Microservices are hot. But what if you have legacy and implementing microservices is nearly impossible. Is this the end of the world? Does this mean you are doomed? No, you can have the best of both worlds. Our legacy Procurios monolith is 15 years old. Today, we find the costs of splitting that codebase into services too high. And to be completely honest; we benefit too much. With event sourcing, discipline and hard work we have both the benefits of services and of our big pile of clean code. So let me tell you how you can build independent, mutually unaware applications. On top of that, I’ll share some secret monolith benefits…

Comments

Please login to leave a comment

It wasn't quite the talk I was expecting: very quick (30 minutes, questions time included, in an 1-hour slot) and more about pretty basic code refactoring rather than independent actual applications.
However Joop seems to be very young and yet a natural born speaker, very clear and fresh, there's a lot of potential.

Rated 2

Samuele Lilli at 16:05 on 11 May 2019

Quite deceptive title and duration (I’m not aware of it was due to speaker or organization).
Good concepts though.

I liked hearing the story and real use cases and how you can have modules in an app and how that is similar to Microservices as its all about decoupling code.

However; This talk has a lot of potential to grow and I was missing more “meat”. I would love to hear about how you would go about splitting it into real Microservices if you did that, and why that would be a bad idea. And why sometimes a monolith is good decoupled code. You did have some of that in your examples; but it was mainly some refactoring I’d like to see more architecture and concepts.

Your energy on stage and the way you presented very clearly was great; you’re a natural on stage. Work on your talks more and tell more stories as you seem like a natural storyteller - Then I think you can be an absolutely amazing conference speaker. Would love to hear more of your talks in future, the way you present is engaging.

The talk defiantly had some potential however there are a couple of areas in which I think it could be improved.

1. The title.
My expectation from this title was something along the lines of a talk detailing the work of splitting a monolith into microservices; consider thinking about a title which better reflects the content of not needing microservices to write clean modular code.

2. The length.
Ran a bit short for the time slot, I think you were expecting more questions but this can be quite hit and miss. There are a number of ways you could go to add new content: consider covering some more theoretical areas behind the choices made or showing how you might evolve the code further in future work.

3. Code samples.
Consider having multiple slides with the same sample on with areas of less interest greyed out and other areas highlighted which you can move between as you talk about the code. Alternatively consider removing less interesting bits all together eg the content of the SQL queries had no huge bearing on the refactoring you were doing.

I did not like this talk, the speaker is young and engaging but sadly the concept expressed do not share anything with the microservices concept.

I simply attended the presentation of the refactor of a legacy code base using nice solutions derived by mixing GRASP pattern and clean code suggestions.

In the end, might be an interesting talk but the target has to be reviewed.