Topic: [HYP] HyperStake | Generous Reward Staking | Advanced Staking Controls & Wallet - page 185. (Read 679332 times)

What is the HYP Wallet rpc port number? is it still 18777??

Its hard to say an exact number, and each coin is different... for example the APEX coin he attacked lets you accumulate unlimited stake weight, HYP only lets you build weight for 30 days per utxo.

So an example from my wallet:
Balance: 4.61m HYP
Total Supply: 89m
% of Supply: 5.1%
Weight: 43,167,600
Statistical Network Weight: 482,756,795 (impossible to actually know network weight, but stats can give us an estimate)
% of Weight: 8.94% (note that unlike other coins without high incentive to stake, my weight % isn't all that different than my % of supply)
Estimated Time to Stake a Block: 16-24 minutes

Summary: although I have the largest known balance of HYP, and almost certainly the largest weight, it would still statistically take me about 20 minutes to stake one block.  This guys requirement is that he needs to stake 6 times in a row and then he will have complete control of the chain, or something like that. This is statistically unlikely for my wallet because difficulty is huge on the HYP network.  HYP does not have the same tiny stake weight that other coins have.

HYP was created to be difficult to stake, and as time goes on it will grow even harder (inflationary supply ensures this). HYP itself should have a very safe buffer against this style of attack, and I would even say that HYP may in fact be one of the safest PoS coins out there.

For other coins... I am not sure. They don't have the same staking activity that HYP does.

Question: in your opinion, what percentage of coins should be staking to provide a good difficulty to secure a coin's network?

Thanks.
Kiklo

Will keep my eyes on this thread. And I am ready to fork if we need to for security.

At the moment I am not too worried. spots issued a challenge to take down MMXIV, which is a much less secure coin than HYP in terms of difficulty to stake. And the OP of the thread said it would be too expensive for him to purchase up the 5-6% of the coin he says he needs in order to reorg the chain. This is actually one of the security measures of PoS that isn't in the code per say... you will lose your own money if you attack the coin.

I also don't think the 5-6% of the coin required would be the same case for HYP.  HYP has way too many active stakers with high weight. Also add in the difficulty of actually acquiring that much HYP, and I think we are in a solid position if what he claims is true. And keep in mind the coin he attacked was an inactive coin with hardly anyone staking it.

There has been some recent events that could affect the security of POS coins.
Their is a group of individuals forming that plan to intentionally attack POS coins.
https://bitcointalksearch.org/topic/cleanup-ill-attack-some-coins-i-owned-apexcoin-for-90-blocks-897493
And it looks like they have been successful in attacking Apexcoin with an attack.
https://chainz.cryptoid.info/apex/

Don't know if it would be a concern for HyperStake, but post it so you would know what they are up too.


 

I am way behind on commits for both HBN and CAPS. But if I can get caught up here soon, and get a chance to play with what you have here, sure I would love to help push this forward more.

So currently on my to do MultiSend to do list:
- Make compatible with former S4C dialog
- Add a sexy table to dialog that lets users edit the vector on the fly
- add custom change address (right now it will return change to the address that staked)
- add enabled/disabled flag to database read/write ( I just tested using an old wallet with 'ats' written to it on the new version, and no errors)
- clean wallet.h


Edit - Tranz let me know if you want to divide and conquer on any of these 

Ahh excellent.

S4C is pretty much removed (will also take it out of wallet.h). Its not on main.cpp so it never gets activated. Hmmm I will test out a wallet with ats on it and see if any errors come up. thanks for the suggestion

It actually does send them all as one transaction with multiple outputs.  Here is an example of a test MultiSend transaction I had yesterday http://hyp.cryptocoinexplorer.com/transaction?transaction=199a112071237f3cbbb0063397f296fee0f7a9c6163f4824e98511f6b3b8a7c3

Just based on reading the code here are my quick thoughts.

You should let the users know they should disable their current S4C before starting up your version. As you do not load the "ats" type anymore and you may get a wallet warning on start up. Although you still saved the s4c types in wallet.h?

The other thing I see a possible improvement would be instead of looping through each address and sending them 1 at a time. You could create an array of addresses and use a sendmany type transaction. This would save a bit on fee for the user, and speed up the send transaction, especially if you are doing A LOT of them. Although it is a small improvement for a fair bit of work, so may not be worth it right now.

Other then that, looks great!

By the way, I am going to add a lot of addresses from the rich list and test sending to lots at once. So don't be alarmed if you get some

Thanks Tranz

First bug found... I did not properly include the lastSentHeight variable so that it won't send twice for a stake that splits... updating it now.

Excellent work PressTab.  I look forward to playing this, and putting it through the wringers. I'll let you know if I find any bugs or suggestions.

I'll update script soon to be using new repo and newest version. Stay tuned.

New HyperStake Wallet Beta Release

This beta release adds:

MultiSend RPC
CoinControl RPC
DisableStake RPC
Making Sync more efficient
Bug and memleak fixes

Please make sure to backup your wallet.dat regularly.

For themes make sure it is located in the same place as your hyperstake-qt.exe.

Please bootstrap for the fastest sync (if you are syncing from scratch). Bootstrap details can be found here: https://bitcointalksearch.org/topic/m.7695751

https://github.com/hyperstake/HyperStake/releases/tag/v1.0.9-BETA


MultiSend RPC

This is a beta release, so help me get out any bugs!! Right now MultiSend vector is saved to database so it will persist through power off. You will however need to disable it each time you start your wallet up (if you don't want it to run) -

I will probably switch the code to leave it enabled through shutdown for the real release of the wallet.

http://hyperstake.wikia.com/wiki/RPC

RPC can be SUPER handy on low power gear since you can pretty well do anything with it and the client daemon that you can do with the qt. (thus taking up less resources)

The problem though is the bin you are using is based off of older code and does not include some more sexier coin control calls. You could compile from source yourself, but this is a very time consuming process for the pi (and a pain in the ass in general). If you want to pursue that route there is help to be had from the community on doing so.




Worse, the most recent bin available is 1.0.6. I wonder if super kasvain could cook another one.

I am guessing the the pi script still links to /presstab/hyperstake but the main repo is now /hyperstake/hyperstake. So you may be on an older version, which had more problems connecting to peers.

yeah.....I don't know what that is so that wont matter for me.  I got it to connect to the network though.  I copy'd the peers.dat and blkindex.dat from my mac wallet and it now connects to the network as well as loading up in about 10 seconds as opposed to a minute and a half

Only if you are making rpc calls. Otherwise, no, they don't make a difference.

I just assume it was the conf file.  I left it on for a few hours while I went to work last night and when I got home it still didn't have any actual connections to the network. I added a few nodes.  We will see if that helps.  Does the rcpuser and password field make a difference as to what is actually entered into them?