loading...

re: How do you know when a technology is ready? VIEW POST

FULL DISCUSSION
 

Technology adoption should be less a matter of timing and more a matter pragmatism.
The right time to start using a new technology is when you need it, and precisely when your old tools aren't cutting it anymore.
There are plenty of industries and workflows that demand implementation of very new tech, and are running things right into deployment fresh out of the womb, so to speak. Usually because the new thing was made for their specific needs. But, these are usually obvious cases, and often fraught with problems over time. Hasty adoption creates weird stacks that don't modernize well.

In general I think devs and their employers are both overeager to adopt new technology because doing so asserts forward thinking and connectivity with the zeitgeist. These decisions are more mimetic than pragmatic, and stem from raucous conversation, hype, and controversy on tech-news reverb chambers. Being able to show off the new tech is pleasing to investors and sexy at dinner parties, but these decisions are guided by social factors that don't really mean they are actually better. If timing is ever a factor in adoption decisions, it probably means you're being pushed by zeitgeist and not honest about you're own needs.

Adopting young technology can put you out of touch with the standards and consensus that take longer to crystallize, and should be done cautiously. My experience has been that software selection is a constant exercise in bandwagon hopping, each with it's own problems and benefits. The right technology is neither the HOT new one, nor the old tested dog, but the one that balances it's explicit strengths and weaknesses with your needs: maturation is rarely a factor.

 

these decisions are guided by social factors that don't really mean they are actually better

We ran into this when working on j2objc a few years back. It was stable, used in production by Google Sheets, Gmail, and more key products. So it was mature and also very low-risk for orgs to adopt. However, neither java nor objc were exciting so we couldn't sell it.

Luckily we are seeing Kotlin Multiplatform has a better "balance of strengths and weaknesses of the technology and the needs of orgs" (as you put it).

code of conduct - report abuse