3 ms·
This doesn't work in all organizations. There is a big risk that some workers won't do their jobs right given the freedom to do so. In a nicely compartmentaliz
by yummyfajitas 13y ago
This doesn't work in all organizations. There is a big risk that some workers won't do their jobs right given the freedom to do so.
In a nicely compartmentalized organization/application this isn't a huge issue - worst case, one compartment performs poorly, the people running it are replaced, and a complete rewrite of one component is performed.
This can be really bad in a more monolithic organization/application. Bad work in one component (e.g., a poorly designed database schema) will leak out into the rest of the system. Because others are building things on top of it, it will be extremely difficult to rip out and replace.
- derefr 13y agoSounds like the big risk is in choosing to have a monolithic architecture.
- einhverfr 13y ago> There is a big risk that some workers won't do their jobs right given the freedom to do so. Find new workers then. The principle is you do the work, you own your work or job, it is yours with a share of the risk and reward that comes with it. > Bad work in one component (e.g., a poorly designed database schema) will leak out into the rest of the system. The solution there is two-fold: 1. Encouraging (though not forcing) getting second opinions, discussions etc ahead of time. 2. Post-review once the component is nearing usefulness by respected peers. "Can we use this? What is the quality?" Is there expense when you have to say "this needs more work" or even "this must be done again?" Sure. However does putting less technically apt managers in charge of the technical decisions cause fewer problems or more ones? You can't just have management and review go away. But management is the midwifery of production, and review of technical peers beats review of managers any day.