Quote from: gmaxwell on Today at 02:12:01 PM

Quote from: aliashraf on Today at 10:55:41 AM

Secondly, it is not how we design algorithms for real world problems, being too much concerned about a case that simply doesn't happen

It is when you write systems where reliability is important or when they must resist attack. As much as possible Bitcoin shouldn't have any quadratic operations in validation, even if one is "probably safe most of the time" it just wastes effort analyzing the attack potential (and risk of getting it wrong or some other change violating the safety assumptions).

We hear excuses like this too much from you: reliability, attack resistance, blah, blah.

We are all aware of the requirements, it is mission critical and subject to adversarial behavior, Bitcoin, so what?
Does it imply programmers are free to impose weird assumptions about a crazy attacker who puts like a million Dollar at stake to keep nodes busy validating his block a few more seconds, jeopardizing his premium?

More generally speaking, I've been observing such talking points so many times, there is always an excuse for bad architectural decisions, experts don't buy them, unfortunately average users/customers do.

I regret saying this, but almost every discussion with you eventually reaches this point where you bring up adversarial behaviors, risks, etc.

Bitcoin is a Trillion Dollars network now, and this fact alone is a source of stress and induces hyper sensitivity problems, having a prominent figure who repeatedly reminds devs/advocates of the risks and threats, intimidating them, is just too much. One could classify it as an act of terror Cheesy

I'm not afraid, nobody should, no matter how BIG is Bitcoin it is nothing compared to human's talent and logic, actually it is made by us, human beings, we won't give up developing and improving it ever, it is why discussions should take place and proceed free from superstition and intimidation.