Quote from: El_Nickio on March 09, 2015, 09:25:46 PM

Quote from: thecarnie on March 09, 2015, 09:16:42 PM

Quote from: El_Nickio on March 09, 2015, 09:03:10 PM

Quote from: thecarnie on March 09, 2015, 07:59:19 AM

Quote from: El_Nickio on March 06, 2015, 09:12:15 PM

Testing your pool - works fine on my laptop with cudaminer but when i try mining with an AMD graphics card with cgminer I just get hardware errors 'invalid nonce count - HW error' - any idea why??

If you get a chance, give that another try -- I followed a couple bug reports thru GIT and I got my CPUminer to act right - so that should solve the GPU miner from having hardware errors... (was getting same issues with CPUminer you were having with ATI/AMD (OpenCL) miner) - maybe this will work. Just set any XNC address for payout - it's running on my 8.7.1 xencoind - so it should payout - so long as it wins some blocks - or I'll have to sort something out for ya for your time - I don't have any big mining hardware and so I don't generate coins anymore --

If this fixes the issue with OpenCL miners, then I'll update the GIT with the p2pool code -- then some more p2pool nodes can go online. I won't be compiling a windows binary though - not anytime soon anyways...

Seems to be working ok now - finding lots of blocks but loads of HW errors on the gpu miner - over 400,000 in about 2 hours!!! Think that its because the difficulty is so low maybe? I'm only losing about 10 khashes because of the errors so its no big deal. My laptop running cudaminer and cpuminer is working fine although i'm going to shut that down now cos its getting hot! I'll leave the gpu miner running overnight to see what happens Grin

Payouts working fine by the way

That nonce count error i was getting was because I had the share difficulty set real low in my .bat file - once i deleted that bit it was ok.

So you're still seeing hardware errors on the GPU mining with AMD/ATI? I saw a steady slew of blocks coming through!

I've found what I think to be the ultimate issue with the version of the client I was working on (with high CPU and other issues) - but I have to make some changes and build on a testnet to know for sure -- but in the end - if it is the issue - there is two options - one is to roll back to the old client and only bring in a couple updates (which won't truly fix the problem it will just make things more stable) - the other is through a hard fork - if it's even possible - that's what I'm looking at now) - but because this coin shares network ID with several other coins - the crosstalk is causing problems for the new client I am working on...

I'm still running the old client and always have been - it always seems to have only two connections though:

{
"addr" : "46.32.253.168:4334",
"
{
"addr" : "54.191.173.227:4334",
"
]
Either of these you?

I don't think a hard fork is a good idea - coins-e are on this one although the price is rock-bottom and after all this time Xencoin still seems to work!

54.191.173.227 = xencoin.liquidbbs.org (seed-node) - not sure who the 46.x is -- the version number shows it's a newer client tho (not mine)
I've seen several of the old client out there -- because of the updates to the core in my client, the protocol version is higher, and I think my node doesn't connect to older clients unless it can't find newer ones (it seems) --

As for a fork, it wouldn't be a big issue so long as the clients were buildable and available -- and we set the fork block far enough out -- the problem I'm seeing though is that a fork wouldn't fix it. It would require a different change - I'll try to document everything in a PM at some point. I may take and see about making a different change and go from there --