[derefr]

> As one technical cofounder to another, I think the most important thing is for you to understand the business side.

But the OP isn't asking for help themselves, and has said nothing about needing help themselves. Maybe they already understand the business side?

[kristiandupont]

Well, it could be a case of "My wife gets upset when I cheat on her. How can I make her understand that it's not a big deal?".

I am not saying that it is but I do think that Stephen Coveys famous "seek first to understand, then to be understood" quote holds a lot of truth. In order to communicate your own perspective on something to someone, you need to have a sense of how theirs differs to yours.

[derefr]

Sure, but then... what's the second step? Because that's what the OP was asking for here.

IMHO it's kind of against the Principle of Charity to assume that someone asking a question about how to do part N of a multi-part process, hasn't already figured out how to do the previous N-1 parts of that process. They must have, to get to the point where they're motivated enough to reach out for help with part N.

In programming terms: if I'm writing a compiler, and I'm asking about codegen, the Principle of Charity would suggest that I've already solved enough other compiler-writing problems that I have an input to feed to a codegen pass!

Ignoring the Principle of Charity in cases like this can lead to some really useless/unproductive advice.

For example, I've been writing a research decompiler for an ISA that has less embedded information than usual (e.g. no reified CALL/RET ops, just plain branch instructions.) It needs some weird heuristics to derive subroutines from coroutines, and to properly scope variables. This means that I can't immediately derive an interprocedural SSA form for the variables, in order to help calculate a dominator tree. So I want to ask people how to calculate a dominator tree from stack variables, because having something like that would help as an input into those heuristics.

But every "solution" on StackOverflow-like sites insists that the context of the problem—the assumptions it makes—are impossible/invalid. They try to mentally rewind to step 1 themselves, and then can't understand how I solved it (because I didn't write out the textbook worth of novel algorithms I used to solve step 1 in the problem description), so they don't even believe there is a step 2. Rather than just... attempting to help me with step 2. Which doesn't actually require any understanding of step 1, other than its outputs.

It's sort of the opposite of the XY problem. Rather than the parent problem X being best solved through a different strategy that avoids problem Y, people believe that the parent problem X itself is unreasonable, because they model an X that is solved through Y as requiring that the solution also flow through a previous sub-problem Z; and they think that Z is something that people just don't bother to do, so nobody could actually be asking in good faith about XY.

I've also suffered through this with doctors, who literally won't believe me when I list off the entire sequence of steps they would try and say that I've tried them to no avail. I only go to the doctor when I run out of ways to try to solve the problem myself, so if I'm here, it's because the problem is a zebra instead of a horse! A horse, I would have fixed on my own!