@FaisalFeroz@iammukeshm Not really. Keeping the domain pure is easy. Hard rule: no adding nuget package dependencies to the core. Only business logic, interfaces, and pocos live there. Everything else goes in a different layer.
@housecor Well, I think youre on the right track. Standing up ephemeral env for testing then tearing them down, feature flags, and canary deployments are best practices. That backed by solid rollback mechanism for recovery. It requiers a sophisticated prod env. Its a system of Confidence.
When you combine the vertical slice architecture pattern with the organization of onion architecture the question of shared logic goes away on its own.
All of those handlers/features/slices get distilled to just "business logic" and live together in a core project. Its elegant. No need for "Domain Models."
This course is a game changer. A light turns on and you realize you've been coding in the dark. It challenges you and changes your perspective down to the core. It levels you up in ways you could not imagine.
Maybe you've heard about the fallacies of distributed computing. But if you want to understand them in more detail, join @udidahan's Distributed Systems Design Fundamentals course. It's free, and available online to complete at your own pace! #DSDF https://t.co/RnksTHMA56
The anxiety a message like this gives you when you have 12 hours to go, lots of work to do, and you still have lots of premium requests remaining in your account, but #GitHub#Copilot has implemented multiple levels of quotas and throttling.
Working full swing on releasing fullstackhero v10 soon! Final stages of work are happening.
In case you have missed, we will have React frontend (admin + dashboard apps), and a .NET 10 supercharged backend, all running seamlessly with aspire!
All of the latest work is on the develop branch: https://t.co/gcJwq7x8r2
Release soon!
@iammukeshm@Meligy I nice way to cover reads would be to add Microsoft DAB. It lets you add reads without having to code to a specific database schema; or other data sources.
https://t.co/zd7Q2Wrw2k
@davidfowl The problem with this article is that its conflaiting process with pipeline. Process stems from a lack of confidence in the system. Enter SCRUM and more human touchpoints. Pipeline is technical. It can eliminate touchpoints but not if there is a lack in confidence.
All right y'all. It's happening. My first book on software development processes is almost complete.
The High-Velocity Engineering playbook will be coming soon. It's intended for small to medium size companies, and those established ones still stuck doing SCRUM.