Microsoft Visual Source Safe - Overview

Overview

SourceSafe was initially not a client/server Source Code Management, but rather a local only SCM. Architecturally, this serves as both a strength and weakness of design, depending on the environment it is used in. It allows a single user system to be set up with less configuration than that of some other SCM systems. In addition, the process of backing up can be as simple as copying all of the contents of a single directory tree. For multi-user environments, however, it lacks many important features found in other SCM products, including support for atomic commits of multiple files (CVS has the same problem as it is built upon the original RCS). SourceSafe inherits its shared functionality using direct remote file system access to all the files in the repository. This, together with a bug where the code is using old memory after a call to reallocate, are contributing factors to why SS databases sometimes go bad.

Starting with VSS 2005, Microsoft has added a client–server mode. In this mode, clients do not need write access to a SMB share where they can potentially damage the SS database. Instead, files must be accessed through the VSS client tools - the VSS windows client, the VSS command-line tool, or some application that integrates with or emulates these client tools.

Read more about this topic:  Microsoft Visual Source Safe