hsrminer - Nvidia mining software for various algos by palgin&alexkap

[palgin]

Ha-ha-ha, haven't been here for quite a long time.

Reverse engineering is cool, but I see no interest for myself in this anymore, especially when it finally came that alexkap died in road accident month ago, which I found out few days ago, so I've lost a colleague and a friend of mine (we were not so close, but he was a person I could rely on)

Info: I've personally spent 1.5 month in hospital (right after New Year, first it was agressive flu that turned to pneumonia and in hospital I got viral meningitis, thank you, hospital). Thanks for everyone who used MY hsrminer during that time, you've helped me to pay all medical bills and survive this nightmare.
Also, hsrminer helped me to attract not only haters, hackers and other sort of garbage, but also good employer, so now I have much less time to spend on developement.

I have Cryptonite nearly ready, also Lyra2v2 with impressive speed improvement.

I see two ways of future hsrminer developement and haven't decided which one to choose:

1) I rework all mechanics, add more layers of code protection and publish new hsrminer with faster 5-7% speed, but not full speed (currently it's significantly more than 5%+   ). Mr hacker hacks this new hsrminer, I fully rework mechanics again and add 5% more speed, and so on, until mr hacker finds something more interesting for himself rather than jerking off all day long. Also, devfee will be lowered because of situation with alexkap.

2) Community betrayed developer, so I won't move even a finger to do something for this "community" anymore and start selling kernels to factory-grade players (currently I have 2 propositions, but price they offer is too low for one-time deal)

Anyway, wish everyone good luck, and have a nice day!

P.S.: links fixed

[1714663641]

1714663641

[1714663641]

1714663641

[Larvitar]

Ha-ha-ha, haven't been here for quite a long time.

Reverse engineering is cool, but I see no interest for myself in this anymore, especially when it finally came that alexkap died in road accident month ago, which I found out few days ago, so I've lost a colleague and a friend of mine (we were not so close, but he was a person I could rely on)

Info: I've personally spent 1.5 month in hospital (right after New Year, first it was agressive flu that turned to pneumonia and in hospital I got viral meningitis, thank you, hospital). Thanks for everyone who used MY hsrminer during that time, you've helped me to pay all medical bills and survive this nightmare.
Also, hsrminer helped me to attract not only haters, hackers and other sort of garbage, but also good employer, so now I have much less time to spend on developement.

I have Cryptonite nearly ready, also Lyra2v2 with impressive speed improvement.

I see two ways of future hsrminer developement and haven't decided which one to choose:

1) I rework all mechanics, add more layers of code protection and publish new hsrminer with faster 5-7% speed, but not full speed (currently it's significantly more than 5%+   ). Mr hacker hacks this new hsrminer, I fully rework mechanics again and add 5% more speed, and so on, until mr hacker finds something more interesting for himself rather than jerking off all day long. Also, devfee will be lowered because of situation with alexkap.

2) Community betrayed developer, so I won't move even a finger to do something for this "community" anymore and start selling kernels to factory-grade players (currently I have 2 propositions, but price they offer is too low for one-time deal)

Anyway, wish everyone good luck, and have a nice day!

P.S.: links fixed

My condolences about alexkap. I lost a friend some time ago, and it's hard. Too hard.

Could I suggest a third way to HSRMiner:
3- Kick hacker, cracker, baker, lacker, (n)er asses with a overkill HSRMiner with a better container, avoiding the reverse-engineering. You didn't betrayed by the ENTIRE community, but just a part. And yes, after I know your situation, I understand the lack of updates and I say: ok, I'll stick at your miner, and support it to become the greater Neoscrypt/Lyra2v2/Cryptonite/Potato/zzzz miner.

I had success using HSRMiner in AwesomeMiner (by switching ccminer executable by HSRMiner one), and it's overkill. It can be a good way to improve the miners base and, of course, fee. We have some private miners, it's bad to see the best miners in few hands.

Best regards, palgin. Don't give up.

[palgin]

My condolences about alexkap. I lost a friend some time ago, and it's hard. Too hard.

Could I suggest a third way to HSRMiner:
3- Kick hacker, cracker, baker, lacker, (n)er asses with a overkill HSRMiner with a better container, avoiding the reverse-engineering. You didn't betrayed by the ENTIRE community, but just a part. And yes, after I know your situation, I understand the lack of updates and I say: ok, I'll stick at your miner, and support it to become the greater Neoscrypt/Lyra2v2/Cryptonite/Potato/zzzz miner.

I had success using HSRMiner in AwesomeMiner (by switching ccminer executable by HSRMiner one), and it's overkill. It can be a good way to improve the miners base and, of course, fee. We have some private miners, it's bad to see the best miners in few hands.

Best regards, palgin. Don't give up.

Thank you, I think I will continue with hsrminer developement, devfee in future versions will be 0.5% and container will be fully reworked, design will remain nearly the same as a tribute to Alex.

Everything can be reverse-engineerd, only time matters, but I have some "features" for these attackers to make their assholes burn, I didn't include this in original code for two reasons:

1) I write only kernels, container was not my business.
2) It's some sort of "cyber weapon", so if something'd go wrong with code it would be a disaster, needs very precise testing on my side.

Now I'll start from the beginning, think I'd be able to give (n)er's a fight

[Just_a_miner]

2) Community betrayed developer, so I won't move even a finger to do something for this "community" anymore and start selling kernels to factory-grade players (currently I have 2 propositions, but price they offer is too low for one-time deal)

It's sad to see that you were sick and lost friend, but don't accuse community of betraying you - community needed help and support, and part of the community choosed what is best for it.

Thank you, I think I will continue with hsrminer developement, devfee in future versions will be 0.5% and container will be fully reworked, design will remain nearly the same as a tribute to Alex.

Do you realise that 0.5% defvee is only 18 seconds per hour? Last time you claimed that your miner had 1% devfee and it's just a lie, because devfee actually is 70 seconds per hour and it's 1.94%

Everything can be reverse-engineerd, only time matters, but I have some "features" for these attackers to make their assholes burn, I didn't include this in original code for two reasons:

1) I write only kernels, container was not my business.
2) It's some sort of "cyber weapon", so if something'd go wrong with code it would be a disaster, needs very precise testing on my side.

Let me give you good advice - never include destructive code in your software - only ordinary users will suffer from it.

[palgin]

Dear sir, my first and last message directed to you personally.

Don't waste your time reading or answering this, you'll be in my ignore list for quite a long time when your opus'll finally be born.

First of all, as I always mention, You came to my thread, not Me to yours, as you can see I've posted nothing at that garbage hole and never tried to stop your little reverse-engineering project.
It's only you splashing your saliva here, GTFO and make your life not so miserable.

Now containment:

It's sad to see that you were sick and lost friend, but don't accuse community of betraying you - community needed help and support, and part of the community choosed what is best for it.

Do you realise that 0.5% defvee is only 18 seconds per hour? Last time you claimed that your miner had 1% devfee and it's just a lie, because devfee actually is 70 seconds per hour and it's 1.94%

Let me give you good advice - never include destructive code in your software - only ordinary users will suffer from it.

1) I'm not accusing anyone, I've just posted two ways of thinking on the same problem, nothing more. And container is a shit when it comes to mining, doesn't matter how good it is, it will cost nothing without proper kernel. Which you can't say about kernel, give it input, you'll receive output, volia, solution, check it against target, volia, proper solution, not so much code, isn't it?

2) The lie is that you do it for community, you do it only because I've hurted your oversized Ego, and don't think you've hidden your personality under this new account

Are you right about devfee? Partially, man, partially, because dev pool connection is not momental, that value highly differs for every user and that doesn't mean that I gain anything from client connecting to server, only fools think so, for some users full dev time was 38-45 secs, for some it could take more as in your case. Even calculating this, kernel speed gave much more profit for end users, much more, none've pushed Neooscrypt so high before me on Pascal architecture and you're just a parasite in that case, you can write hundreds of reveals and investigations, that will change nothing.
Maybe in your perfect world users follow devs, idea, community, etc., in real world the only thing users follow is money.
In your case users switched for functionality, nothing more, so your crown is made of chocolate, man, be careful, chocolate's melting sometimes.

And let's make it bold!!!! 14528731478% DEVFEE, beware, liar, thief, criminal, freedom, ZOG, HOLD THE DOOR, hold the door, holdthdor, Hotdor, Hodor!!!

3) Destructive code is the one and only thing that can stop your kind from your business, so look carefully when you'll start new freedom-fight campaign against me, or you'll make a lot of trouble to your fanbase and for yourself personally (let's make it a little surprise and see).
Because You'll be a creator of "destructive" binary, not Me, my binary will be as clean as a perfect diamond prism.

And there will be no conclusion to all this shit, it's just some kind of open PM which everyone can read and understand from his/her point of view.

Sayonara, my "friend".

[likwidkhaos]

palgin-

i don't often post but i was digging around for a stable maxwell build of the neoscrypt kernel and saw your recent setbacks.  as these forums are often filled with toxicity, i'll keep it simple:

thank you for your development work on hsrminer_neoscrypt, it's made me a bit of cash.  i'm okay with your dev fee and i'll continue to use your software as long as it's profitable.  i get mine, you get yours.  fair's fair.

i'm hoping you're feeling better now and over your sickness.

most importantly, i'm truly sorry about the loss of your friend and colleague, may he rest in peace.  it is very unfortunate.  i'm in hopes that you only recall the good that was channeled between you and he, and realize the impact your collaborations have made on our world.

salut,

-khao

[Just_a_miner]

It's sad to see that you were sick and lost friend, but don't accuse community of betraying you - community needed help and support, and part of the community choosed what is best for it.

Do you realise that 0.5% defvee is only 18 seconds per hour? Last time you claimed that your miner had 1% devfee and it's just a lie, because devfee actually is 70 seconds per hour and it's 1.94%

Let me give you good advice - never include destructive code in your software - only ordinary users will suffer from it.

1) I'm not accusing anyone, I've just posted two ways of thinking on the same problem, nothing more. And container is a shit when it comes to mining, doesn't matter how good it is, it will cost nothing without proper kernel. Which you can't say about kernel, give it input, you'll receive output, volia, solution, check it against target, volia, proper solution, not so much code, isn't it?

So "container is shit"? Interesting, probably that's why you decided not to spend time on coding own "container" and just took ccminer  

2) The lie is that you do it for community, you do it only because I've hurted your oversized Ego, and don't think you've hidden your personality under this new account

I understand that you feel yourself little dizzy after hospital, but you have mixed something up. I didn't even know who are you before I've started to use hsrminer in december. And with my fork I've actually helped people and they are grateful to me.

Are you right about devfee? Partially, man, partially, because dev pool connection is not momental, that value highly differs for every user and that doesn't mean that I gain anything from client connecting to server, only fools think so, for some users full dev time was 38-45 secs, for some it could take more as in your case. Even calculating this, kernel speed gave much more profit for end users, much more, none've pushed Neooscrypt so high before me on Pascal architecture and you're just a parasite in that case, you can write hundreds of reveals and investigations, that will change nothing.
Maybe in your perfect world users follow devs, idea, community, etc., in real world the only thing users follow is money.
In your case users switched for functionality, nothing more, so your crown is made of chocolate, man, be careful, chocolate's melting sometimes.

And let's make it bold!!!! 14528731478% DEVFEE, beware, liar, thief, criminal, freedom, ZOG, HOLD THE DOOR, hold the door, holdthdor, Hotdor, Hodor!!!

Too much words, too much emotions, too little facts. Okay guys, here we go. There are two work timelimits inside hsrminer - one is for user pool, and one is for devfee pool. Mr. Palgin set user pool work timelimit to exactly 3530 seconds, it's 58 minutes 50 seconds, and he set devfee pool work timelimit to exactly 70 seconds. How it all works - after miner works on user pool for 3530 seconds, it disconnects and switches pool to devfee one and mine there for 70 seconds. 3530+70=3600 - we get full hour. So every hour miner doesn't work for user pool at least 70 seconds  - it's minus 1.94% income for user.

Mr. Palgin is saying us that he doesn't get enough devfee due to non-momental pool switching, etc - but user doesn't need to care about that and suffer from that- user already lost 1.94% of income. And it doesn't matter for user if miner is actually mining whole 70 seconds during devfee period, or just doing nothing, trying to connect to devfee pool - user ALREADY lost 1.94% of his income.

Furthemore, after switching from devfee pool back to user pool - user experiences same problems that Mr. Palgin was complained about - miner can't instantly connect to user pool and start work, instead it waits work from user pool and due to badly optimized switch-pool code wait time can vary anywhere from 5 to 30 seconds. So 70 seconds of devfee period plus up to 30 seconds to get back to user pool and start working - 75-100 seconds, and it's 2.08% - 2.75% of lost income instead of claimed 1% for any user that uses original hsrminer.


This is real world and not cinema. As Mr. Palgin said: "in real world the only thing users follow is money". I let myself to add that in real world it's not easy to separate good guys from bad ones, as almost everyone is somewhere in between. So Mr. Palgin pretends to be a good guy, and calls me "attacker", "hacker", i.e. bad guy. Well I'm flattered.

But let's see all the lies that Mr. "Good Guy" Palgin told us:

1)

Product contains devfee 1% (0.5% for me, 0.5% for alexkap)

Confirmed Palgin's lie, any original hsrminer user lose from 2.08% to 2.75% of income due to devfee pool switching  - see above. But real devfee % would scare users away, so Mr. Palgin decided to claim that devfee is 1% which is obvious lie. Mr. Palgin is a smart guy, he creates fast cuda kernels, he knows math well, he couldn't make silly mistake calculating devfee %. Claiming fake low devfee % was done intentionally to attract more users.

2)

F.A.Q (will be updated basing on frequent questions in this topic):

Q: hsrminer looks quite like ccminer...
A: True, in early builds it really was ~30% ccminer code (especially monitoring), but now only data output is quite the same. Also we'll migrate to new frontend after algo coverage.

Confirmed Palgin's lie, I've thoroughly reverse-engineered hsrminer and know every byte of it and it contains 99% of ccminer code. Mr. Palgin just erased code parts that related to other algos (as ccminer is multialgo miner), changed user messages and add logo. Of course there are kernels, but we are talking about "shitty container" here, right? As for kernels I'm sure they were not written from scratch too - Mr. Palgin took open-source ones like tpruvot or klaust and improved them.


3)

9xx family and lower not supported in this release.

Confirmed Palgin's lie. Even without source code I was able to add support for GTX 970, 980, 980 Ti to my fork and it's workability confirmed by users - 1, 2

But 9xx GPUs doesn't give much hashrate and won't give Palgin much devfee, so he decided - why bother about it?



4) Yes, it's not over yet. Do you guys wonder why original hsrminer doesn't have API and benchmark, doesn't have "-r" option, doesn't have many useful ccminer options despite the fact that it's actually ccminer with different kernels and logo?

It's because Mr. "Good Guy" Palgin intentionally disabled API, deleted parts of the code that handle those useful options and especially "-r" option. It was done because API and those options could interfere with his devfee. So I as the "Bad Guy" had to add all those features back in my fork, working without access to source code, so people could actually use this miner instead of just paying devfee.

Mr. Palgin claims that all those features are just unnecessary things, and main thing is his fast cuda kernel. Well Mr. Palgin, miner without ability to exit if connection to pool was interrupted ( -r option) is just a joke because even fastest cuda-kernel won't pay for those losses user would get in case of pool's inaccessibility - his rigs would idle without work for hours - it's my personal experience with original hsrminer.


So Mr. Palgin, users follow money? All we can see from 1), 2), 3), 4) is that you are the one who care about money (devfee) most.
If users followed your logic - they all would instantly switch to my fork because of reasons described in 1), 3) and 4). Yes, 2) doesn't count here, it's just little addition to your "Good Guy" portrait  

Get over yourself - fix topic description and miner logo - your claimed 1% devfee is blatant lie, bruh.

[EVERYUSERNAMEISTAKEN]

Looks like the kernels have become popcorn.

Delicious.

[SCSI2]

Sounds like you guys should collaborate on the miner and split the devfee. Justaminer has spent quite some time figuring out what the end-users want, including API and reliability. I doubt that he had any ill intent by patching it in the beginning based on posts in respective thread. Slapping a devfee on a patched release was questionable, but that was done before Palgin's return.

Palgin's kernel optimization work is amazing. No questions about it.

We all would benefit if you two combine your knowledge and experience and make a great miner to try and beat that excavator which currently has ~10% advantage on 1080/1080Ti.

I don't approve patching someone else's proprietary binary, but when the only dev went AWOL, there was some excuse for it. If I were the original dev in such situation, I would expect an apology from someone who hacked my binary without my knowledge and would be open to suggestions and collaboration. But that's just me.

In any case, try to be respectful and open-minded. The last thing we need right now is a dev war in this space.

With respect,
SCSI

[Elder III]

We had all of our business rigs on hsrminer for a couple of months (until the algorithm wasn't as profitable as others) and we didn't have any problems with it. I still have a personal desktop using it for 3 months now and no problems either.  I don't have any issue with the devfee either - too many freeloaders these days that don't have the knowledge to create their own miner but expect others to give it to them for free.

[sir_blacks]

Sounds like you guys should collaborate on the miner and split the devfee. Justaminer has spent quite some time figuring out what the end-users want, including API and reliability. I doubt that he had any ill intent by patching it in the beginning based on posts in respective thread. Slapping a devfee on a patched release was questionable, but that was done before Palgin's return.

Palgin's kernel optimization work is amazing. No questions about it.

We all would benefit if you two combine your knowledge and experience and make a great miner to try and beat that excavator which currently has ~10% advantage on 1080/1080Ti.

I don't approve patching someone else's proprietary binary, but when the only dev went AWOL, there was some excuse for it. If I were the original dev in such situation, I would expect an apology from someone who hacked my binary without my knowledge and would be open to suggestions and collaboration. But that's just me.

In any case, try to be respectful and open-minded. The last thing we need right now is a dev war in this space.

With respect,
SCSI

Agreed 1000% & i will say for sure all community it have more to win if in stead a war between 2 brilliant person it start combine your mind & force !!! it create more for bouth of you & for all community

cheers

[MagicSmoker]

...
First of all, as I always mention, You came to my thread, not Me to yours, as you can see I've posted nothing at that garbage hole and never tried to stop your little reverse-engineering project.
...

All true, but you also failed to respond to *hundreds* of posts in your own thread for 2+ months so you really shouldn't be surprised that pretty much everyone thought you had abandoned your own miner. Also, popping in to post on another thread back in late January only made the absence from this thread all the more glaring.

While I personally think what Just_a_miner did was a bit shady - especially adding on his own devfee after making a big deal out of removing yours - the fact remains you were AWOL for 2+ months and Just_a_miner appeared to be trying to fix some of the shortcomings of your miner. I'm not saying you two ought to kiss and make up as @SCSI2 is suggesting, rather that you might want to back off on the self-righteous indignation because you let the community down first, not vice versa.

[iTiamo]

Hey palgin, heard getblocktemplate was supported in your miner. How to set coinbase address when solomining?

[mez]

Dear sir, my first and last message directed to you personally.

Don't waste your time reading or answering this, you'll be in my ignore list for quite a long time when your opus'll finally be born.

First of all, as I always mention, You came to my thread, not Me to yours, as you can see I've posted nothing at that garbage hole and never tried to stop your little reverse-engineering project.
It's only you splashing your saliva here, GTFO and make your life not so miserable.

Now containment:

It's sad to see that you were sick and lost friend, but don't accuse community of betraying you - community needed help and support, and part of the community choosed what is best for it.

Do you realise that 0.5% defvee is only 18 seconds per hour? Last time you claimed that your miner had 1% devfee and it's just a lie, because devfee actually is 70 seconds per hour and it's 1.94%

Let me give you good advice - never include destructive code in your software - only ordinary users will suffer from it.

1) I'm not accusing anyone, I've just posted two ways of thinking on the same problem, nothing more. And container is a shit when it comes to mining, doesn't matter how good it is, it will cost nothing without proper kernel. Which you can't say about kernel, give it input, you'll receive output, volia, solution, check it against target, volia, proper solution, not so much code, isn't it?

2) The lie is that you do it for community, you do it only because I've hurted your oversized Ego, and don't think you've hidden your personality under this new account

Are you right about devfee? Partially, man, partially, because dev pool connection is not momental, that value highly differs for every user and that doesn't mean that I gain anything from client connecting to server, only fools think so, for some users full dev time was 38-45 secs, for some it could take more as in your case. Even calculating this, kernel speed gave much more profit for end users, much more, none've pushed Neooscrypt so high before me on Pascal architecture and you're just a parasite in that case, you can write hundreds of reveals and investigations, that will change nothing.
Maybe in your perfect world users follow devs, idea, community, etc., in real world the only thing users follow is money.
In your case users switched for functionality, nothing more, so your crown is made of chocolate, man, be careful, chocolate's melting sometimes.

And let's make it bold!!!! 14528731478% DEVFEE, beware, liar, thief, criminal, freedom, ZOG, HOLD THE DOOR, hold the door, holdthdor, Hotdor, Hodor!!!

3) Destructive code is the one and only thing that can stop your kind from your business, so look carefully when you'll start new freedom-fight campaign against me, or you'll make a lot of trouble to your fanbase and for yourself personally (let's make it a little surprise and see).
Because You'll be a creator of "destructive" binary, not Me, my binary will be as clean as a perfect diamond prism.

And there will be no conclusion to all this shit, it's just some kind of open PM which everyone can read and understand from his/her point of view.

Sayonara, my "friend".

Keep up the good work, ignore the prick =) thanks for the great miner

[trucobit]

I use your miner and I am grateful, whether I use it alone or in Nemos. I have not used the other miner copied from this.

It seems to me in very bad taste that that person has to come here again and again to discredit anyone, when he makes use of his work. For that reason, do not use the alternative fork miner of this one. He is a very shameless person.

I do not think he has lost many people, and little by little he will have more users again.

Luck and encourage, I expect new versions of your hsrminer that I like so much.

An idea for the same. Many times loses too much time changing difficulty again and again, especially in Suprnova, to avoid that or to define a fixed difficulty for us would be a breakthrough.

regards

[jugger1028]

LoL.. Why is it still closed source? I'd say there are no innocent players here, it's closed source for a reason. And a dev dying and him getting sick while at the same time while "friends" sent messages of having marital problems? Yea.. everything is very coincidental and shady.

[dyablor]

Hello,

My miner works for 1-2 even 10 or more hours, but constantly crashes with the below error:

Cuda error in func 'neoscrypt_cpu_hash_1060' at line 1408 : the launch timed out and was terminated.
Cuda error in func 'neoscrypt_cpu_hash_1060' at line 1408 : the launch timed out and was terminated.
 INFO : [10:45:30] : Connection interrupted
 INFO : [10:45:30] : Share difficulty reset -- 64 (0.00098)
 INFO : [10:46:20] : New block #65039 received, new difficulty is 1230.424
 INFO : [10:46:47] : New block #65040 received, new difficulty is 1229.887
 INFO : [10:46:53] : New block #65041 received, new difficulty is 1234.741
 INFO : [10:47:26] : New block #65042 received, new difficulty is 1317.929
 INFO : [11:04:09] : Connection interrupted
 INFO : [11:04:10] : Share difficulty reset -- 32 (0.00049)
 INFO : [11:08:22] : New block #65043 received, new difficulty is 981.775
 INFO : [11:08:34] : New block #65044 received, new difficulty is 977.713
 INFO : [11:08:41] : New block #65045 received, new difficulty is 1058.435
 INFO : [11:10:32] : New block #65046 received, new difficulty is 1051.208
 INFO : [11:12:31] : New block #65047 received, new difficulty is 1038.114
 INFO : [11:13:36] : New block #65048 received, new difficulty is 1148.740
 INFO : [11:13:40] : New block #65049 received, new difficulty is 1154.018
 INFO : [11:14:33] : New block #65050 received, new difficulty is 1209.880
 INFO : [11:14:47] : New block #65051 received, new difficulty is 1269.202
 INFO : [11:15:06] : New block #65052 received, new difficulty is 1270.820
 INFO : [11:15:10] : New block #65053 received, new difficulty is 1291.227
 INFO : [11:15:43] : New block #65054 received, new difficulty is 1298.237
 INFO : [11:15:48] : New block #65055 received, new difficulty is 1312.554
 INFO : [11:18:04] : New block #65056 received, new difficulty is 1290.903
 INFO : [11:18:06] : New block #65057 received, new difficulty is 1312.083
 INFO : [11:21:59] : New block #65058 received, new difficulty is 1272.708
 INFO : [11:22:28] : New block #65059 received, new difficulty is 1641.071
 INFO : [11:22:49] : Connection interrupted

Any clues on why this is happening?

PS: I have 2 rigs with the same settings, mining the same thing, on the same pool, and only one is crashing. The other one works perfectly.

[Elder III]

Hello,

My miner works for 1-2 even 10 or more hours, but constantly crashes with the below error:

Cuda error in func 'neoscrypt_cpu_hash_1060' at line 1408 : the launch timed out and was terminated.
Cuda error in func 'neoscrypt_cpu_hash_1060' at line 1408 : the launch timed out and was terminated.
 INFO : [10:45:30] : Connection interrupted
 INFO : [10:45:30] : Share difficulty reset -- 64 (0.00098)
 INFO : [10:46:20] : New block #65039 received, new difficulty is 1230.424
 INFO : [10:46:47] : New block #65040 received, new difficulty is 1229.887
 INFO : [10:46:53] : New block #65041 received, new difficulty is 1234.741
 INFO : [10:47:26] : New block #65042 received, new difficulty is 1317.929
 INFO : [11:04:09] : Connection interrupted
 INFO : [11:04:10] : Share difficulty reset -- 32 (0.00049)
 INFO : [11:08:22] : New block #65043 received, new difficulty is 981.775
 INFO : [11:08:34] : New block #65044 received, new difficulty is 977.713
 INFO : [11:08:41] : New block #65045 received, new difficulty is 1058.435
 INFO : [11:10:32] : New block #65046 received, new difficulty is 1051.208
 INFO : [11:12:31] : New block #65047 received, new difficulty is 1038.114
 INFO : [11:13:36] : New block #65048 received, new difficulty is 1148.740
 INFO : [11:13:40] : New block #65049 received, new difficulty is 1154.018
 INFO : [11:14:33] : New block #65050 received, new difficulty is 1209.880
 INFO : [11:14:47] : New block #65051 received, new difficulty is 1269.202
 INFO : [11:15:06] : New block #65052 received, new difficulty is 1270.820
 INFO : [11:15:10] : New block #65053 received, new difficulty is 1291.227
 INFO : [11:15:43] : New block #65054 received, new difficulty is 1298.237
 INFO : [11:15:48] : New block #65055 received, new difficulty is 1312.554
 INFO : [11:18:04] : New block #65056 received, new difficulty is 1290.903
 INFO : [11:18:06] : New block #65057 received, new difficulty is 1312.083
 INFO : [11:21:59] : New block #65058 received, new difficulty is 1272.708
 INFO : [11:22:28] : New block #65059 received, new difficulty is 1641.071
 INFO : [11:22:49] : Connection interrupted

Any clues on why this is happening?

PS: I have 2 rigs with the same settings, mining the same thing, on the same pool, and only one is crashing. The other one works perfectly.

Are you using the test version of the miner? I have found that it works better for systems with GTX 1060 GPUs in them.  If you're not running at stock settings, try that first to see if it eliminates the crashing.

[gradinkov]

Well, it's surely been said a million times already, but:

PLEASE add an API!

[angelbbs]

Well, it's surely been said a million times already, but:

PLEASE add an API!

API works on a fixed port 4001