Well, as with all homebrewn solutions, you don't know if you are reinventing the wheel until you're done. At first, it always starts with "the current solutions

Well, as with all homebrewn solutions, you don't know if you are reinventing the wheel until you're done. At first, it always starts with "the current solutions that are available do not fit me, but I still could use them to achieve what I want". There was nothing forcing 37signals back in 2004 to roll their own framework in order to support developing their apps, but they did anyway. And for every Rails out there, there are thousands of internal frameworks with big ambitions that just turned out to be inferior to what's already available. You just can't know it when you start developing. It takes a bit of ego and ambition to go that path, but sometimes it pays off. And my guess is that if it paid off in the past, you're more likely to try it again.