Topic: ◈◈Bitcredit ◈◈ Migrating to UniQredit◈◈ - page 54. (Read 284545 times)

If you are talking about the miner, then yeah, please submit a pull request with fixes, that would be the quickest and best resolution.

As for the credit engine , you are also correct, it is meant to be only active for BNs, if you look at the dev branch you'll see that the working proof of concept for how we will handle trust and credit ratings, which is only a job for those offering the services...ie BNs.

NOTE!!!

Things are looking pretty I/O intensive so maybe the smaller PCs will have trouble hosting a fully functional BN. I have since begun a more thorough design and am working on code ~8 hours a day, you should see improvements with the upcoming update, as well as features like the p2plending i have been promising for so long.

Colored coins need work, a lot of it, not sure we can make this update.

the internal miner still only checks NONCE 1 for each random number! this is part of why it is so slow...
aren't the banknodes being paid to figure the credit? shouldn't that be part of what they do? 
  a normal wallet doesn't need to figure others credit...

Ok great you'd have to get an idea of what i am trying to accomplish, so first is to get on the banking branch and do a full fresh sync. You'll see a lot more .dat files in the ratings folder as the client syncs (and also see how slow it's become).

At any point past say 20K blocks you'll be able to analyse the data and maybe make sense of it if you look @ the source.

What i have done, is basically, for every address on the chain, each client maintains a DB of trust and credit ratings as well as pertinent info like average income, expenditure, total earned ever vs total spent, how long the address has existed ...etc etc

This data , along with a more detailed analysis of specific types of transactions, forms the basis of the credit scorecard for BCR. The data and methods used while far more detailed and far more advanced than the previous iteration of the Trust Engine, still are primitive, we haven't started using probability/differentials, and other traditional mathematical/statistical methods. This is where the problem is.... while this is certainly an advancement, i still label it as alpha, it too primitive and clumsy.

As it stands the engine can tel you quite a bit about a users' financial activity but it is in no way enough. I need to extract more data and perform more in-depth analysis to produce better working results if this will ever become a viable reality.

So, I need someone who can help me

1) Integrate useage of a relational DB
2) Populate the DB with every scrap of data we can mine off the chain
3) Integrate high level mathematics and statistical tools
4) Produce a single report detailing each addresses stats/relations
and finally
5) Create a recommendation tool

Already 2 and 4 are in primitive stages, which can for proof of concept be used to illustrate P2PLending.

So.....any takers?

i did some mysql and of course sql language studies at school but i don't know what exactly you need that somebody do... if you want to send a pm to me to see what is needed maybe i can give an help....

p.s. all the studies were done with phpmyadmin but i suppose that the basics are the same right?

Is there anyone good with mysql and c++?

Merging trust rating and credit rating into a single rating. For now many calculations are crude , will be refined.

While one can gain trust, you can also give yourself an advantage by being a BitAngel (donator) , using an active BN, taking up office, and getting voted up by trusted peers.

This is a one time offer:- 100K BCR for a complete write up of BCR, i mean everything.
Another 100K BCR for a complete how to guide for everything in the wallet. 50K extra if you do it with images. These offers expire when i release the update :- 2 weeks.

the orphans part had a little problems when i tried low end cpu.. hope all is ok..
mixing 0.8 and greater source, who the fuck did it? crave? pharma?
pharma scammers messed up more than you can imagine..
lets continue this at 8-bit thread..

great minds think alike, i am going to use the same persistence method i use for databases.

I don't know exactly how it works currently, but when the client checks the block explorer for bids, could it ignore the result and keep the existing data if it had reduced? eg. bids last time checked were x.yz BTC, now it's 0.00 BTC, but there hasn't been a superblock, so we'll stick with x.yz BTC...

Would this add some resilience to the system, prevent a non-responsive explorer at the wrong time screwing up a superblock payout?

Yeah, BTC will have to do, along with a fiat solution soon. Oh yeah...also i'm just super excited that i figured out a nice way to data mine and represent the info in a useable way so the trust engine is getting a re-do as well as the recommendation wizard for BN owners. Also gonna try and root out DS and IX stuff from the code. This may be the first time an update actually reduces the size of the codebase. Gonna delay the next update by an extra week or two so we can have a test only version of p2plending.

I was going to suggest just sticking to BTC as well, it's simple and cheap enough to convert anything else to it. Multiple cryptocurrencies might have made sense if the plan was to keep multiple currencies as reserve assets but if the eventual goal is to have assets in the form of something more tangible then bidding with BTC alone will serve fine.

I really wanted to keep bidding in at least 3 coins in order to make it versatile, but the reality is there are no reliable Dash explorers. Of the ones that always seem to be working, they do not have a viable API. So i regretfully inform you that Dash bidding will cease @ the next update. We will now only work with BTC.

While i wait on some confirmations on some items on the to do list. i have brought together a few lines of code that demonstrate how p2plending will work. the default is micro loans and we will observe the system form there onwards.

No we don't have those problems.

are these issues also affected here?
https://github.com/bitcoin/bitcoin/pull/3514
https://github.com/bitcoin/bitcoin/pull/1382

He is refering to shadowcash as far i could see...
Anyway talking about the coin, i want to know if the older versions will stop working at the next upgrade or not (I synced with an old version weeks ago and soon enough i will reopen the wallet)

Great, i'm actually writing some backups for the bidtracker, hopefully this will increase reliability. Actually yesterday's miner is actually one of the good guys, just that the blockexplorers let us down.

I may have a working solution for the 40 block limit, as this is a protocol change, i want to first put a "warn only" function in the rc and then if we find that it is reliable and won't freeze up the chain, we enable the DoS code.

I changed nothing in those post.

Don't even know how.

EDIT: I understand there are a lot of people looking to do dubious things. But thanks for pointing that out because I just went in and changed all my passwords just in case.

This forum is not known to have been hacked in the past or anything.

I have been around no for a while and have no intentions to do anything to ruin it.

Can you please show where this happened in another thread?

That was me, but I used an electrum BTC address as my DASH bids weren't getting picked up due to blockstrap's unreliability, although the BTC and LTC explorers seem OK.  Couldn't be bothered downloading 250GB of BTC blockchain.

Imported the privkey without issue but no new address / BCR.      I've read that you can import electrum keys into Core without problems, just not the other way around, so not sure what the problem is.     62Di3cmoJo3J619tz7qamJsUezavKkGBNW has plenty of shiny new BCR in it but I've no idea how to access it.


Nevermind, examined the tx and two inputs were used to make the bid, imported the other privkey and the won amount showed up immediately. 

Would be nice to see the 40 block limit in the next update, yesterdays superblock went to a noncompliant miner.

Once again confirming that bidtracking is functional but the block explorers are the problem. I was actually working on getting you guys p2p lending in the next update 

but i'll have to put that on hold and fix the issue of bad explorers. Maybe i can fit it all in one update if i can make the target deadline. Seems today's playout block paid out to the only BTC bidder, lucky guy.

Do you have miningkeys.dat with at least one address with +50K BCR?