I really love git and now I have setup an remote repositorie. Thats pretty cool but I cant find the plain data like. My Idea was:
local development > commit, commit, commit and if ready for the production > push to remote > and the git repositorie is the virtual host in apache and the data is live.
Is my thiniking really wrong or how can I realize that?
Thanks for you help!
Your thinking process seems to be correct, very similar to how I would do it.
I think this document will help you with understanding GIT, it was a huge help for me, especially with a subversion background: http://progit.org/ebook/progit.pdf
Depends--are you okay with potentially untested code going to production? For example--Dev A pushes changes intended for production, tests them out and they look fine. Meanwhile, Dev B tests in some code--but oops, he made a typo and now the application immediately fails on requests. Dev A, meanwhile, pushes to production, bringing in Dev B's broken changes.
You need a deployment strategy--that is, you need to always be sure that what you're pushing to production is (a) tested and (b) reliable. I'm of the opinion that production should NEVER be running "straight from git," if only because that makes it very easy for bad things to happen--i.e., if someone gains access to your frontend, they can now do whatever they want to your code base. At best, they can steal the code and your history. At worst, they can make changes to your code base--likely without you realizing it.
You could, for example, have a separate repository specifically for production--that you can fork from and then merge changes as you need--or deploy the code as a separate archive. This would give you a lot more control--and insight--into what is going to prod.
If you are interested in git workflow, you should definitely checkout A Successful Git Branching Model and Distributed Workflows from the GitBook.