Topic: [BBR] Boolberry: Privacy and Security - Guaranteed Since 2014 - page 380. (Read 1210752 times)

there appears to be a small number (1?) of miners that have an overabundant supply of BBR

it suggests a proprietary miner, so the sooner we get a way to nullify this overabundance, the sooner the price will stabilize

Until then, to pay for the mass mining costs, daily sales to recoup the costs are needed by this theoretical big miner.
BBR tech is clearly the best and if any cryptonote is to solve the blockchain bloat problem (it does loom ahead), it will be boolberry

James

P.S. Big miner guy, plz fill my order on polo!

not really man, its just startign with GPU miners coming in

Dear friends!

Thanks to everyone who was involved in testing stratum poolserver! Test server found 5 blocks and was tested under hight load. Now we have working stratum pool server + miner. It's able to work with trustedshares option disabled , but if you want to safe your cpu resources - leave this enabled.

This is still not perfect, have a lot of TODOs(and i guess some bugs), we going to improve it with saving/loading scratchpad to file (local chache), uploading inital scratchpad from thirdparty storage (sourceforge or something else). Next version will be in next few days.


PS: We don't need hashrate on this server anymore, so you could disconnect your miners.

Is there a BBR blockchain explorer?

James

From the cryptonote site: "In our version the sender uses the receiver's public address and his own random data to compute a one-time key for the payment."
As long as BBR did not change this, it seems that there is no need to transmit the paymentID.
Of course, the problem is that the receiver of the funds wont have a way of knowing who sent it!
Since my use case wont care about that, it is good

almost dead . leave away.

Is there any need to keep going back to genesis block all the time? Usually, you process the blockchain and once you are caught up you only need to keep a much smaller amount around. So a set of memory mapped (month long) subsets of the blockchain would allow total reprocessing at HDD read speeds.

Ive build caching systems that can also solve this as long as I know the furthest back I need to keep things. Then I can just purge the cache of blocks we will not need (until next restart)

James

(1)  I think it is possible to do such an analysis IF the payment ID is reused, unfortunately.  Someone should correct me if I'm wrong.  If it's only used once, then it's safe.

Payment IDs are wholly unnecessary, but there's a speed issue with having huge #s of destination addresses (and the wallet doesn't support it yet).  In general, though, one could have a lot of dests - one just needs to do more work analyzing new transactions.  It shouldn't be too horrible.

For blockchain - it's really beyond memory mapping;  the right long term solution is to incorporate something like LevelDB as a true external database.  The scratchpad doesn't need to be as big as the full blockchain, so it can remain memory mapped for mining.  But even mmap -- on 32 bit systems it's limited to 32 bits.

1. My concern is that in order for Alice to Pay Bob using payment ID, if that ID (random or not) is public information and the ID ends up in the blockchain, couldnt an attacker start creating a correlation predictor based on payments to the public ID? Also it is possible to send payments without any payment ID at least it worked in my tests.

2. Sorry, I am still learning cryptonote so I just assumed the dest address is there, I guess it is the payment ID address that is there and unless that was communicated privately, info is leaked

3. After fiddling with things all weekend (half a dozen places that assumed non-zero fees!), I did manage to get it working. My use case wont have any mining, just very low difficulty (big sleep in miner loop) so that we get blocks every minute.

After problems in every step of the process I finally got it to accept a no-fee tx and then finally release it tx from the mempool and once that worked, all seemed pretty happy. I've already premined 99.2% and when I get to 99.999% the block reward will be small enough to discourage any cheating. The purpose of pNXT is to map 1:1 to NXT so that people can send NXT from its web GUI and have it use boolberry cryptonote pNXT for the transfer.

James

P.S. Good point about 32 bit mode, it seems we need to use a memory mapped blockchain file for it to have a chance.

P.P.S. zerofees should encourage a lot of tx, which provides best privacy, at least from a practical point. I guess I need to think about regular purges. With pNXT it is a matter of forcing everyone to redeem it and reset the blockchain, but that does sound rather messy...

oh ok. still have issues with 'search'.
can u check?

bbr.extremepool.org is no more.

New pool with new code is http://boolberry.extremepool.org

mining at http://bbr.extremepool.org/ but when trying to 'search' it get freezed and never finds me

Oh, sorry.
Payment id is just an array of 32 bytes binary data, randomly generated by receiver(i believe it randomly generated). It is encoded as hexstring to be shown(in exchage web interface) and than it decoded back into 32 bytes of binary data to be placed into transaction extra.
In other words it's just an ID info placed into transaction

It seems to pick up the number of CPUs by itself, so you don't really need the -t option.

(I generally use -t `nproc` for all the various cpuminers, because some of them either don't pick it up, or assume things about hyperthreading and AES-NI that don't apply to everyone. Of course try different numbers, but usually the default is the best.)

I read jl777's question differently:

Is the payment ID in cleartext, or is it encrypted so that only the intended recipient can decode it?

And if it's encrypted, I assume it's done in such a way that it does not enable linking of multiple transactions using the same payment ID?

1. Payment ID is generated by payments receiver and included into transaction as a prove that sender is really created this transaction. This is not the best solution, as i know this was made as workaround. I have ideas how to go without payment_id but sender need to communicate with receiver directly.

2. I don't know what you mean here, cryptonote blockchain don't have user addresses in it. fake_outs_counter select amounts to be mixedin into ringsignature (output keys is used for this).

3. Fee can't be 0(network rule). Also, pools will not gonna mine transactions with low fee, and according to last commits will not even relay it. DEFAULT_DUST_THRESHOLD also can't be changed  since it used in blockreward formula. Don't know how you gonna avoid "dusting"... i've removed dust from block reward, but dust will appear from exchanges or bbrdice, or any other payment services.

Extreme Pool

Directions for mining using stratum (Linux only):

1.  Get and install the new minerthe new miner  
cd Cpuminer-multi

2.  Now that the miner is downloaded and installed, using the following command to mine on Extreme Pool.

I have vps host with macos, when script will be available i'll integrate it into project build system to get builded automatically.

Even if we build 32bit version of BBR, i can't grant that this will work fine since there a lot of places that potentially buggy in 32bit.
Another reason is that 4 GB will be not enought for normal work after say ... 2-3 months. As i know monero already had this problems.
That's why i'm not going to support 32bit. But, probably i'll fix this issue after i close all critical tasks.
But, if someone will build 32-bit, please, before announce - run coretests and unittests to ensure that currency core is not broken.

Extreme Pool

Since the test pool has found blocks now, and my independent tests confirm the pool and new stratum miner should be working fine.
I am announcing the new boolberry public pool is open @ http://boolberry.extremepool.org

If you care currently mining @ bbr.extremepool.org please move your miners over to http://boolberry.extremepool.org.

Get the latest stratum miner for boolberry!

Enjoy!