Git

What it means

Git is the most widely used version control system for source code. It records every change, lets several developers work in parallel and allows rollback to earlier versions. Linus Torvalds created it for Linux kernel development, and today virtually every software team uses it.

Each change is saved as a commit with a description and author, so the project history is complete and traceable. Developers work in branches where they can try new features without risking the main version, then merge finished work back.

Git underpins collaboration through pull requests, in which colleagues review changes in a code review before merging. Automated testing and deployment through CI/CD, the core of the DevOps approach, build on it.

How it works in practice

Repositories are usually hosted on platforms such as GitHub, GitLab or Bitbucket, which add issue tracking, code review and automation. Companies can choose a cloud service or their own server, for example to protect code or meet customer requirements.

In software development agreements, it is wise to agree that code is kept in a repository the client can access. This makes a change of supplier easier, and for critical systems it is complemented by source code escrow.

Git history also serves as evidence of authorship and progress, for example in disputes or when documenting development costs that may be capitalised as intangible assets or claimed under the R&D super-deduction.

What to watch out for

Passwords, API keys and credentials must never go into a repository. Even after deletion they remain in the history, and if the repository is made public attackers can exploit them. Sensitive data belong in a secrets manager, and a leaked key must be rotated immediately.

When using open source libraries in a project, keep track of their licences. Some licences require derivative code to be published, which may conflict with the company's business model, so maintain an overview of the components used and their licences.