GNU Arch is a distributed revision control system, meaning that it does not require the use of a central “repository” for people to collaborate. However, it does allow for such a cooperation model and the next subsection will detail how to do this.
The following subsections focus on distributed revision control—that is, how to exchange data among branches located at different remote sites in order to efficiently collaborate. This is a salient feature of GNU Arch and is extremely useful in development patterns that are inherently distributed, as is common with Free Software.
From a practical viewpoint, one may also argue that the distributed approach can supersede the centralized approach even in cases where developers are tightly connected (e.g., within a company). First it removes the need to set up a single central repository which in effect is a single point of failure. Second, it does not require complex access right management which is usually implied by the use of a central repository. Finally, it allows each developer to synchronize its working branch(es) with other people's work when she/he sees fit.