Tracking filesystem entries as first-class VCS citizens (was: Bazaar still below the radar when evaluating VCS tools)

[Paul Moore]

On 24 February 2010 09:40, Ben Finney <ben+bazaar at benfinney.id.au> wrote:
> "Stephen J. Turnbull" <stephen at xemacs.org> writes:
>
>> Mercurial and git (especially git) take the point of view that they
>> are tracking content, and a "file" is a way to associate a name with
>> content, and not much more than that. Bazaar considers that a file is
>> an abstract type which has a "name" attribute and a "content"
>> attribute.
>
> Bazaar is tracking more than just the name, but is also tracking other
> attributes of the filesystem entry. Ones I know of include the
> executable permission and the entry type (directory, normal file,
> symbolic link, etc.).
>
> These are important data about filesystem entries, that I want
> reconstructed when applying changes from the VCS to the working tree.
>
> AFAICT, that means Git and Mercurial (by the above description) don't
> track some important information about files that Bazaar does. In other
> words, I don't agree with Git or Mercurial that “a file is a way to
> associate a name with content”, since there are other things a file is
> for.

FWIW, I don't believe that Stephen is right about Mercurial. My
understanding is that Mercurial tracks file renames in much the same
way as Bazaar. Git is the odd one out in tracking content rather than
names.

Paul.