[ANN][BLC] Blakecoin Blake-256 for GPU/FPGA With Merged Mined Pools Stable Net

[mogrith]

Well if it was Blakes7coin could probably get it into SciFiCoin exchange that is opening in Jan. 1/2

[kr105]

Blakes7coin

?

[BlueDragon747]

still getting quite a few (high-hash) errors on the pool 

[kramble]

still getting quite a few (high-hash) errors on the pool 

Yep, its going through cycles of working OK, then high-hash, then working OK again. The "high-hash" message is in the JSON returned by the pool, not generated locally in cgminer. Interestingly I've seen my worker change between diff=2 and diff=4 on the pool dashboard, but cgminer always reports diff=2 as the target.

I've also taken a look at the calc_diff code in cgminer, and the diff1target in submit_share. I think this is just cosmetic, the miner is actually using the standard target sent by the pool (which must be for a 0x000000007fffffff hash I guess since its diff=2).

[mogrith]

Blakes7coin

?

 Blake's 7 is a old BBC SciFi Show http://blakes7.com/

1/2 because I don't really think BlueDragon wants to take the coin that way. The scficoin exchange if for Sf Themed coins. (ok OFF is more horror but I'm not going to argue with an Elder God)

[mogrith]

Well I did not disconnect overnight like I normally do. No orphan blocks mined. Seems much better now.

[kramble]

Blake's 7 is a old BBC SciFi Show http://blakes7.com/

1/2 because I don't really think BlueDragon wants to take the coin that way. The scficoin exchange if for Sf Themed coins. (ok OFF is more horror but I'm not going to argue with an Elder God)

BlueDragon did mention that he's developing a MMORPG or similar and wanted to use BlakeCoin as the in-game currency, so it may be relevant after all. And I'm old enough to have seen the entire Blake 7 on its initial transmission (wobbly sets and all).

[mogrith]

Blake's 7 is a old BBC SciFi Show http://blakes7.com/

1/2 because I don't really think BlueDragon wants to take the coin that way. The scficoin exchange if for Sf Themed coins. (ok OFF is more horror but I'm not going to argue with an Elder God)

BlueDragon did mention that he's developing a MMORPG or similar and wanted to use BlakeCoin as the in-game currency, so it may be relevant after all. And I'm old enough to have seen the entire Blake 7 on its initial transmission (wobbly sets and all).

If BlakeCoin becomes SF themed on it's own that works just fine. The exchange can be another way to get Blake let the market decide how many Blakes will buy a bar of Gold Pressed Latinum.

[kramble]

If BlakeCoin becomes SF themed on it's own that works just fine. The exchange can be another way to get Blake let the market decide how many Blakes will buy a bar of Gold Pressed Latinum.

Fascinating, that prompted a little google-fu. Apparently Latinum cannot be "replicated" (though I assume Gold can be), so it has some similarity to cryptocurrencies. Of course currency would be an anathema in a post-scarcity culture. Anyway this is getting seriously off topic, so I'll stop there.

[mogrith]

Well I got orphan blocks again.

Not sure if less often because of Diff reporting is better or if Higher Diff means I'm finding few blocks to notice.

[kramble]

Well I got orphan blocks again.

Not sure if less often because of Diff reporting is better or if Higher Diff means I'm finding few blocks to notice.

It'll be the higher diff. The changes I made to cgminer 3.1.1 won't have affected solo mining at all.
Specifically:
gen_hash() lines 5392-94 is only used in the stratum pool not for solo mining.
diff1targ line 5607 just affects the final check before submitting a share. I should revert this as its just plain wrong (in the sense that it allows sub diff 1 shares to be submitted, but these will never be generated by the FPGA except as HW errors).

I'm not sure that the (unimplemented) changes to calc_diff() to are actually required as I think this just affects the displayed diff and as that is currently correct, it seems pointless to change it (especially as I don't fully understand the code anyway).

The pool seems to have been behaving itself overnight as I didn't see any rejected blocks in the log. I'm a little concerned that to date I've found 17 blocks (425 BLC) on the pool but my earnings (both confirmed and unconfirmed) are just 225 BLC (almost exactly half). It could just be Lady Luck doing her hot/cold thing again, but it would be useful to cross-check with other users' experience.

[BlueDragon747]

The pool seems to have been behaving itself overnight as I didn't see any rejected blocks in the log. I'm a little concerned that to date I've found 17 blocks (425 BLC) on the pool but my earnings (both confirmed and unconfirmed) are just 225 BLC (almost exactly half). It could just be Lady Luck doing her hot/cold thing again, but it would be useful to cross-check with other users' experience.

I think pool mining comes down to amount of share per round rather than blocks found?

solo is better if you find the blocks quickly and don't get any orphans or wallet issues  

Edit:
still getting the (high-hash) issue on the pool periodically 

[kramble]

The pool seems to have been behaving itself overnight as I didn't see any rejected blocks in the log. I'm a little concerned that to date I've found 17 blocks (425 BLC) on the pool but my earnings (both confirmed and unconfirmed) are just 225 BLC (almost exactly half). It could just be Lady Luck doing her hot/cold thing again, but it would be useful to cross-check with other users' experience.

I think pool mining comes down to amount of share per round rather than blocks found?

solo is better if you find the blocks quickly and don't get any orphans or wallet issues  

Yep, but in the long run (evening out the variation) the earnings should be identical (less the 3% pool fee). My concern comes from the fact that the pool is accepting shares at diff=2 (even though my worker sometimes shows as diff=4 on the dashboard), and given that the earnings to date are roughly half, I'm wondering if only half the shares are being credited? It could just be that my FPGA cgminer is not submitting all the shares, though I tried modifying it to sumbit diff=1 shares and they were all rejected high-hash (which explains what that message means). Anyway if you're seeing the correct earnings/blocks ratio then it must just be variation as you've got a lot more blocks than anyone else  

EDIT yep, its high-hashing right now. Unfortunately I didn't save the overnight log (just grepped it for BLOCK!) so I can't check how much was actually rejected, actually, belay that, its in my terminal history..

 [2013-12-16 11:52:07] ZTX0 0100401300-1 | 212.0MHz | (5s):422.5M (avg):426.0Mh/s | A:1754 R:1016 HW:440 U:0.0/m
 [2013-12-16 11:52:07] ZTX1 0100401300-2 | 212.0MHz | (5s):422.7M (avg):423.6Mh/s | A:1823 R:1045 HW:58 U:0.0/m
 [2013-12-16 11:52:07] ZTX2 0100401300-3 | 208.0MHz | (5s):414.7M (avg):418.5Mh/s | A:1757 R:946 HW:418 U:0.0/m
 [2013-12-16 11:52:07] ZTX3 0100401300-4 | 204.0MHz | (5s):406.3M (avg):407.6Mh/s | A:1792 R:963 HW:99 U:0.0/m

pi@tvpi ~/blakemine/cgminer-ztex $ grep BLOCK log.txt
 [2013-12-15 20:27:15] Rejected 0001f401 Diff 33.6K/2 BLOCK! ZTX 0  (high-hash)
 [2013-12-15 20:32:48] Accepted 000f79c7 Diff 4.23K/2 BLOCK! ZTX 0
 [2013-12-15 22:20:11] Accepted 000ed150 Diff 4.42K/2 BLOCK! ZTX 1
 [2013-12-15 22:21:13] Accepted 00085b65 Diff 7.84K/2 BLOCK! ZTX 2
 [2013-12-16 06:19:00] Accepted 0014d800 Diff 3.14K/2 BLOCK! ZTX 0
 [2013-12-16 07:23:40] Accepted 0007393c Diff 9.07K/2 BLOCK! ZTX 0
 [2013-12-16 07:30:00] Accepted 000e03e4 Diff 4.67K/2 BLOCK! ZTX 1
 [2013-12-16 07:36:11] Accepted 0000a9a6 Diff 98.9K/2 BLOCK! ZTX 1
 [2013-12-16 09:37:23] Accepted 00113296 Diff 3.81K/2 BLOCK! ZTX 0


So it was actually bad for just under half the time. Strange that there was a eight hour gap in the block submissions though. Next time I'll save the log rather than overwriting it.

EDIT2. Its possible that the ztex's just went SICK->DEAD, then they restarted at around 06:19 ... I've just had that happen right now (running cgminer interactively rather than in background). I'm going to have to do a work around for this as the SICK/DEAD logic is more appropriate to the GPU cards than FPGA. Perhaps I'll just exit cgminer on detecting SICK and run it in a loop in a bash script so it restarts (and reloads the bitstreams) automatically?

[BlueDragon747]

The pool seems to have been behaving itself overnight as I didn't see any rejected blocks in the log. I'm a little concerned that to date I've found 17 blocks (425 BLC) on the pool but my earnings (both confirmed and unconfirmed) are just 225 BLC (almost exactly half). It could just be Lady Luck doing her hot/cold thing again, but it would be useful to cross-check with other users' experience.

I think pool mining comes down to amount of share per round rather than blocks found?

solo is better if you find the blocks quickly and don't get any orphans or wallet issues  

Yep, but in the long run (evening out the variation) the earnings should be identical (less the 3% pool fee). My concern comes from the fact that the pool is accepting shares at diff=2 (even though my worker sometimes shows as diff=4 on the dashboard), and given that the earnings to date are roughly half, I'm wondering if only half the shares are being credited? It could just be that my FPGA cgminer is not submitting all the shares, though I tried modifying it to sumbit diff=1 shares and they were all rejected high-hash (which explains what that message means). Anyway if you're seeing the correct earnings/blocks ratio then it must just be variation as you've got a lot more blocks than anyone else  

yeah I noticed the diff 4 on the dashboard before it gets the high-hash error then dashboard drops back to diff 2 but the pool still gets the high-hash error until a new block is found on the network?

I do get less on the pool vs solo but it should be more consistent, but due to the high-hash issue on the pool the average per hour is lower  

I do get the correct payout for my round share ratio on the pool

[Spayse]

Thanks for making A POOL i have a feeling this coin will hit the exchanges now and  in a big way!  too bad this pool just came out    ive found with a little tuning i was able to double my mhs from the default conf file. is that the blake hashing formula? (im about three times faster than btc mining) now 1 issue  high hash reject wtf what do  i need to do? i lose about35%





0
 

[BlueDragon747]

its great to see a working pool but the high-hash error needs to be fixed by kr105 please its so close to working perfect  

the modified blake-256 algorithm is just under 3x hash rate in my testing compared with Bitcoin on the GPU and just over 2x on the FPGA, Nvidia cards are slow needs a Cuda Miner  

[kr105]

its great to see a working pool but the high-hash error needs to be fixed by kr105 please its so close to working perfect  

the modified blake-256 algorithm is just under 3x hash rate in my testing compared with Bitcoin on the GPU and just over 2x on the FPGA, Nvidia cards are slow needs a Cuda Miner  

I will back to work on this soon, I'm out on a business trip for a few days and will be back very soon!

[svost]

downloaded and compiled from https://github.com/kR105/cgminer

run cgminer -k blake256



Why WU = 0 

[BlueDragon747]

downloaded and compiled from https://github.com/kR105/cgminer

run cgminer -k blake256

Why WU = 0 

for the pool this is my cgminer.conf:

{
"pools" : [
   {
      "url" : "stratum+tcp://stratum.blakecoinpool.org:3333",
      "user" : "UserName.Worker",
      "pass" : "WorkerPassword"
   }
],

"intensity" : "9",
"auto-gpu" : true,
"expiry" : "120",
"failover-only" : true,
"gpu-threads" : "2",
"log" : "5",
"no-restart" : true,
"queue" : "5",
"scan-time" : "10",
"worksize" : "128",
"temp-hysteresis" : "4",
"blake256" : true,
"vectors" : "1",
"no-submit-stale": true,
"kernel-path" : "/"
}

I only run cgminer no option switches as it is all in the cgminer.conf

[BombaUcigasa]

I can report the pool works correctly at least 50% of the time and that at this time each blakecoin is worth about 0.01 USD or so...