Interlude. Before GitHub

When I tell young developers that I caught the days of CVS, they look at me like someone who remembers steam locomotives. CVS appeared in 1986, it was devised by the Dutch programmer Dick Grune, and it was one of the first systems that allowed several people to edit the same code at the same time without overwriting each other’s work. Then, in 2000, came Subversion, which everyone just called SVN, and at the time it seemed the height of civilisation. Git appeared in 2005, and GitHub, without which programming is hard to imagine today, opened only in April 2008. I went through that whole path in the ordinary course of work: first we lived in CVS, then we moved to SVN and were glad we no longer had to be afraid of renaming a file, and then Git quietly took over, and the first two are now remembered mainly by dinosaurs like me.

Those who grew up afterwards sometimes imagine that before GitHub there was little code by other people and it was hard to get. In fact there was an abundance of it; it just lay scattered across many private warehouses and did not yet have one big square. There were the huge source trees of GNU, there were the BSD systems, where almost everything was open, from the kernel to the last utility, there were the servers of various universities and projects, from which code was fetched through that same CVS or SVN, and there were mailing lists where it was discussed and fixed. The archaeology of programs began long before it had one shared museum.