The past years
The past years brought some ups and downs for all of us. These days are not so relaxed either. Nevertheless we slowly but steadily grew our business. Bit by bit, App by App.
Attracting good people (aka talent) is not easy for us. Especially here in Germany there is high demand for DevOps and developers. We have high technical standards. We compete with big tech. Over the past years the wages have been rising. But working in a small company has its advantages as well, and we pitch that to potential candidates.
With the COVID-19 pandemic we became a remote company. Hiring internationally helped us. That’s also a good fit to our international client base. Diverse cultural backgrounds are sometimes challenging. Only being connected by video calls and software is hard.
Of course, increasing head count does not necessarily raise productivity by the same amount. It comes with overhead. And it takes time to get people boarded.
Domain knowledge was traditionally also personal knowledge here. A cultural change is ongoing to share more. A knowledge base including standard procedures helps us to capture tacit knowledge into more accessible codified formats.
Keeping everyone aligned and moving forward in the same direction takes extra time and effort too. New patterns are required to stay productive.
Vision and alignment struggle
fortrabbit was founded by three friends with equal rights. That works as long as there is a strong alignment in vision and actual implementation between the partners. Over the years it turned out that goals and ideas drifted apart. Finding direction became extremely difficult if not even impossible. We have spent a lot of time discussing instead of trying ideas. It was painful.
We failed to find a shared direction, but we settled the conflicts by shifting power. Not the optimal solution, but something that enables us to move forward again.
More clients = more responsibilities. Maintenance, support and day to day operations are keeping us busy. And then there is the hosting platform itself. It is a home-grown system that has reached maturity. It’s not perfect, but it is battle proven.
In recent years we have invested a lot by updating underlying software, refactoring code and improving monitoring and stability.
It’s still not as ideal as we wish it to be. There is a very long backlog of small improvements and documented quirks. We are also maintaining a long list of new features and big improvements.
We have to admit that we were not able to keep up the pace of the early days. We can not afford to introduce breaking changes since we have websites in production. Supporting and maintaining old and new features on the same platform would become too complex a game to play.
Core parts of the platform are now 10 years old. There is a lot of custom code where open source solutions are available today. The expectation of developers evolved, what was new and unique back then is standard today. Back then Let's Encrypt certs was a nice bonus, now it's a requirement.
A New Platform to be build
We have finally settled to do a bigger rewrite project.
It’s easier to build the spaceship on the ground than building while flying in space.
This approach certainly has its risks. But it also frees us from constraints and it enables us to approach features from new perspectives. That decision was already made more than 2 years ago. Since then we have been discussing and defining the feature set of the new platform.
Now, half of the productive team time is going into the new platform, with more to come.
At its core the new platform will continue our mission to provide PHPower to the PHPeople — much improved, with good backward compatibility. We are looking forward to be able to migrate over 90% of all existing clients.
We are working on a final feature specification. Part of the process will be internal and external client feedback rounds. We started doing client interviews, asking for feedback on a visual mockup.
Ping us if you are interested.
At the same time, we have started approaching the project from the infrastructure side. On that level we are discussing and stress testing various implementation models.
As usual, we will not provide estimates on when something is going to be finished. Even when we think something is only a week away to launch, we already know from past experience that it will take much longer.
We can say that it will take at least a year from now for the new platform to launch. Then there will be a long transitioning phase to migrate existing clients over.
More details will be provided along the way. Thank you for your attention.
p.s. I struggled a while to put this online.