[TYPO3-dev] Best practice: How to store a complete TYPO3 installation in git

[Patrick-Emil Zörner]

On 8 May 2013 18:09, Andy Grunwald <andygrunwald at gmail.com> wrote:

> i know that many people / companies working with git and TYPO3.
> Today i asked me (during a coffee break) which kind of ways are exists to
> store a _complete_ TYPO3 installation / project in git.
>

Our organisation just did it because I decided to do so about two years ago.


> Every problem in development can be solved by x ways. I think this
> question, too.
> But what are advantages / disadvantages of the different ways?
>

I do not think there is a recipe. You have to look at the requirements of
your organisation an decide what makes sense. Especially what to just
backup and what to put in a version control system.


> Some (created in 2 min.) thoughts are:
>
> * Store the complete TYPO3 core source in Git?
>

Does not seem very useful to me.

* Use a "build" script (e.g. to download TYPO3 source or extensions from
> TER, ...)
>

Use submodules. You should make your own forks if you do adjustments though.


> * What about the database / content / changes? Tools like Liquibase [1]?
>

We currently only store sys_template TV-O and TV-DS.


> * What about content assets? Set them to ignore?
>

What do you mean tt_content and pages?


> * What about TYPO3 Source updates?
>

I have no clue what you are trying to do.


> I think there are many more questions.
>

Make a plan what you want and what makes sense to you first ;-)

The way of storing a project into git it direct related to the way to
> deliver it to the server.
> What are solutions for this?
>

Every developer in our organisation including designers has his own
test-account. They present the results in a branch. If it looks good to the
one who submitted the ticket a message is sent via our git host requesting
to include changes. The person in charge looks at the changes and decides
if the changes OK. If so the next decision is if to just merge or cherry
pick (some changes are OK but some are shit).

A short google search opened a stackoverflow post [2]
> But this far from "complete".
>

Happens ;-)


> Maybe some of you can post their experience and the way how you work with
> git and TYPO3 projects (private / customers).
> E.g. * What about your .gitignore? * What structure do you use? * Are
> custom scripts involved? * If yes, what does these custom scripts do?
>

Wow! I could write pages about that.

Global and local ignore file is far to complicated and I do not know what
you want. Could write a assay but the mailing-list is not the place.

Yes we use a bash-script I wrote that does a lot of things. I can only tell
you what I do but again I do not know what you want.


> If there are some useful answers we / i can create a full "Best practice"
> post / topic in a Blog or on Buzz about this topic.
> In my opinion this is a topic which is interesting for many people.
>

Write one ;-)


> If someone of you using Subversion, the same question can be applied to
> this CVS as well.
> So your answer is welcome, too.
>

If you do not use git or mercurial you are doing something wrong!

I offer you to talk to you on skype and keep in contact. In return I would
like that you write an article and publish it (how we need to find out
first).

greetings

paddy