittle SVN environment. If all goes well, we'll take the plunge and make the switch permanent.

Yes, the technical sell for Git is the easy part, the cultural sell will be harder. It's up to me to make the business case to the bean-counters and make the technical transition painless. So far, so good.

I've posted this before, the scripts I am using are available at -
http://github.com/innerhippy/svnAndGit
The more eyes on this the better...
Cheers
Will
On 03/07/10 12:37, David Bainbridge wrote:
Show 49 quoted lines
> Hi William,
>
> I have been following this thread with interest so I thought that I
> would just throw in my thoughts!
>
> While maintaining synchronization with Git is part of what is needed I
> suspect that this will not entirely convince the management of your
> company that Git is the way forward.
>
> They probably see Svn as a safe repository ... The company's assets
> (intellectual property) are on a central server that is backed up, and
> the contents of that repository can be audited and so on. They may be
> thinking about things like SOX compliance too.
>
> So if you want them to accept Git as a replacement for svn then you
> need to understand and address these concerns. This means that you
> will have to have a conversation with them. To a large extent this a
> people thing ... technical solutions won't necessary convince them.
> They are running a company based on the knowledge and information they
> own - and they want to make sure that it doesn't get lost, stolen,
> corrupted, or whatever. And they are accountable to the shareholders
> for this.
>
> Also, you say that they have been using Svn for donkey's years, so
> from a corporate perspective it probably does what they want and need.
> Otherwise THEY would have decided to change it.
>
> I am in a similar situation and while developers clearly want to use
> gIt, the motivation from a corporate perspective is less clear and can
> be perceived as introducing risk. So we are looking at the wya in
> which repositories are set up, the topology of git repository
> networks, use of Gitosis. Gitolite and Gitorious, and so on, to
> provide some security in the corporate environment.
>
> Every company will have a different view of this so there is no
> 'right' answer. A lot depends on the type of product you produce and
> how long it will need to be supported. If you have products that need
> to be supported for 10 years or more then promoting a tool that is 5
> years old may also raise some eyebrows! You need to have the answers
> ready :-)
>
> Get it right and you will be seen as a hero who understands the
> business. Get it wrong and you will consigned to the religious nerd
> category who just wants to promote his favourite tool ... which I
> would hope is not the case :-)
>
> Good luck with this ... you are not alone!
>
> Dave Bainbridge
David Bainbridge· Jul 4, 2010, 22:01 UTC · re: William Hall · lore

Re: SVN migration

Hi Will,
You seem to have all the bases covered :-)

It seems like they have been continuing to use SVN just because it was there, and there was no one willing or able to take them somewhere else, even when the business was changing around them.

The geographical distribution is an interesting one, because somehow someone needs to understand what progress is being made with the product development. With Git it can be easy for people to hide their 'dirty little secrets' for far too long. This doesn't really matter in open source development, but is an issue for in a commercial situation.

I tend to prefer a central repository and then judge progress by what is there. If it isn't there it hasn't been done ...

Unfortunately Git books are not so hot on this aspect of deploying Git :-(
Maybe there are some other ideas out there ...
All the best,

David Bainbridge Sweden

On 4 July 2010 19:55, William Hall <will@gnatter.net> wrote:
Show 101 quoted lines
> Hi David,
> Thanks for your thoughts!
>
> I agree with your points. To an extent, management don't really care how we
> implement SCM - as long as it's effective and secure they will trust the
> "tech-wranglers" to do the right thing and not impede upon the company's
> workflow. Fortunately the industry in which I work is VFX, so "cutting edge"
> software is at the core of what we do. I am not imposing Git because of my
> own personal preference, I honestly believe that SVN is simply not the right
> tool for the job - which will increasingly involve multi-site collaboration
> that spans departments as well as timezones. The ability for two disparate
> teams of developers to collaborate effectively without polluting the global
> codebase is essential.
>
> The limitations with SVN are becoming more and more apparent - especially
> now that we have now embarked upon a fairly radical shake-up of our existing
> software stack.
>
> I have explained all this with senior management, some who have heard of Git
> (and its reputation) and they pretty much say "about time too".
>
> The hard part is that we have two tiers of developers - core software
> techies (C++, python) and scripters (python, MEL - these are the people who
> make VFX movies, for example, happen). The former will have no problem with
> Git, the latter probably just don't care - they just want to check stuff in
> and out.)
>
> What I need to do is create this hybrid system that enables the scripters to
> pretty much carry on as usual, and to provide the necessary tools to do SCM
> more effectively - ie without the overhead of a brittle SVN environment. If
> all goes well, we'll take the plunge and make the switch permanent.
>
> Yes, the technical sell for Git is the easy part, the cultural sell will be
> harder. It's up to me to make the business case to the bean-counters and
> make the technical transition painless. So far, so good.
>
> I've posted this before, the scripts I am using are available at -
>
> http://github.com/innerhippy/svnAndGit
>
> The more eyes on this the better...
>
> Cheers
>
> Will
>
>
>
>
> On 03/07/10 12:37, David Bainbridge wrote:
>>
>> Hi William,
>>
>> I have been following this thread with interest so I thought that I
>> would just throw in my thoughts!
>>
>> While maintaining synchronization with Git is part of what is needed I
>> suspect that this will not entirely convince the management of your
>> company that Git is the way forward.
>>
>> They probably see Svn as a safe repository ... The company's assets
>> (intellectual property) are on a central server that is backed up, and
>> the contents of that repository can be audited and so on. They may be
>> thinking about things like SOX compliance too.
>>
>> So if you want them to accept Git as a replacement for svn then you
>> need to understand and address these concerns. This means that you
>> will have to have a conversation with them. To a large extent this a
>> people thing ... technical solutions won't necessary convince them.
>> They are running a company based on the knowledge and information they
>> own - and they want to make sure that it doesn't get lost, stolen,
>> corrupted, or whatever. And they are accountable to the shareholders
>> for this.
>>
>> Also, you say that they have been using Svn for donkey's years, so
>> from a corporate perspective it probably does what they want and need.
>> Otherwise THEY would have decided to change it.
>>
>> I am in a similar situation and while developers clearly want to use
>> gIt, the motivation from a corporate perspective is less clear and can
>> be perceived as introducing risk. So we are looking at the wya in
>> which repositories are set up, the topology of git repository
>> networks, use of Gitosis. Gitolite and Gitorious, and so on, to
>> provide some security in the corporate environment.
>>
>> Every company will have a different view of this so there is no
>> 'right' answer. A lot depends on the type of product you produce and
>> how long it will need to be supported. If you have products that need
>> to be supported for 10 years or more then promoting a tool that is 5
>> years old may also raise some eyebrows! You need to have the answers
>> ready :-)
>>
>> Get it right and you will be seen as a hero who understands the
>> business. Get it wrong and you will consigned to the religious nerd
>> category who just wants to promote his favourite tool ... which I
>> would hope is not the case :-)
>>
>> Good luck with this ... you are not alone!
>>
>> Dave Bainbridge
>

← back to recent threads