[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [gits] best practice question

In general, I find that having a branch that I don't intend to merge back into master can be tedious. In most projects I've seen, there will be one git repo for the project, a stable "master" branch, and then one or more branches for unstable/experimental features. Some especially large projects will even have entirely different repos for components. Some examples: https://git.kernel.org/cgit/ http://git.musicpd.org/ https://github.com/git/git/branches/all

If I find myself wanting to have a separate branch that I don't intend to merge back into master, I'll either split that piece off into a new repository entirely, or I'll just add a new directory to the main project repo. In your case, I'd do the latter because: A) You want to share content between the documents, B) The documents are closely related to each other, and C) The shared content will likely not be useful in other projects.

Just to give an example of a time when I would split something into a separate repo: If I had a small library (of code or data) that I was using in a project that would be useful in other projects, I would put that content into a separate repository and then use git's `submodule` feature to include that repository's contents into the original repository.

Hope that helps,
Josh Kunz

On 11/13/14 10:54 PM, Suresh Venkatasubramanian wrote:
so I’m using git for a paper I’m writing. I recently branched off a
version for a document we’re preparing for a workshop: the main document
(in prep for a conference later on) is still in the master branch.

But I’m constantly encountering things that should be in both branches
(like relevant papers, bib entries, and common notes). so far I’ve been
cherry picking updates as necessary, but that’s a little painful.

Is there a ‘best practice’ for how to do this properly ? or should I
just not have created a branch

Suresh Venkatasubramanian Web: http://www.cs.utah.edu/~suresh/
Blog: http://geomblog.blogspot.com/
Associate Professor,
School of Computing, University of Utah

Why is it that buying a thought costs a penny, but freely given thoughts
are two cents ? -- Heard on the internet.