Is it okay to have a special commit for each version increment?

Question

I have a manifest.json file in my project and a build process that uses that file and creates the deliverable to be sent to the customer (they are beta testers). So here's what I do. I do a number of commits for development and bugfixes each time and when I think I'm ready I create another commit just to increment the version in the manifest.json file. It looks like this:

@@ -1,7 +1,7 @@
 {
   "manifest_version": 2,
   "name": "Blah Blah Blah",
-  "version": "0.4.2",
+  "version": "0.4.3",
   "description": "Description goes Blah Blah Blah ha ha!",
   "content_scripts": [
     {
~
~

Is this okay? Any other suggestions. Am I polluting commits?

Answer 1

Is this okay? Any other suggestions. Am I polluting commits?

It looks perfectly fine to me, and this is useful to set special tags on it and such.

If you're concerned about disk space for example, when doing this way, Git will record only the manifest.json itself, the rest being already referenced in existing objects. So the footprint of such a commit will remain minimal.

EDIT (to include what we've discussed in comments below).

It's always a good idea to prepare "cooked" commits before submission, as well as forming "logical units". It's even required when submitting patches to a maintainer (at least in Linux kernel project), so this one can reject some of your patches if needed, while keeping on merging the rest.

In that way, separating technical enhancements and administrative stuff is a good thing. As regards the Linux kernel once again, Linus Torvalds itself does the same on Linux's main branch. The following commit simply switch to "5.13-rc1" version.

Commit 6efb943b861 on kernel.org
Commit 6efb943b861 on github.com