Hacker News new | ask | show | jobs
by bluejekyll 3295 days ago
But code is written by humans... have you ever seen code with no bugs? In 20 years of professional development I have not.

Given that humans are imperfect, and could even potentially act in bad faith, isn't it reasonable to have an exception clause? I get the argument to not have one; that it's impossible to have favorites and central figures manipulate the system, but nothing is perfect.
3 comments
kalleboo 3295 days ago
> isn't it reasonable to have an exception clause

It absolutely is. Which is why you don't say "code is law". Which is why Ethereum is dumb.

link
nebabyte 3294 days ago
If you think laws are written without exception clauses, I have a bridge to sell ya.
link
kalleboo 3294 days ago
Laws are interpreted by courts.

There is no court of Ethereum aside from "can I convince the developers + 50% of miners to do a hard fork"

link
jjeaff 3294 days ago
I have a solution. Perhaps you take the etherium users, and they can vote and elect arbitrators, let's call them judges. Then those judges can hold "court" and a selection of etherium users would act as a "jury" to decide on how to handle exceptions. Of course, we will also need to appoint people to enforce those laws. Maybe we should start with a constitution to get things all lined out...
link
bluejekyll 3294 days ago
Actually I would suggest starting with something less strong than a "constitution", something that just defines the federation of etherium exchanges in broad strokes, call that the "federalist papers" or something.
link
microcolonel 3295 days ago
https://github.com/seL4/seL4 https://www.mitls.org/
link
bluejekyll 3295 days ago
sel4 is amazing, but it is not bug free:

https://github.com/seL4/seL4/issues/36

It might be close... for miTLS I don't have access to the issues, but let's assume it's bug free now for sake of argument; it hasn't always been bug free, that is in earlier unproven releases.

"Program testing can be used to show the presence of bugs, but never to show their absence!" - Edsger Dijkstra

link
microcolonel 3295 days ago
Minor nitpick: the x86 port was never verified, the 32-bit ARM one was.
link
WalterSear 3294 days ago
Not while claiming the opposite, it's not.
link