Yeah, perfect. Thanks.

Sometimes things don't work. Hi everybody. I'm Christian from GSI. First

me is Dennis Klein. He's long doing our group. So later on, if you have some

complex questions, he might be able to answer them also. We use the software

development for experiments group in the IT department of GSI and FAIR. I stress

that because I'm not going to talk about GSI as a whole or FAIR as a whole

because we can't. It's too heterogeneous. So I'm only talking about our group.

The software that we are doing, and I'm going to show it in the next slide, is

mostly library and framework styles. So we are providing the base stuff that

other experiments use for their analyzers, for their reconstruction, for

the simulations, whatever. Lately we are also being involved in some

infrastructure things. Basically we are deploying some of the software for our

supercomputers. So we are doing some container stuff and all that things. Here

you can see the software stacks that we are doing and we are doing exactly the

middle box where the name Alpha is printed on. So that's the stack that we

are, it's our responsibility. Whereas at the top you see the software that

uses our stuff and at the bottom you see the stuff that we use. We are actively

working also with both of them. So you can see contributions on all of the

things like on Boost or CMake from us. And also we are working with the

experiments. So if they have some issues we will help them and we will look at

their software. And even some of their software is as a test case in our CI.

So that we make sure that when we change our software that it still at least

compiles and runs some tests for few selected of cells.

So let's go. Introduction. Most of the aspects that were already named

are really great. I'm going to repeat some. So even if it's what some people

say it's all exactly what we would say also.

What I'm going to concentrate on in this talk is on the workflows and tools that

we are doing. Here is a link on our workflow

documents. So if you want to know exact details what we suggest people to do

you can see it there. And this talk will also conclude some

recommendations and lessons learned in some places.

And finally. Version control. We are using only Git.

Nothing else finally. Most of our software is on GitHub.

At least our public software. I will go into detail

later on. Some of our things are on GitLab.

On a private instance. I will go to introduce

also. We are really using it extensively.

Everybody is creating repositories on an as needed basis.

Either privately or for the group and even for really small projects

that are only a few hundred lines or thousand lines maximum.

We have repositories. We are trying to have really good commit

messages. So if you look three years later you

need to know what happened.

Test tweets were also named lately and

it's a constant point where we are improving still.

Our test tweets are there. They are working but sometimes

even the test tweets break in some places and that's what we are

working on. Our big projects are CMake and CTest based.

Which already was mentioned by the people. CMake is really the thing

that for CCD++ projects are the going thing.