2

The company I'm working for is using a third-party web application (PHP/MySQL) which can be extended by modules. We need to track our own changes to the core application and to additional modules, keep track of which modules are installed and at the same time maintain the ability to easily update the application and its modules to the latest vendor version.

At this time we're using one git repository containing everything. To maintain updatability we're using a vendor branch strategy. This works fine for the main app, but pretty soon becomes messy when multiple modules enter the stage.

However I was not able to find a feasible way to manage such a project with git, especially because both git's submodules and git's subtree merging seem to require a subproject to be contained in one directory. But that's not true for most web applications I've seen, as it is with this one which has following directory structure:

app/
  `-code/
      `-path/to/plugin/code
  `--i18n/
      `-path/to/plugin/i18n
  `--assets/
      `-path/to/plugin/assets

So does anybody know how this Problem could be solved without sacrificing too much usability (as submodules do IMHO)

Update:

Please also consider that I'm on Windows Vista and symlinks don't seem to work with Git for Windows.

raphinesse
  • 19,068
  • 6
  • 39
  • 48

1 Answers1

2

If your environment supports symlinks, I would recommend the submodule approach, combined with soft links in order to get the proper directory structure.

That way, you get the benefits from:

From the comments:

basically you have two directories structures:

  • one resulting directly from a git checkout, with its submodules directories.
  • One, manually created to fit the expected structure, with 'ln -s' (or mklink for Windows Vista/Seven) commands in it to link the correct directories from the first structure to the expected place in the second structure.

The OP raphinesse objects:

It would be really nice to be able to just clone the repository and have the required structure to run the app.
Having to recreate the whole hierarchy with symlinks in a separate directory every time someone clones the repository seems like a lot of work to me.

To which I reply:

It can be a script included in the repo itself, that the user would execute in order to generate the correct directory structure. a One-click solution in that case.

The OP raphinesse agrees:

Yeah, I just thought about adding a script which initializes and updates the submodules as well as setting up the relevant links "post-checkout" in the working directory of the main repository (In case the UAC-Prompt doesn't make any trouble). Of course the symlinks set up this way would have to be included in the .gitignore file.

Community
  • 1
  • 1
VonC
  • 1,262,500
  • 529
  • 4,410
  • 5,250
  • I see the benefits of a component approach and therefore would be willing to accept the restrictions git submodule imposes on the workflow. I'm also familiar with the concept of symlinks and occasionally use them on Linux machines. However I don't quite see how I should use them to workaround the fact that a submodule expects to be in _a dedicated subdirectory of the source tree_ as the man page states. Could you please clarify that for me? – raphinesse Aug 05 '11 at 13:12
  • @raphiness: basically you have two directories structures: one resulting directly from a `git checkout`, with its submodules directories. One, manually created to fit the expected structure, with '`ln -s`' commands in it to link the correct directories from the first structure to the expected place in the second structure. – VonC Aug 05 '11 at 13:58
  • I forgot to mention that I'm on Windows Vista. And as I just found out, [symlinks don't seem to work](http://code.google.com/p/msysgit/issues/detail?id=224) with Git for Windows. *sigh* – raphinesse Aug 05 '11 at 14:59
  • @raphiness: a/ Windows Vista has symlink. b/ You wouldn't use them in the git working tree, but in a separate non-git directory structure which will link to the git working tree. – VonC Aug 05 '11 at 15:09
  • a) I know. b) It would be really nice to be able to just clone the repository and have the required structure to run the app. Having to recreate the whole hierarchy with symlinks in a separate directory every time someone clones the repository seems like a lot of work to me. – raphinesse Aug 08 '11 at 09:15
  • @raphinesse: It can be a script included in the repo itself, that the user would execute in order to generate the correct directory structure. a One-click solution in that case. – VonC Aug 08 '11 at 09:18
  • Yeah, I just thought about adding a script which initializes and updates the submodules as well as setting up the relevant links "post-checkout" in the working directory of the main repository (In case the UAC-Prompt doesn't make any trouble). Of course the symlinks set up this way would have to be included in the .gitignore file. I think this could be an acceptable solution. Or do you see any problems with that? – raphinesse Aug 08 '11 at 09:42
  • @raphinesse: no, I don't see any issue at first glance. That might work. – VonC Aug 08 '11 at 09:43