4 ms·
The flip side, though, is that intrinsic measures of software quality are terrible. Most boil the ocean looking to define what makes something good in the smal
by taeric 9y ago
The flip side, though, is that intrinsic measures of software quality are terrible. Most boil the ocean looking to define what makes something good in the small, and completely fail to generalize into the large.
Worse, too many are interested in building a general solution when they haven't even done a single concrete solution. I have yet to find a proud developer that said, "I used the boring choices for this problem and got a solid solution."
This is akin to expecting an artisan to give better results for a standard kitchen than a builder would get from buying Ikea.
And don't take that as some sort of value judgement either way. Builders are there own to of Craftsman. Different doesn't need a hierarchy.
My gut for why this is? In a fiercely competitive field, leverage from success is the primary driver. And you get more leverage and success by being different. The upside is huge. Naturally, so is the downside. Survivor bias kicks in and people think they succeed from wiser choices, so they are incapable of reflecting and making less risky choices as they age as an org. Which requires an influx of young risk takers that have not hit a downside yet.
- apineda 9y agoThis is difficult to read. I'm sure you have some good points in there.
- taeric 9y agoI'll blame writing on my phone while on the bus. With my stop coming up. :) That said, probably nothing too deep. I certainly wasn't aiming for it. My main points: * Most of us in engineering focus on intrinsic measurements of our code's quality. These don't generalize well to product quality * Nothing wrong with being a builder, instead of an artisan. Especially when it comes to building. * Most people that succeeded fall to survivor bias and think they know why they succeeded. They are usually wrong.