My point here is that ultimately it doesn’t really matter. You can make either work at scale (Spotify for microservices, Google for monorepo). So don’t even bother thinking about it too much, just make a decision and start investing in tooling. Which I think is your same statement here.
Ultimately I’m going to ask whoever proposed micro services vs monorepo is: how are you going to qualify shared changes? How are you going to publish artifacts? How are you going to document things and make them reusable? There are ways to do so in both approaches. If someone isn’t answering those questions and weighing them against the effort required to implement only the tooling changes without the application code changes, they’re missing the point and are at risk of just wasting everyone’s time remaking the same mistakes that their predecessors made (not investing in tooling) that prompted the discussion about moving from one to the other in the first place.
Ultimately I’m going to ask whoever proposed micro services vs monorepo is: how are you going to qualify shared changes? How are you going to publish artifacts? How are you going to document things and make them reusable? There are ways to do so in both approaches. If someone isn’t answering those questions and weighing them against the effort required to implement only the tooling changes without the application code changes, they’re missing the point and are at risk of just wasting everyone’s time remaking the same mistakes that their predecessors made (not investing in tooling) that prompted the discussion about moving from one to the other in the first place.