There were three key takeaways I appreciate getting from this talk:
1. Spend lots of time upfront identifying a well-designed API.
2. Abstract your API from the implementation at all costs, even if your implementation is currently an API provided by an external vendor, so that swapping out the implementation in future doesn't seem an unapproachable proposition with respect to enterprise continuity.
3. SOAP with WSDL's problem is how likely it is for the level of detail implied by a WSDL to make you de-facto coupled to the specific implementation of the service that generated it.
I had a hard time following the benefit of splitting the Experience layer and Process layer in your three tier model. Just a little more explanation of that would have been helpful for me.
I felt like the first 12-15 minutes of this talk was bogged down in politics. I fully expected the slides to start with the "Your services WILL fail" slide and then build up an argument for doing microservices right.
Many of the ideas could have used diagrams and common nomenclature such as "this is the standard Adapter pattern."
I thought this was very clever to have a session where the solution is opposite the title of the session. It was also interesting to learn about microservices as I've not had the chance to work with them before. I felt like the familiarization with the technology helped me gain better perspective on where the industry is heading. And that was very worthwhile.
Comments
Comments are closed.
There were three key takeaways I appreciate getting from this talk:
1. Spend lots of time upfront identifying a well-designed API.
2. Abstract your API from the implementation at all costs, even if your implementation is currently an API provided by an external vendor, so that swapping out the implementation in future doesn't seem an unapproachable proposition with respect to enterprise continuity.
3. SOAP with WSDL's problem is how likely it is for the level of detail implied by a WSDL to make you de-facto coupled to the specific implementation of the service that generated it.
I had a hard time following the benefit of splitting the Experience layer and Process layer in your three tier model. Just a little more explanation of that would have been helpful for me.
I felt like the first 12-15 minutes of this talk was bogged down in politics. I fully expected the slides to start with the "Your services WILL fail" slide and then build up an argument for doing microservices right.
Many of the ideas could have used diagrams and common nomenclature such as "this is the standard Adapter pattern."
I thought this was very clever to have a session where the solution is opposite the title of the session. It was also interesting to learn about microservices as I've not had the chance to work with them before. I felt like the familiarization with the technology helped me gain better perspective on where the industry is heading. And that was very worthwhile.