[ddingus]

There is a difference!

The incentive was robust code that would work well, get it done, go to the moon.

Here, machine time is expensive, puts emphasis on code that works, but just barely...

Let's just say NASA would check for the "yup, you are gonna burn some money" case, and reject it.

[FDSGSG]

Well, there's about $3B worth of WETH issued and only about $1.1M worth of those have been lost due to mistakes like the OPs.

I think that people are so unlikely to fuck this up that such a check would be rather pointless.

[ddingus]

Again NASA would disagree.

Money matters a lot. Should this mess endure, people will forever be saying a little bit of gas would have been worth it.

And we've people with six figure arguments as to why such a check makes sense.

The idea of minimal code, focused on speed coupled with financials leads me FAR away from all this.

I will watch with great interest and entertainment.

[shadowgovt]

Should this mess endure, people will just keep using dollars and other fiat currencies.

[FDSGSG]

>Again NASA would disagree.

Yeah, because NASA has been utterly fucked by the congress. Because of politics it's better for NASA to spend 5x the money on 1 reliable spacecraft than to build 5 slightly less reliable spacecraft out of which only 1 fails.

Even if the economics of it don't make sense, NASA can't afford to be seen failing because because politicians will not want to fund them.

I guess my point is that NASA is an exceptionally badly managed entity, not something you'd want to aspire to. (Of course the people working at NASA are not the ones to blame for this.)

>Money matters a lot. Should this mess endure, people will forever be saying a little bit of gas would have been worth it.

That gas would probably add up to more money than has been lost here.

>And we've people with six figure arguments as to why such a check makes sense.

The gas fees of such a check would probably be higher than the losses averted, especially in the long run. And any "losses" are essentially distributed among all ETH holders anyway.

[ddingus]

"The gas fees of such a check would probably be higher than the losses averted, especially in the long run. And any "losses" are essentially distributed among all ETH holders anyway. "

This is also how high toxicity systems get made.

[FDSGSG]

The real solution is better client software. Doing these checks on the client-side is free, that's where they should live.

It would be stupid to put these checks in the contract, they would be very expensive and only help people using unsuitable client software.

[ddingus]

"It would be stupid to put these checks in the contract, they would be very expensive and only help people using unsuitable client software. "

You feel all client software will be suitable?

I don't. There is NO WAY. Lots of people will do ANYTHING for a bit of margin, hoping for volume.

[ddingus]

Sidebar:

How come such a check is so damn expensive?

It should not be.

[FDSGSG]

I agree with you big time. This is hopefully something we'll see addressed by smart contract platforms in the coming years or decades.

Clearly Ethereum is far from a ready product, but I think we can expect to see massive improvements when proof of stake goes live this year.

[ddingus]

Yes! It is all very interesting, but way too rough, IMHO.