[ANN] Nxt :: descendant of Bitcoin

[Fuserleer]

gotta side with snailbrain here....I was intrigued at first by NxT, but some of the posts BCNext has made just don't seem right.

Plus this is a brand new project, built on new code, with these "advanced" features, and launch is Jan....no extensive beta testing or any stuff in place to ensure commercial grade quality.

[1714250770]

1714250770

[1714250770]

1714250770

[cunicula]

gotta side with snailbrain here....I was intrigued at first by NxT, but some of the posts BCNext has made just don't seem right.

Plus this is a brand new project, built on new code, with these "advanced" features, and launch is Jan....no extensive beta testing or any stuff in place to ensure commercial grade quality.

Again, this is a PR fail. It is reminiscent of the Sunny King approach. He is the only person who has released an innovative altcoin. He wasn't any more skilled at hyping it than BTCnxt. If anything much worse.

I have tested his brain and found that it functions pretty well. I am not an expert on coding, but I can differentiate smart and stupid people. My bet is that this will indeed turn out to be quite innovative.

the people who produce commercial grade quality != the people who are innovative

the process works like this:
1) person makes something innovative which is buggy and user unfriendly.
2) person who is more focused on process effeciency turns it into commercial grade quality.

My opinion is that (1) and (2) are rarely the same person. They require different skillsets.

case in point
1) Satoshi
2) Sunny King

[Fuserleer]

How does PR constitute an unknown quality of client through lack of extensive testing?

I'm a damn good coder, I'll hazard I can dance with the best of the best and I've worked on super cutting edge stuff in the past....but...I've overseen and spent a year now developing eMunie, and so far it's been through 6 months of HEAVY testing.

I won't launch it until I am CERTAIN 100% that it is reliable, thats got nothing to do with PR, that my obligation to people that will invest wealth into something I created, and if there is even the slightest problem I missed through skimping on testing to destruction...those people that had faith lose money.

PR doesn't even come into it

Edit: And saying that people who test to commercial grade are not innovative...check out eMunie....nothing in here comes even close to it.

[BCNext]

Guys, could we focus on pure math instead of personal opinions?

This forum is full of haters, I see this.  Do not expect me to answer politely to all these veiled troll questions and imhos.

[Fuserleer]

I'm not hating, I'm just pointing out that you have a lack of testing policy and people are going to entrust you with their wealth.

I think its a perfectly valid question about testing in all honesty

[BCNext]

Okay, what way should we test Nxt?  I can even reveal the source code before the release if most of the investors agree to this.

[Fuserleer]

Well surely that is for you to determine the right use-cases and worst-case scenarios to test for, not for me to tell you, your code works different to eMunie, so different test cases are needed.

Simply, you want to test it to destruction!!  You want to MAKE it break, WAAAAAAY above any load that it will ever receive, then you fix that weak spot and test it to destruction again...keep doing that until either it doesn't break, OR the breaking point is so far outside of normal use it's irrelevant.

For example, for one entire week with eMunie, we generated 70k transactions per day, on a 10 node network....it took it all, just about.  This was after many iterations of the same test, pushing the failure point out further and further.

That will NEVER happen in real world scenarios where 10 machines are creating 70k transactions per day even at VISA scale levels (I know you don't like VISA scale references but you gotta think about it).

Other things, connectivity issues you cant control, so, pull the plug at a crucial moment and make sure that it continues with no errors, any DB's you use kill the drive halfway through a write and make sure it recovers etc etc

You're a software engineer right?  You should know this stuff!

[BCNext]

I do know how to test software.  Right now 3 blockchains are being generated by 3 different computers with intensive transaction flow.  Sometimes I switch the network on and watch blockchain re-organization (longest chain wins).

At the beginning of December we will start the final test when new features is not be added and I have 3 weeks to fix found bugs.

[Fuserleer]

Well, I hate to be the bearer of bad news, but with all the features you've announced, with a 3 months development time INCLUDING testing and bug fixes...well, I think you'll find you wont have enough time and there will be some real nasties down the road with lots of grumpy investors.

[BCNext]

All the features?    They require a little coding.  Nxt core is completed, adding new features on top of it is very fast and easy.

The only thing that slows me down is the client.  I have to cooperate with the client developer.

[Fuserleer]

Thats just coming across as arrogant.

So a REAL P2P exchange, that many many smart people have been looking at yet not done, is trivial in your opinion?

[BCNext]

Let me show this using Nxt DNS as an example.

What I have to do to add Nxt DNS

1]  Add a new subtype of messages -- SUBTYPE_MESSAGING_DOMAIN
2]  Add code for processing "resolveAlias" request -- ~30 lines of code

That's it.  Now an Nxt node processes transactions that set domain aliases and requests from JavaScript that asks for real address.

[Fuserleer]

What about management, entry clashes, expiry, abuse handling, spoofing and all that kinda thing?

I wish you the best of luck I really do, but I just see a whole load of pitfalls ahead that you don't seem to realise are right there.

[BCNext]

So a REAL P2P exchange, that many many smart people have been looking at yet not done, is trivial in your opinion?

REAL P2P exchange from the scratches would take a lot of resources.  The same on the already implemented code required 2 days of coding.

[BCNext]

What about management, entry clashes, expiry, abuse handling, spoofing and all that kinda thing?

Ah, I see your point now.  Nxt lets to trade assets represented by colored coins for NXT in atomic transactions.  All the rest is up to asset issuers.

[Fuserleer]

So it isn't a real P2P exchange that you are implementing as we are?  So what is it, a proxy to a website built in?

And yes, it DOES take a lot of resources, time and effort to do....I know this

[BCNext]

It is real.

Here is an example

I want to run a business.  I issue an asset (let's call it Krusty Krab tokens - KKT).  I set quantity of KKT to 1000.  I announce that I sell KKT for 20 NXT each and pay 5 NXT monthly interest within next 6 months.  I create an ask order -- 1000 KKT for 20,000 NXT.  You create a transaction that sends me 500 NXT for 25 KKT.  It's up to you what to do with your 25 KKT (you can place an ask order and sell them for 27 NXT) but each month I will check the blockchain to see who owns KKT.  And I will send 5 NXT for each KKT to those who owns them.

[Come-from-Beyond]

Look at https://bitcointalk.org/index.php?topic=303898.msg3253052#msg3253052

[snailbrain]

and thanks for those 20btc everyone

how secure can Nxt be when he can't secure his own computer?

joke.. (a bit)

[Come-from-Beyond]

and thanks for those 20btc everyone

U wouldn't get them back anyway.

I'm that guy who is developing the client, so I will publish updates until BCNext is back. If I disappear one of the 2 remaining guys will continue the work until Nxt is released as planned.