This is one of those @unclebobmartin post’s I disagree with almost completely: blog.cleancoder.com/uncle-bo…. The deployment model is NOT a detail.

Oct 3, 2014 · 10:19 AM UTC

11
22
30
Replying to @stilkov
@stilkov don’t you think it _should_ be a detail?
10
@unclebobmartin In an ideal world – yes. I have never seen it work that way though, but I’ve seen many projects fail that thought it would.
1
This tweet is unavailable
@evandor34 Don’t even get me started.
@moru0011 don’t depend on synchronous calls.
Replying to @stilkov
@stilkov @unclebobmartin Agreed. An "agnostic" component model is a recipe for disaster.
1
Replying to @stilkov
@stilkov @unclebobmartin strongly agreed
1
Replying to @stilkov
@stilkov Very 1990s, @unclebobmartin. No DHT, no Consensus, no Failover, no Replication, no Latency, no CAP. Distributed system, clean cake
2
1
2
Replying to @stilkov
@stilkov @unclebobmartin @chadfowler I think his point was that you shouldn’t start a design defined by the deployment model. I agree.
2
Replying to @stilkov
@stilkov @unclebobmartin +1
Load more