3 ms·
"but need to spend time debating how to protect ourselves from marketing “attacks”." I don't have a complete solution to this, but I know I've got two for you:
by jerf 9y ago
"but need to spend time debating how to protect ourselves from marketing “attacks”."
I don't have a complete solution to this, but I know I've got two for you:
1. Beware any solution where the same thing is demo'ed every time. It's a sign that either nobody is actually using it, or it's over-tied to the specific thing being demo'ed.
2. Any time you see a solution that massively outperforms some existing and well-developed solution, you must always ask what was dropped from the existing solution to get that speed, on the assumption that the well-developed solution is probably already pretty optimized to be whatever it is. It isn't necessarily bad to drop things, heck 90% of "cloud" technology basically consists of "dropping the things that don't work well in a distributed environment", but you need to know what those things are.
#2 in particular would have saved a lot of people in the context of MongoDB. It doesn't mean Mongo is a bad choice for everything, or at least, not anymore, but you need to know what you're giving up to get there. (And in particular I'd suggest "We can use this technology without having to have discipline!" is... double-edged at the very least. You may not want a bondage-and-discipline tech at a startup, but tools that offer some gentle-but-solid defaults and guidance may help you focus your cognitive energies less on establishing all the rules from scratch and more on whatever your actual problem is.)