The meta question suggested by this article: When a product succeeds wildly (like Jira has) its creators make a ton of money. That's good. But what should th

The meta question suggested by this article: When a product succeeds wildly (like Jira has) its creators make a ton of money. That's good. But what should they do with the money? Is it smart to roll the money back into the product by adding new features? We all know the advantages of that approach. But it obviously causes feature bloat and complexity bloat, especially when backward compatibility is the prime directive. That's the curse of the customer base. For most of the last generation we've been able to ride Moore's Law and the rest of the cost curve plunge. It simply didn't matter if each feature we dreamed up took more cycles / RAM / disk / network. We could count on the resources being there. That set us free to try to do more and more complex things. But Moore's law only helps indirectly with managing complexity. For hackers it makes possible stuff like "Show Definition..." in the IDE. This complexity bloat afflicted Microsoft Windows back in the XP / Vista days. It afflicts what I'm working on now. What else could successful product creators do with their money? * Obviously, return some of it to their investors. * Invest it in new complementary products. * Use it to finance big organization-wide pushes to refactor to improve performance. * Reduce prices and drive more widespread adoption. The movie industry is an interesting model. They put together a team to make a movie, then disband it. When they have a big hit the participants with residuals rake it in. They then invest in more movies (some of them are atrocious sequels, but many are good new stuff.) Imagine if they did what we hackers do now. _The Graduate_ would be on version 17, and the guy would saying "I have one word for you, son: cloud computing" instead of "I have one word for you, son: plastics." The only reason anybody would buy a ticket would be tradition. We hackers should look at Jira carefully. Is its bloat what we're doing to OUR users? We can certainly dish out the bloat and complexity. Can we take it? Jira makes us confront that.