[ANN][RIC] Riecoin: constellations POW *CPU* HARD FORK successful, world record - page 137.

gatra

hero member

Activity: 583

Merit: 505

CTO @ Flixxo, Riecoin dev

Quote from: aamarket on September 08, 2014, 04:07:44 AM

Everybody's so excited about 'superblock' - so I try to be a bit negativist.

I really doubt 30 minutes for record breaking 6tuple, the Riecoin network diff was more than 1700 ( I guess ), now it is less than 1450.
Please bear in mind it is exponentially difficult to find 6tuplets with higher diff ( I posted some rough estimation before ).

Is the network 'slowdown' for (possibly much more than) 30 minutes - and the need for everybody to update his clients -
really worth the risk of 'hard fork' problems ?

Do you have any math support for "30 minutes time to find superblock" other than "At current rates, I expect it would take about 30 mins" ?

I like Riecoin, and I wish all the best to it.

Sure! there's always math to be had when it comes to ric.
The assumption is that the difficulty raised to the ninth power is linear (6 because according to Hardy-Littlewood's k tuple conjecture the average space between 6 tuples is O(log^6 p), and 3 because testing for primality takes about O(log^3 p), so 6+3=9). This has been working nicely: the average time between blocks is about the expected 150 seconds, which means the assumption is working so far.

Current diff is 1441, but we want a block with 1857. So (1857/1441)^9 = 9.8 times harder. If we now have 2.5 minutes then 9.8 times harder would take 24.5 minutes (24.5 in average! it should be between 18.5 and 32 mins with 90% confidence, we have about 5% chance of it being more than 32 minutes and 0.2% of it taking more than an hour - this is calculated assuming we are following a Poisson process).
When difficulty was 1700, it would have been 1.7 times harder, meaning that choosing a block interval target of 5 minutes instead of 2.5 would have made us break the record a couple of months ago.

aamarket

sr. member

Activity: 259

Merit: 250

Everybody's so excited about 'superblock' - so I try to be a bit negativist.

I really doubt 30 minutes for record breaking 6tuple, the Riecoin network diff was more than 1700 ( I guess ), now it is less than 1450.
Please bear in mind it is exponentially difficult to find 6tuplets with higher diff ( I posted some rough estimation before ).

Is the network 'slowdown' for (possibly much more than) 30 minutes - and the need for everybody to update his clients -
really worth the risk of 'hard fork' problems ?

Do you have any math support for "30 minutes time to find superblock" other than "At current rates, I expect it would take about 30 mins" ?

I like Riecoin, and I wish all the best to it.

northranger79510

sr. member

Activity: 308

Merit: 250

Riecoin and Huntercoin to rule all!

Keep up good work Gatra! Cheesy

bsunau7

member

Activity: 114

Merit: 10

Hi,

Another 5% from making sieve2 code ~30% smaller and starting to optimize memory access (pre-fetching). Once again to see this improvement you'll need to increase the sieve size (-m). I am running it at 15million which means I am sieving against the first 970k primes.

Link is, as always; https://mega.co.nz/#F!6Y8ywRoZ!5ogMJEc2Teq5ckt9WClNpg

Note: This runs the chip hot, very hot. The CPU is sitting on 94-104c (thermal throttles at 104c) even with a heat sink and a 12v noctua fan. Stock systems will probably clock right down to save themselves (and thus give bad performance).

I do like the idea of a monthly super block, so much so I am considering putting a bounty of the first super block not mined by a pool.

Regards,

--
bsunau7

esuncloud

member

Activity: 98

Merit: 10

Quote from: gatra on September 04, 2014, 02:01:55 PM

Regarding the superblock thing, I think that the easiest would be to force one block (let's say block 122000) to have the desired difficulty and then return to normal. At current rates, I expect it would take about 30 mins (expected time) to solve the world-record block. Less if we are lucky, more if we aren't. This would require everyone to update to the new client but nothing else would need to be changed.
There will be a delay on transactions, but only for less than half an hour. Think of it as "expected downtime". We could make that block have a reward of 600 to accomodate for the extra difficulty.

What do you say?

I am surprised that it will only take 30 mins to solve the world-record block. It is absolutely worthy to do a hard fork for this idea.

fairglu

legendary

Activity: 1100

Merit: 1032

Quote from: gatra on September 04, 2014, 02:01:55 PM

Regarding the superblock thing, I think that the easiest would be to force one block (let's say block 122000) to have the desired difficulty and then return to normal. At current rates, I expect it would take about 30 mins (expected time) to solve the world-record block. Less if we are lucky, more if we aren't. This would require everyone to update to the new client but nothing else would need to be changed.
There will be a delay on transactions, but only for less than half an hour. Think of it as "expected downtime". We could make that block have a reward of 600 to accomodate for the extra difficulty.

What do you say?

Great idea!

Actually, why not go one step forward and have regular world-record superblocks?

Say once a week, first super-block would have minimum world record difficulty (or maybe below, to make sure the stall is reasonnable as expected), then next week, super-block with world record difficulty + a tiny bit, then + 2x tiny bit, etc.

A weekly reasonable "stall" in exchange of weekly world records, exploring the prime frontier, going boldly where no blockchain has gone before.

dga

hero member

Activity: 737

Merit: 511

Quote from: gatra on September 04, 2014, 01:52:17 PM

hi!

I tryed the uninitialized variable fix and yes, it now seems to work well under windows. I also made the stratum protocol properly reconnect after losing the connection to the pool.
Source is (under the stratum branch) here: https://github.com/gatra/fastrie

dga: would you please merge this?

thanks and regards,
gatra

Yup. I'm traveling right now but will try to get this done on Sunday.

Ellieo

member

Activity: 60

Merit: 10

Quote from: gatra on September 04, 2014, 02:01:55 PM

At current rates, I expect it would take about 30 mins (expected time) to solve the world-record block.

What do you say?

I don't mind even if it takes a day to find the world-record prime block. The bigger prime the better.

LEAP FORWARD RIECOIN!!!

The Dark Knight

newbie

Activity: 2

Merit: 0

Quote from: gatra on September 04, 2014, 02:01:55 PM

Regarding the superblock thing, I think that the easiest would be to force one block (let's say block 122000) to have the desired difficulty and then return to normal. At current rates, I expect it would take about 30 mins (expected time) to solve the world-record block. Less if we are lucky, more if we aren't. This would require everyone to update to the new client but nothing else would need to be changed.
There will be a delay on transactions, but only for less than half an hour. Think of it as "expected downtime". We could make that block have a reward of 600 to accomodate for the extra difficulty.

What do you say?

Great idea in my opinion!

gatra

hero member

Activity: 583

Merit: 505

CTO @ Flixxo, Riecoin dev

Regarding the superblock thing, I think that the easiest would be to force one block (let's say block 122000) to have the desired difficulty and then return to normal. At current rates, I expect it would take about 30 mins (expected time) to solve the world-record block. Less if we are lucky, more if we aren't. This would require everyone to update to the new client but nothing else would need to be changed.
There will be a delay on transactions, but only for less than half an hour. Think of it as "expected downtime". We could make that block have a reward of 600 to accomodate for the extra difficulty.

What do you say?

gatra

hero member

Activity: 583

Merit: 505

CTO @ Flixxo, Riecoin dev

hi!

I tryed the uninitialized variable fix and yes, it now seems to work well under windows. I also made the stratum protocol properly reconnect after losing the connection to the pool.
Source is (under the stratum branch) here: https://github.com/gatra/fastrie

dga: would you please merge this?

thanks and regards,
gatra

F22Raptor

full member

Activity: 126

Merit: 100

i haven't seen this coin for a while,god bless me that this coin isn't dead.i will read posts to decide invest or not

nonce-pool

full member

Activity: 149

Merit: 100

Quote from: curefrankosflue on September 03, 2014, 05:46:31 PM

Hey nonce-pool, thanks for answering. in case you have time to answer again I would appreciate your help.

I tried what you suggested and managed the miner compile on Ubuntu 14.04. I got an executable xptminer file, although the miner didn't mine.
It got just 0s for all chains and no shares even after some 30 minutes. Of course I triplechecked whether I was mining to the right address and the other syntax was right.

2ch/s: 0.000 3ch/s: 0.000 4ch/s: 0.0000

I switched back to Ubuntu 13.10 as some other things did not work properly and Ubuntu 13.10 has done more or less well so far.
I tried to download and compile the miner there.

All steps went well until the command:
LD_LIBRARY_PATH=/usr/local/lib make -j4

As far as I understand the line LD_LIBRARY_PATH.... compiles the xptMiner to make an executable file.

It compiled many of the source code files but in the end I got the error message:
/usr/bin/ld: cannot find -lcrypto
/usr/bin/ld: cannot find -lssl
collect2: error: ld returned 1 exit status
make: *** [xptminer] Error 1

maybe I am missing some other libraries? or are these missing source code files of the xptminer? as I said I'm a newbie to Linux

Thank you very much for your help.

Michael
ana: curefrankosflue

Looking back through the bash history, didn't really need to do anything special to get it to work on 14.04, just follow bsunau7 advice for 13.10 and that should get you going.

also adding to your command line when running increases performance considerably (not sure what the best number here is, still trying)

Code:

-s 50000000

bsunau7

member

Activity: 114

Merit: 10

Quote from: curefrankosflue on September 03, 2014, 05:46:31 PM

It compiled many of the source code files but in the end I got the error message:
/usr/bin/ld: cannot find -lcrypto
/usr/bin/ld: cannot find -lssl
collect2: error: ld returned 1 exit status
make: *** [xptminer] Error 1

maybe I am missing some other libraries? or are these missing source code files of the xptminer? as I said I'm a newbie to Linux

Thank you very much for your help.

Michael
ana: curefrankosflue

Try installing libssl-dev if ubuntu 13 does not have it. As root:

Code:

# apt-get install libssl-dev

Next try setting LD_LIBRARY_PATH before you run xptminer. It sets the path which the executable will look for it's libraries. Doing it on the make line sets it for that execution of "make" and not the executable created.

Something like

Code:

# export LD_LIBRARY_PATH=/usr/local/lib
# xptminer

Should work. The export makes it active for the life of the shell. There are a few ways to make this "sticky" my preference is to add it into dynamic linker, there are are lots of hits on google for this, search for something like "adding /usr/local/lib to library path" should give you a pretty good start!

Regards,

--
bsunau7

cinnamon_carter

legendary

Activity: 1148

Merit: 1018

It's about time -- All merrit accepted !!!

Huge thanks to dga for his work on the miner which I am running and having a bit of fun experimenting with.
Also thanks Gatra for making it all possible by creating Riecoin.

Most people have no idea how much work and costs are involved in coding, testing and creating these things.

curefrankosflue

newbie

Activity: 71

Merit: 0

Quote from: nonce-pool on August 30, 2014, 05:51:57 AM

Quote from: curefrankosflue on August 29, 2014, 02:38:11 PM

hey, just a question, in case you have time to answer.

how did you manage to compile the Riecoin miner on your Odroid? I have a U3 and have been trying to compile it without luck. I followed the instructions in the github directory:
https://github.com/gatra/fastrie/tree/stratum/xptMiner

Prerequisites:
Ubuntu: sudo apt-get -y install build-essential m4 openssl libssl-dev git libjson0 libjson0-dev libcurl4-openssl-dev wget

Building:
wget http://mirrors.kernel.org/gnu/gmp/gmp-5.1.3.tar.bz2 tar xjvf gmp-5.1.3.tar.bz2 cd gmp-5.1.3 ./configure --enable-cxx make -j4 && sudo make install cd git clone https://github.com/clintar/xptMiner.git cd xptMiner LD_LIBRARY_PATH=/usr/local/lib make -j4 ./xptminer -u username.riecoinworkername -p workerpassword

I'm running Ubuntu 14.04 Lts on the Odroid.

Any help would be greatly appreciated. I'm still new to linux, so the problem might very well be sitting in front of the screen

Thanks for your time.

What was the error? If I had to guess apt-get install libjansson-dev and try again.

Hey nonce-pool, thanks for answering. in case you have time to answer again I would appreciate your help.

I tried what you suggested and managed the miner compile on Ubuntu 14.04. I got an executable xptminer file, although the miner didn't mine.
It got just 0s for all chains and no shares even after some 30 minutes. Of course I triplechecked whether I was mining to the right address and the other syntax was right.

2ch/s: 0.000 3ch/s: 0.000 4ch/s: 0.0000

I switched back to Ubuntu 13.10 as some other things did not work properly and Ubuntu 13.10 has done more or less well so far.
I tried to download and compile the miner there.

All steps went well until the command:
LD_LIBRARY_PATH=/usr/local/lib make -j4

As far as I understand the line LD_LIBRARY_PATH.... compiles the xptMiner to make an executable file.

It compiled many of the source code files but in the end I got the error message:
/usr/bin/ld: cannot find -lcrypto
/usr/bin/ld: cannot find -lssl
collect2: error: ld returned 1 exit status
make: *** [xptminer] Error 1

maybe I am missing some other libraries? or are these missing source code files of the xptminer? as I said I'm a newbie to Linux

Thank you very much for your help.

Michael
ana: curefrankosflue

northranger79510

sr. member

Activity: 308

Merit: 250

Riecoin and Huntercoin to rule all!

I still think Riecoin has what it takes to go to $10 and succeed but might be away. Apologies for not being active as of recent.

dga

hero member

Activity: 737

Merit: 511

Quote from: The Dark Knight on September 01, 2014, 04:54:51 PM

I am delighted to confirm that the bug is fixed. The miner is finding shares just as quick as its Linux equivalent. Works like a charm on Windows.

Awesome. Thanks for testing.

I've put b16 online in the usual spot

http://www.cs.cmu.edu/~dga/crypto/ric/

including windows .exe files. Woo.

bsunau7

member

Activity: 114

Merit: 10

Another version of the little ARM miner project.

A 5-11% improvements depending on sieve size (the bigger the sieve bigger improvement) by re-factoring part of the code in assembly. To get the 11% I had to increase my sieve by about 30% (I am now sieving using the first 770k primes), so you'll need to increase '-m' accordingly.

Link is https://mega.co.nz/#F!6Y8ywRoZ!5ogMJEc2Teq5ckt9WClNpg.

Regards,

PS The last time I had to do assembler was for m88k, which should give you an idea of how long ago that was.

--
bsunau7

The Dark Knight

newbie

Activity: 2

Merit: 0

Quote from: dga on August 31, 2014, 01:31:12 PM

Fixed a bug in b15 reported by github user HenryHu. This doesn't seem to affect the Linux version, but I wonder if it's related to the hang on Windows. The fix has been pushed to git in https://github.com/dave-andersen/fastrie/commit/79e21cecb9bea07caac30c6a6b46735f558faee8

(It's a one-line fix that should not mess up Gatra's current effort at porting to stratum)

I've put a temporary alpha test binary in http[Suspicious link removed]

if someone with windows wants to see if that one actually works or not. It probably doesn't, but it's worth checking.

Thanks to HenryHu for the report and precise identification of the bug!

(If anyone tests it and notices that this actually fixes a real bug on Linux, LMK and I'll push out a quick b16 build)

I am delighted to confirm that the bug is fixed. The miner is finding shares just as quick as its Linux equivalent. Works like a charm on Windows.

Topic: [ANN][RIC] Riecoin: constellations POW *CPU* HARD FORK successful, world record - page 137. (Read 685207 times)

Topic: [ANN][RIC] Riecoin: constellations POW CPU HARD FORK successful, world record - page 137. (Read 685207 times)