[simgidacav]

Wait, what? How could someone reconstruct single commits? There was no Git back then! The patch program is quite old (1985), but those patches go back to 1972.

[pjc50]

https://en.wikipedia.org/wiki/Source_Code_Control_System dates from 1972. The evolution appears to have been SCCS -> RCS -> CVS -> svn|hg|bzr|git (cambrian explosion) -> git (mass extinction).

[davidgerard]

I'd place the explosion around the time of CVS - but that was an explosion of proprietary VCSes. They didn't have to be as good as git, they just had to be better than CVS.

[pjc50]

Mm, although I felt that Rational Clearcase, MVS, and Visual Sourcesafe were mostly worse than CVS.

[pjmlp]

Given my experience on the enterprise, you can still use the present form.

Although I do like gated checkins in TFS without having to roll our own solutions.

[ajross]

SCCS was contemporary with early unix development, so certainly it's technically possible that there are revision histories going back that far. But I'd have to believe that the early commits in this repo are just release checkpoints. It looks like the most recent stuff (1994) dates from early proto-FreeBSD work.

[autoreleasepool]

Here's the first "synthetic commit" from June 20, 1972

https://github.com/dspinellis/unix-history-repo/commit/67174...

[a_bonobo]

The accompanying publication explains it more: http://www.dmst.aueb.gr/dds/pubs/conf/2015-MSR-Unix-History/...

>Curating and processing available snapshots as well as old and modern configuration management repositories allows the reconstruction of a new synthetic Git repository that combines under a single roof most of the available data

[bitcointicker]

Last page starts here:

https://github.com/dspinellis/unix-history-repo/commits/mast...

Fascinating!

[cyphar]

Because Git wasn't the first version control system. Nor was it the first bringover-modify-merge model of distributed source control.