Well 0.6 can mine 4 different Equihash algorithms, 0.7 alpha currently only one, but do not worry, the other will come back soon

Hi Folks,

sorry for the silence - I do not write so much when busy with coding.
So, what I got for you is lolMiner 0.7 alpha 2 for mining BEAM.

Changes over alpha 1:
- Memory consumption reduced from 3.9 to 3.4G (so now more stable on 4G cards even with screen attached)
- Performance increased about 15%
- Experimentally the API now shows submitted and rejected shares per card

Download link:
https://github.com/Lolliedieb/lolMiner-preview/releases/tag/0.7_a2

MD5 Hashes of release files:
Windows: 12efcea8836eb7ec482a056c4b653fac
Linux: c2529dd0e2dc2816a5cd51506938b075

Notes:
- HiveOS, Pimp OS and others are informed
- No idea yet about Nvidia performance - any comment welcome


Regarding all here where the OpenSource miner was stable but this one not: Check out new version - the reduced memory usage should help in those cases where one card was slow (4G cards).

Just updated the miner ... somehow I fucked up the config file and there was an other small issue I mitigated

I do not know what they did when integrating this, but this -usercfg= is an outdated style. Try "--usercfg /home/ethos/lol_config.stub.json".

Hey Guys,

I was pretty busy recently to prepare the first alpha version of lolMiner 0.7 for you

The release files can be found here:
https://github.com/Lolliedieb/lolMiner-preview/releases/tag/0.7_a1

Internally the miner changes a lot, especially its getting a complete new mining backbone for more stability. Therefore the currently only supported coin is BEAM (Equihash 150/5) at a fee of 1%.

Also the miner is currently only optimized for AMD cards – I guess for Nvidia the cuda alternatives are quicker and the new backbone code produces some high CPU load with them (on AMD is extremely low on contrast). Still, a 4G Nvidia GPU will run if you want to test it.

With regard to the overall schedule: The next weeks I will introduce the other algorithms plus some new stuff that is still a surprise. What I can unveil: all lolMiner binary releases will have fee lowered for all algorithms to 1%!

Hope that serves you well, have fun mining Beam

It seems that -usercfg is also old style ... try "--usercfg lol_config.stub.json"

Right, sry for being so silent here.
As HaloGenius pointed out correctly I was called three weeks ago to help Beam team getting a miner software ready for their mainnet launch - which was done yesterday. While this turned me very busy last days it had some positive side effects:
1) There is now an OpenSource Equihash miner for Beam in the public
2) While writing it I learned a lot about my own code again
3) Of cause got a compensation by Beam team for my work. This also ensures I got enough funds to continue my work here.

When storm at Beam is calmed I will continue work with lolMiner almost immediately, hoping for a 0.7 release in February. I will change large parts of its back end (but I assume configuration to be compatible) and of cause will add Beam as well as some nice surprises for you. Stay tuned

Can you send me your full config in a private msg?
Config file (if used) + command line string you use to start? I will check it.

Suprnova is also off by has rate, they sometimes confuse numbers.
Well raking a lot fee is relative... for mnx I take 1%... do you know how much I earned in MNX last month? I can tell you, it was presisely 84.4$ (I know because I registered this as buisness and need to pay taxes on this). Now tell me how much does a software dev take that develops such specialized stuff?

Overall spend approx 300 hrs last 6 weeks in lolMiner miner dev, earned approx 1150$ in total from fees in that time (before taxes). If I would not need the bugs or without external jobs I do on top I would have stoped for long...
What I see is your uses only single minus sign, but parameters of lolMiner 0.6 start with two minus signs and without the "=" sign, e.g. --coin AION ... then it should work

Seems that the --coin parameter or its config file counterpart is missing when starting miner. Can you check that?

Just a small hint if someone is reading here:

Someone is atm mining to my fee acc directly because using the default user_config.json file.
That very generous, but I guess you prefer mining into your own wallet Hope this msg arrives the user

Ah, then you mean 2G instead of 2M ...

well the 7450 is an almost 7 year old low end card. On the memory hard algorithms like 144/5 or 192/7 it would perform horrible like maybe 1-2 sol/s if even. You could try to run it with AION or MNX, that two algorithms may fit for the card. But overall I doubt that its worth the energy the card consumes.

Yes, the 0.6 works with MNX

For AMD the code is the same as 0.5, because the code I tested in alpha 4 turned out to be slower on Vega cards then the 0.5 code. For Nvidia the code equals the one of alpha 3 which is quicker then 0.5.  Hope to find something even quicker in 0.7 but did not start that task yet.

You confuse me... 2M? What do you meam by this ^^

hohoho (= ho^3)
Sry users for the long silence, I was very busy recently. Still I found some time to prepare lolMiner 0.6 for you


Here is the download link:
https://github.com/Lolliedieb/lolMiner-releases/releases/tag/0.6

Changes over 0.6 alpha 4:

- Fixed a bug with defect shares when running on >4 GPUs
- Slightly improved AION performance on AMD
- Added support for edit ExchangeCoin (shortcut: EXCC) (Equihash 144/5) pow structure change that happens today (!)
- Rules back to the alpha 3 MNX code for AMD. Worked better for most cards
- Reduced Equihash 144/5 fee to 1.5%
- lolMiner Awesome OS integration and HIVE 0.6 integration will be updated automatically


Changes over 0.5 stable:

- New codes for all coins on all cards (better speed)
- New API and statistics module (API updating every 10 seconds)
- Better stratum stability
- Improved stability for many GPU systems
- and many more small stuff I can not remember


One note regarding the EXCC support:

the work was funded by the EXCC team. If was a bit more to do, because EXCC has a different block header and also an own stratum format. Due to the funding the mining will be at 0 fee in the 0.6 stable branch. The miner supports the new official EXCC pool software that will get public with the fork. No proxy or similar required.

I wish you guys a very marry Christmas and a good start into 2019. I will come back some time in January with an alpha for lolMiner 0.7 Stay tuned.

Just got aware of this ... I think I got an idea why this happens, will fix it asap.


Is alpha 3 working fine for you but alpha 4 does not? Just wonder, because the file for 144/5 did not change and there were no complains about alpha 3 into this direction. Something must be broken with the kernel file 

Yes, may be problem on their side, do not know, but even on this small data base 18% package loss is too much. Hope it is resolved now, happy mining



Great



So, alpha 4 making trouble here? If yes, can you check what happens with only GPUs 0-3 running and then slowly take in more GPUs? Would like to know when the trouble starts. I my self do not have so many GPUs to test.
(By the way it is possible that this is only a display problem and they are mining fine due to new stats routine ... but needs to be investigated first.

Well loosing it from time to time can happen when connection gets closed (e.g. pool thinks miner is offline because did not send shares for a while) ... but it should recover quickly!

But you are not using 0.6 alpha 4 there ... because that one has changed the waiting time for reconnect from 5 to 3 seconds

Hey folks, just released lolMiner 0.6 alpha 4

You can find the download here:
https://github.com/Lolliedieb/lolMiner-preview/releases/tag/alpha4

Change log:

MNX AMD 96/5 code changed

- improved performance on AMD by approx 6%, 580 4G stock now at approx 13700 sol/s
- reduced memory consumption, so work batch can be chosen higher

Reworked stratum interface

- More connection stability
- Better response time
- A new target is now only shown when it is really different from the previous
- Slight change to internal AUTO144_5 routines (better support)

Reworked mining code

- When more then 4 GPUs are used, more host side miner threads will be launched
- (Aiming to improve stability for >4 GPUs)
- Statistics are moved to API / statistics module
- Some small bug fixes (e.g. miner now runes fine with the remaining GPUs when some GPUs fail the compatibility check)


The most important here  is the new stratum and mining code on host side. The stratum interfaces are now much more stable, especially the alpha 3 AION problems are fixed.
I hope some miner with more then 4 GPUs can test the new mining side code - I hope to have at least one of the instability issues with many GPUs fixed.

Awesome Miner news:
Also had a recent discussion with the awesome miner team. They are looking into this release and we will pushing 0.6 integration forwards.

Hive OS news:
lolMiner 0.6 series is moved into Hive OS 6 core. Currently the release is using alpha 3 but will be updated to alpha 4 soon.
One important note when moving from HIVE 5 to 6: The miner now uses the command line to pass parameters to the miner. So extra argument field now now uses the --param value syntax instead of "PARAM" : VALUE .

Nicehash news:
As some may have noticed nicehash recently added Equihash 144/5 (what they label wron as ZHash, because ZHash is rather the name of 144/5+BTCZ personalization).
Anyways, you can mine there with alpha 4 by using
Note: In one test I observed approx 2% rejected shares. I parsed the return code and see they claimed the share was above target, but verified multiple times it was in fact below (although close to the the target). Will investigate it further and contact nicehash team when needed.

Future plan:
If the changes work smoothly I will go through some of the codes again to speed them up and we will have 0.6 final the next 10 to 14 days Stay tuned.

----------




Check out the new version - it now should continue fine. The waitforrewind is a two edged blade - it resets some internal timers in the failover support code. Its better when it works without touching this parameter.

Also on my 4G 580 it works, but the older generation cards sometimes do not allow to allocate more then 1/4th of its mem for computing - but this does not seem to be the problem here.

Short status update: new Straum interfaces now tested internally if they are stable - at least they are more responsive and easier for me to maintain. If they are fine will make alpha 4 release after this weekend.
Also did a first step into the direction of resolving the stability issues with more then 5 cards. Miner now detects if there are more cards to mine on and if yes it starts multiple mining handlers each one doing a subset of the cards. Do not know yet if this helps - because I do not have more then 4 cards - but at least a test with 2x2 cards worked well enough - hope this makes the miner more stable for your large rigs, too

Also coupled off the statistics printing of the miner process itself and shifted it into the statistics module (that before did api) ... so when miner disconnects and waits for new connection you will get an updated hash of 0 sol/s during that time but this is not critical - only miner will be stopped during the time the stratum reconnects.  Well, lots of small stuff but feels god