How do I store my previously-written code?

published Feb 17, 2006, last modified Jun 26, 2013

Slashdot | How Do You Store Your Previously-Written Code? asks a fairly simple question.

With a fairly simple response:

Subversion . I do Subversion for a living. You may choose another version control system, such as DARCS or SVK , but I use Subversion. I organize my code in:


form, which allows me to perform most common repository operations (merge, branch, tag) fairly quickly and without worries.

After I've pronounced a release stable enough, I tag the release (by copying the trunk/ into tags/release-x.y.z . Then, I check it out, and make a tarball out of it, which I place on a separate folder.

Each project gets its own repository. Each project also gets a unique folder named after the project's name, which usually contains a TO-DO list in text form, a checkout of the trunk/ for that project's repository, and assorted files and documentations (or symbolic links to my knowledge base folder, Information collection ). In each project folder, I also have a subfolder which contains each release in source tarball form, and (if the project needs compilation for a target platform) the compiled installable packages ( .tar.gz , or RPM )

I help myself by using Trac . It couples fairly well with Subversion, letting me know how my project is evolving, and letting me see colored differences for each revision.

All in all, use a version control system. That, along with regular backups, ensures you have full productivity and complete history of your work.