I think there is a new concept in what I (and tons of others) have been advocating for a while now: Small apps, not services, including UI
4
6
.@andyhedges @kellabyte There is zero respect left for SOA among the majority of people I talk to, especially in non-enterprise communities
1
2
2
@andyhedges @kellabyte I agree with both of you, and always advocated for the same thing. But not sure this was ever part of mainstream SOA.
@andyhedges @kellabyte Sure, provided two services need to share state. Seems to be in line with the “App” concept in 12factor.net
Another difference in the microservices debate seems to be – as @boicy phrases it –it’s about building for replacement rather than reuse
7
4
2
@andyhedges But that’s in line with his claim (“in the last few years”)
@kellabyte In MSA, my guess is it would rather be something like “CreateOrder”
1
1
@kellabyte In the SOA approach I advocated, a service was a mid-sized, autonomous unit, with maybe a few dozen classes, e.g. OrderManagement
1
@boicy @tastapod @kellabyte @martinfowler As tempting as it is to lose the literal “micro” meaning, it becomes a lot less distinctive then
1
@andyhedges Yes, with “before” I wasn’t referring to @martinfowler’s article, but the SOA discussions of 200x.
@clemensv @tastapod @kellabyte Entirely true, I always liked Microsoft’s take on the principles (if not the implementation).
1
@andyhedges Not sure how to google for that. And I was referring to the last couple of years, when the microservices term was created.
More vintage stuff: RT @clemensv: @stilkov @kellabyte I agree with your list infoq.com/articles/tilkov-10…
1
1
@kellabyte I know. I used to be a big advocate before I became thoroughly disillusioned, particularly with the vendors’ take.
1
1
@kellabyte I’m not sure I know what SOA principles are. In fact I’m not sure anyone does.
3
I personally don’t recall anyone advocating tons of tiny (a few 100 LOC) services before, one for each single feature.
7
1
3