On the way to the post-Ethereum world

[kushti]

A note: I worked for Nxt(core dev) in 2014 and now working on Scorex ( https://bitcointalk.org/index.php?topic=1103640 ).

Nxt core devs were sceptical about smart contracts inclusion into core. I myself was also sceptical about some things included into Nxt core  

Bitcoin developers switched off many opcodes for a reason ( http://bitcoin.stackexchange.com/questions/25225/what-was-the-vulnerability-in-v0-3-4-that-allowed-an-attacker-to-steal-coins ). Then Vitalik complicated security conundrum with its "Turing-complete" approach.

Now Ethereum guys are making the biggest error in their career going to do a hardfork, because not all the contracts are equal and some of them are too big to fail. There are some talks(copied from Bitcoin community) on "miners will decide", but it is unlikely miners will have any reasonable discussion. Instead, they will just follow the "digital Lenin".

It is time to go another path. The core must be as simple as that. But it also need to be powerful enough to support protocols and applications built outside the core.

There are many designs possible, and there are many open questions and concerns about them. Nevertheless, as the topic is going to be hot, I would like to propose one design in July or August (after Scorex 2.0 release).

[CIYAM]

It should be noted that Automated Transactions (AT) has been running for over 1 year on two separate blockchains without incident (and yes I will accept the fact the two blockchains involved don't amount to a huge amount of capital but certainly ATs have been handling amounts above 1 BTC regularly).

There is a reason why we didn't provide a high-level language and developed each "smart contract" ourselves (with literally months of testing).

Getting smart contracts "right" is a very difficult thing to do so you need the "attack surface" to be as small as possible.

[iamnotback]

There is a reason why we didn't provide a high-level language and developed each "smart contract" ourselves (with literally months of testing).

https://bitshares.org/technology/industrial-performance-and-scalability/#smart-contracts

[kushti]

There is a reason why we didn't provide a high-level language and developed each "smart contract" ourselves (with literally months of testing).

Getting smart contracts "right" is a very difficult thing to do so you need the "attack surface" to be as small as possible.

Interesting. And what do you think about Solidity? Isn't it a language helping developers to make deadly errors?

[CIYAM]

Interesting. And what do you think about Solidity? Isn't it a language helping developers to make deadly errors?

For a start Javascript is about the worst possible language you could use to code anything in (so if Solidity has much in common with that then it is doomed from the start).

But in more general terms I think the problem is that "any high level language" is at this stage a bad idea as people simply don't realise how easy it is to find and exploit bugs in code (so you don't want anyone but experts writing such things).

The approach that was taken with AT was to simply do "machine code" (now assembly) and get every nuance correct in each smart contract.

[iamnotback]

For a start Javascript is about the worst possible language you could use to code anything in

Maybe you are not aware of ASM.js.

[CIYAM]

For a start Javascript is about the worst possible language you could use to code anything in

Maybe you are not aware of ASM.js.

Not sure how that is even slightly relevant - the problem with .js is its dynamic type system and processing (static languages are far easier to secure).

[spartacusrex]

For a start Javascript is about the worst possible language you could use to code anything in..

x1000..

[iamnotback]

For a start Javascript is about the worst possible language you could use to code anything in

Maybe you are not aware of ASM.js.

Not sure how that is even slightly relevant - the problem with .js is its dynamic type system and processing (static languages are far easier to secure).

C/C++ code can be compiled to ASM.js.

ASM.js removes those dynamic types and GC.

ASM.js is statically compiled.

[CIYAM]

C code can be compiled to ASM.js.

ASM.js removes those dynamic types.

I'm not really sure why that would be better than AT (as it is about 10x as complex).

And having high-level languages such as C/C++ is not a good idea as I pointed out.

[iamnotback]

C code can be compiled to ASM.js.

ASM.js removes those dynamic types.

I'm not really sure why that would be better than AT (as it is about 10x as complex).

Write once, run every where.

[CIYAM]

Write once, run every where.

AT is "write once" but in any case that stupid "meme" was from Java and is generally parodied as "write once, debug everywhere".

[iamnotback]

Write once, run every where.

AT is "write once" and that stupid "meme" was from Java and is generally parodied as "write once, debug everywhere".

I am happy that you don't value the WWW. Figure it out.

[CIYAM]

I am happy that you don't value the WWW. Figure it out.

Plenty of times my browser crashes due to .js (and I use NoScript always).

(if you are trying to be clever then you might want to try a bit harder)

[iamnotback]

I am happy that you don't value the WWW. Figure it out.

Plenty of times my browser crashes due to .js (and I use NoScript always).

(if you are trying to be clever then you might want to try a bit harder)

I see you don't understand what "no GC" means. Goodbye wannabe expert polyglot coder.

[CIYAM]

I am happy that you don't value the WWW. Figure it out.

Plenty of times my browser crashes due to .js (and I use NoScript always).

(if you are trying to be clever then you might want to try a bit harder)

I see you don't understand what "no GC" means. Goodbye wannabe coder.

What a "wanker" you are.

Not sure why you just come here to troll me with nothing intelligent to contribute at all.

Seems like your supposed IQ has gone on holiday (about 10 years ago I think).

Why don't you do us all a favour and actually "fuck off" for good?

(the only supporters you have here are your sockies)

[iamnotback]

What a "wanker" you are.

Every time I win the technical argument, you refuse to mea culpa. Stubborn ox.

Next time try to not making sweeping statements about things you don't completely understand, e.g. JavaScript.

[CIYAM]

What a "wanker" you are.

Every time I win the technical argument, you refuse to mea culpa. Stubborn ox.

Except you have never won a single technical argument against me (you idiot).

Try actually arguing rather than insulting next time (but I know you won't as you never do).

[iamnotback]

What a "wanker" you are.

Every time I win the technical argument, you refuse to mea culpa. Stubborn ox.

Except you have never won a single technical argument against me (you idiot).

Wake me up when you realize the relevance of "no GC" to your point about the browser crashing.

[CIYAM]

Wake me up when you realize the relevance of "no GC" to your point about the browser crashing.

What on earth does GC have to do with that?

Try answering this question rather than another insult if you possibly can (as there is no sensible answer I am pretty certain that you won't).