Right, there is no measurements involved and most decisions are gut-based. In the end most systems will work somehow, it’s just that the pain points in development are different.
The older I get the more I feel like this is a feature, not a bug. I have spent too much time over engineering systems that end up being canned because leadership decides to go in a different direction. These days I am much more concerned with striking a balance between perfect and good enough.
Get something out into production, start generating some revenue and finding the stuff people like about it, then see what is giving you the most problems and fix that. That’s going to be a much more successful path than never shipping anything because of endless polishing and optimizing.
If I ever find myself spending a long time debating some technical decision (aka bike shedding), I have learned to hit pause and say let’s just pick a path and go with it. As an aside this also works for deciding what restaurant to go to or what kind of pizza to order.
Completely agree. I'm personally inclined to be more toward the "good enough" end of the spectrum than the "perfect" end of the spectrum - but it's still a judgement call. Ultimately, one needs to turn out a product, make money, and fix it up where there are gaps or deficiencies. It is a feature; a good one. Anything else is just hubris and vanity.
If you feel you need micro-services to force an organizational change, don't equivocate, just lead with that, and get back to business.
Even if there is measurement, it’s pretty easy to cherry pick stats or question the existing stats. I’ve seen it time and time again often “data driven” only matters if the data agrees with the leader/devs preconceived beliefs.
spaceywilly|11 months ago
The older I get the more I feel like this is a feature, not a bug. I have spent too much time over engineering systems that end up being canned because leadership decides to go in a different direction. These days I am much more concerned with striking a balance between perfect and good enough.
Get something out into production, start generating some revenue and finding the stuff people like about it, then see what is giving you the most problems and fix that. That’s going to be a much more successful path than never shipping anything because of endless polishing and optimizing.
If I ever find myself spending a long time debating some technical decision (aka bike shedding), I have learned to hit pause and say let’s just pick a path and go with it. As an aside this also works for deciding what restaurant to go to or what kind of pizza to order.
kayo_20211030|11 months ago
If you feel you need micro-services to force an organizational change, don't equivocate, just lead with that, and get back to business.
softwaredoug|11 months ago