Topic: [ANN][SIM] Simcoin - A Simple Coin - page 65. (Read 157733 times)

But I'm only saying that because I'm in the tank.

Thanks! I'll do my best.

yeah, well but 75 isn't much better either

Thanks! I'll do my best.

----

And with 75 investors we surpassed NXT. It's good, because people still bash NXT for small number of initial investors.

Total investments at this point: 43 BTC

Invested 1.5 BTC because your tech really gets to the heart of why cryptos haven't spread that much: security and speed. Add ease of use to your "paranoid" security and you'll have one helluva coin.

Allow me to introduce myself to everyone as Investor #151.

Running three instances gave 3x 2500 keys/sec (7500 total) using around 35% of six cores (2 virtual) - this was the best I achieved, running four or more instances gave no overall increase.

Machine was essentially idle during tests.

I've updated the code for error-protected addresses that I did for NXT and converted it into the new format.

As a bonus I also wrote a simple address generator. It's not user-friendly, it's just for you, early adopters

https://simcoin.info/address_gen.zip

It's a simple console exe that generates random addresses and stores them in a CSV file.

It doesn't look for user-defined words because it's much faster to check a word against millions of pre-generated addresses than to generate those addresses again. What if the next day you need another address with a different word? You just try to find it in your database of addresses, instead of generating them all over again. This also helps if you just want to pick a better looking random address.

Don't go crazy with it, if some issue comes up - I will not hesitate to change how keys are generated.

The program will append new addresses and passphrases to "out.csv" file it creates on the first run. Make sure nobody else can read it.

If you use it, I would appreciate if you post here how many addresses per second you got and what kind of CPU your machine has. My, rather old, 2 GHz PC generates them at the rate of about 1400 keys per second. This means I can find a 6-character word in 4-8 days. The code is not particularly optimized, so the speed can be significantly improved later.

If your CPU has multiple cores, you can run several copies of the program, just make sure you run it from different folders. And leave one core free for the rest of the system.

To exit press Ctrl-C.

So you would have more people to dump to?

No way. This is not "get rich quick" scheme. It's "wait a bit, get very rich" scheme

I was getting about 1900 Keys/sec on a 3 year old Intel dual core 2Ghz laptop.

OK...  too busy to run a clean test and/or change computers...  but with 35 tabs open in Chrome, watching a video, and a traditional crypto wallet running I get 1500 keys/sec with Intel® Core™2 Duo Processor T7250 (2M Cache, 2.00 GHz, 800 MHz FSB).

It increases the size of the database anyway. If you have data representing that information, you have to store that data. It is as simple as that. You cannot have information without it corresponding data.

It will inevitably increase bandwidth usage. Question is: who much is acceptable?

As some of you know, I was the most vocal in NXT community about lack of checksums and helped develop Reed-Solomon protected addresses, which are yet to be implemented in NXT. Well, in Simcoin they will be there from the start and will be shorter - 12 characters instead of 17 (while providing the same level of protection):

There will be no "darkNXT", i.e. no ability to send money to an arbitrary address and lose it. There will be no genesis account where you can send money to destroy it. There will be no reduced security for new addresses as in NXT.

In NXT and Bitcoin the best way to handle customer deposits is to generate unique addresses for them. This increases the size of the database. In Simcoin transactions will be done similar to Ripple - with a short text field where you can put some special code to link this deposit to your account. Or you can register with your address as a username, which is possible because they are so short and mnemonic.

There will be ability to send messages, similar to NXT Arbitrary Messages, but it will be done right: there will be a type tag and all messages will be automatically and transparently encrypted. In fact, ALL traffic will be automatically encrypted to help with privacy, security and to make it harder for governments to filter our packets.

I also will not repeat the same mistake BCNext did: he abandoned his creation to an idealistic vision of completely distributed and "headless" development. It was sad to see great ideas of NXT disintegrate like this. Instead, the source code will remain closed until my vision is implemented exactly as I planned it to be. I might be wrong in some decisions, but at least I will be coherent.

Transaction speed will be a priority. The goal is ambitious: to achieve <1 sec confirmation. Complete confirmation, no backsies. Either your transaction is confirmed and engraved in stone in less than 1 second, or it will be rejected and you will have to try it later.

I believe it is possible. Here's a network simulation video: