|
Source Control
Provides multiple users with an option to store SQL scripts and other files in a source control repository.
Aqua Data Studio provides clients for two source control tools: Subversion (SVN) and Concurrent Versions System (CVS).
What is the difference between CVS and Subversion?
Revision Numbers
- Revision numbers are by file within CVS. CVS employs RCS (Revision Control System) as a backend, and each file has a matching RCS file within the repository. The repository is generally organized according to the structure of a user's project tree.
- Subversion repositories appear as a single filesystem. Each commit generates a completely new filesystem tree, so the repository can be compared to an array of trees. Revision 57 would refer to a particular tree, which is the way the filesystem appeared after the 57th commit.
- In CVS, you may talk about 'revision 9 of source.d.' In Subversion, you would say 'source.d as it appears in revision 9.' In CVS, revisions 8 and 8 of source.d are always different. In Subversion, it's quite likely that source.d was not changed between revisions 8 and 9.
Commits
- In Subversion, commits are atomic - they all happen at once or none do. The entire change is applied (or rolled back) and is only visible to other users once finished.
- In CVS, a commit alters each file sequentially until it is done. This can cause problems. If a network connection dies in the midst of a commit, this can set the CVS repository partially changed, and often corrupted and unstable. If a user updates their working copy during the commit of someone esle's change, they could end up with a partial commit from the CVS server.
Handling of Binary Files
- In CVS, each revision of a binary file is stored individually, which can use a lot of storage. In Subversion, all files are stored as binary and a very efficient binary Compare algorithm (Vdelta) computes the differences between them. Multiple versions of binary files take up much less space than when using CVS.
Directories
- In Subversion, directories are versioned, just like files. Subversion stores the state of files in a repository, which tracks the history of both files and directories. CVS is unable to track history for a directory.
Renaming Files
- CVS does not support the direct renaming of repository files, and cannot trace version history of files that have been renamed.
To rename a file in CVS, you are forced to do the following:
$ mv myfile1 myfile2
$ cvs remove myfile1
$ cvs add myfile2
$ cvs commit
The new myfile2 has no record or indication of a common history with the old myfile1 (now in the Attic).
SVN's move command can be used to rename files.
$ svn move myfile1 myfile2
$ svn commit
Here, the common history of myfile1 and myfile2 is kept.
Metadata
- Properties support is a Subversion feature allowing users to append metadata to any versionable object within a repository. This can include things such as mime type, whether the file is executable or not, etc.
Branching and Tagging
- In CVS, branching and tagging are extremely processor intensive for large repositories. Their processing is proportional to the number of files being branched or tagged.
- Subversion makes both branching and tagging constant time operations. It doesn't differentiate between filesystem space and 'branch' space. Branches and tags are just directories within the filesystem. Branches and tags are created by copying the directory you are tagging.
|
|
