Which is the better repository structure for my project?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



I have a web application ProjectFoo managed by subversion, and now I
want to use git to manage it.

ProjectFoo have
* 4 editions, the differences between each edition are trivial things
just like images & CSS files.
* 2 development lines, stable & unstable, so we can always do bug
fixing in latest stable branches, and add new features to unstable
branches, test them and finally deploy them as new stable branches
like 1.2.

The repository structure of project_foo looks like this:

|
+- trunk
|
+- branches -+- stable -+- 1.0 -+- trunk
             |          |       |
             |          |       +- edition_1
             |          |       |
             |          |       +- edition_2
             |          |
             |          +- 1.1 -+- trunk
             |                  |
             |                  +- edition_1
             |                  |
             |                  +- edition_2
             |
             +- unstable -+- edition_1
                          |
                          +- edition_2

I've found git-svnconvert.rb can convert subversion repository which
have strange branches structure like above to git repository, and I've
done some modifications & fixing to make it works with the latest
version of git. It even can split a subversion repository to several
git repositories by branch.

I've imagined 2 different structures of the new repository:
1. Follow the old structure of subversion repository. The known down
side is, with single git repository, I can not test 4 editions at the
same time.
2. Split to several repositories by edition, each repository contain 2
branches (stable & unstable) of an edition, and a repository of
"trunk" edition used to do the main development, other edition of
repository just track "trunk" repository and pull changes from it.

Which one is better?

I've tested method 2, when do git pull on edition_1 repository, it
seems do not known which changes already made on edition_1, and looks
like try to merge all changes on trunk/master (from "trunk"
repository) to edition_1/master, actually these changes already merged
to edition_1 in subversion repository.

The question is, when repo_1 start to remote tracking repo_2, is there
any method to tell git that repo_1 is "already up-to-date" with
repo_2, so that git pull will just merge new changes on repo_2 to
repo_1?


-- 
Best Regards

Rainux
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[Index of Archives]     [Linux Kernel Development]     [Gcc Help]     [IETF Annouce]     [DCCP]     [Netdev]     [Networking]     [Security]     [V4L]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [Fedora Users]

  Powered by Linux