With the API pulled out of the original site repository, the vagrant VMs don't work to well.I am looking at solutions for this and would like to have some feedback.My first solution is to create a repo for the VM, then add the other repos as submodules. I see 2 potential issues with that.The first one is that some changes might be lost. According to the documentation (http://git-scm.com/book/en/Git-Tools-Submodules#Issues-with-Submodules), when doing `git submodule update`, changes that are not in branches will be lost. I know we are suppose to always submit changes in branches, but I'm sure not everyone does.The second issue I see, is that the submodules would point to the main repos, not the developers forks. That looks like it would make it a pain to commit changes.Another solution I see is to only create the vm. Then have the people do the clones by themselves inside the vagrant folder. That does not sound too friendly, but would probably works fine.And the last one I thought of is having people symlink their repo in the VM vagrants folder. Again, not to friendly. And it would not work in Windows.I'm going to think of it more and try to play with the submodule solution this weekend, but I would like to have feedback. Maybe someone has a better idea.Thank youEric Hogue--
You received this message because you are subscribed to the Google Groups "joindin-developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to joindin-develop...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
I thought about this more. I think the best option is to have the VM in a repo. Add the joind.in, joindin-api and joindin-web2 folder to .gitignore. Then have the users clone their fork in there. We should probably have a small script that would extract the user name from the VM repo, clone the 3 repo from his forks and add the upstream remote for the user.Any thought?