- Current CTO/VPs built/helped build original monolith
- Nobody wants to tell the CTO that their code is shit (and/or is from a different era and needs a complete overhaul), unrelated to that fact it's a monolith. CTOs are too busy doing marketing/getting funding to make a decision on microservices vs monotolith, so the newly-hired architects gets to call the shots
- Everyone cheers on microservices because it fits within the story of a fast growing, serious, technical company and nobody wants to be that lone dissenting opinion/criticise the CTO.
Nobody is seriously and truthfully recommending microservices because they believe them to be the best trade off and superior choice. It's because they like their job, they like hiring people, and it fits within the narrative.
And it just so happens during the massive overhaul that you get to rewrite a ton of code and improve it, while just calling it a migration to microservices
So it's a way of not hurting the feelings of the CTO, going along with the crowd and a way of rewriting a ton of old bad code with an excuse supported by almost everyone
Nobody wants to tell the CTO that their code is shit
It's a pattern because it's a factual inevitability. Whether you're an individual lead engineer or a CTO/Founder, eventually you always look back and conclude things could have been done better and watch in pleasure/horror as you reap the benefits/drawbacks of how you laid down patterns and processes that the team dutifully followed.
Even with good patterns and processes, the shifting landscape of requirements and priorities and business objectives can lead an application into the weeds.
Initially, you may think you're building a single, focused organ like a liver. A few months in you realize that liver needs to power a specialized toaster for English muffins. Eventually, your core offering becomes a conveyor belt that takes toasted rye bread to the buttering rack, and now you need the toaster that only fits two round English muffins per minute to produce 10,000 square pieces of rye per minute while still relying on the liver for power.
- Current CTO/VPs built/helped build original monolith
- Nobody wants to tell the CTO that their code is shit (and/or is from a different era and needs a complete overhaul), unrelated to that fact it's a monolith. CTOs are too busy doing marketing/getting funding to make a decision on microservices vs monotolith, so the newly-hired architects gets to call the shots
- Everyone cheers on microservices because it fits within the story of a fast growing, serious, technical company and nobody wants to be that lone dissenting opinion/criticise the CTO.
Nobody is seriously and truthfully recommending microservices because they believe them to be the best trade off and superior choice. It's because they like their job, they like hiring people, and it fits within the narrative.
And it just so happens during the massive overhaul that you get to rewrite a ton of code and improve it, while just calling it a migration to microservices
So it's a way of not hurting the feelings of the CTO, going along with the crowd and a way of rewriting a ton of old bad code with an excuse supported by almost everyone