Git Tutorial #1 - Reset current branch to delete local commits with no data loss, and replace with 1 better remote commit
When you are working on a project thats undergoing changes so rapidly that you have no chance to do milestones or take breaks to put out stable versions, you should still commit LOCALLY (Commit, not Commit&Push) often to back up your work, and have old code to go back to during troubleshooting. Taking the time is annoying so just write as little description as possible to save time.
Then after you have gotten to a milestone or a stable point, you can throw all those away, and Commit and Push ONE single solid commit with complete description and thought put into it when you have time.This is best to do with local commits only, otherwise there can be some weird situations.
I mean you can do it, but its bad practice to change the history.
Step 1 - Highlight the last commit you want to keep.
Step 2 - Right click on it and choose Reset current branch to here.
Step 3 - This is where you make important decisions. In this tutorial we want to leave our files unchanged, and just delete the commit indexes.
Step 4 - Now it will show the last commit as the past one you picked, and that "New" files can be committed. Still no file data has changed. At this point you want to the final Commit, and (most likely Push to github too).
Step 5 - You see you've effectively turned all the junk into 1 well formatted commit.
Can the text/descriptions for the old ones can be copied to the new text? No. There is no easy way to aggregate them all like some other git merge commands do. This is why I said dont put that much effort into the temp names. You could copy/paste em manually if it was that important.