Topic: [ATTN: POOL OPERATORS] PoolServerJ - scalable java mining pool backend - page 5. (Read 31150 times)

not yet... I'm waiting to prove it with an nmc block on testnet.

so is the single line of code updated on the poolserverj site?

ok chill out for a minute Davinci...

Firstly you knew all along I didn't not have them means to fully test it.  A full test means mining a real block and it's well known by many including you that I only have 50MH at my disposal so there was never any chance that I could a release a fully tested version within 24 hrs.  It was at best a guess and hope that the I'd gotten every minor detail of the protocol right.  What I did release was a couple of thousand lines code with one wrong line (see below). That was 1/2 written over the previous few weeks and 1/2 written in a 12 hour coding marathon...

Secondly lets not forget that after this flurry of activity and urgency it took you several days before you even got around to testing it.  Unfortunately you seem to have a habit of coming online just before I go to bed so we haven't exactly had good communications.  You should probably be aware I've been working non-stop on this for the past 5 days.  You carry on as though I've abandoned the project whenever you can't get hold of me for a couple of days.  If I'm not answering on IRC it chances are I'm alseep or I've got head buried so deep in some code that I don't want to break my concentration.  3 other people have been helping me with various test/debug processes when I haven't been able to get a hold of you so you don't worry about idle hours...

I don't know why your miners haven't been able to connect... Other people have been able to.  I'm sorry about the message last night then immediately disappearing.  That was extenuating circumstance as some clown got drunk and drove into a power pole in my street and we lost power for several hours.

The good news is after rolling up sleeves and hacking at the namecoind code yesterday to dump some useful trace info and see what's going on from that end I think I've found the problem.  And as these things often turn out the be it was a single line of code.  I still need to prove it but all the tests I've done so far indicate it will work when a real nmc solution is submitted.

As to your offer, whilst I am a big fan of passive income your comment 'until it's sold to someone else' makes that a little vague.  It could be sold tomorrow for all I know.  So I will accept the other offer of the original bounty.  I feel this is reasonable considering 99% of the work was completed in the original timeframe and that it was impossible to complete the last 1% without outside assistance.

Thanks if you are supporting the PSJ give the bitcoin to shadders if you are supporting MMP in PSJ you can donate here...
17nk7MqLLNy9Kw3NGqpa4G6qsPtWkUTuUX
http://blockexplorer.com/address/17nk7MqLLNy9Kw3NGqpa4G6qsPtWkUTuUX
I appreciate the support.

I'am going to cover 25 BTC of davincis bounty for a working solution. It seems a little bit tiny but there are plenty pools out there running poolserverj who could be interested as well. Maybe someone else is going to add some btc, i think we should support shadders and davinci here.

After talking to shadders on irc, i understand that not only poolserverj has to do the job well. I would like to see a working 4diff patch for vinced's bitcoind/namecoind in order to eliminate the duplicate work bug and multithread support.

Maybe we should talk that through on irc?

Shadders,

Our original deal was 100 BTC for the Merged-Mine-Proxy (MMP) to be incorporated into PoolServerJ (PSJ).  Then I requested you do it in 18hrs for 300 BTC or 24hrs for 200 BTC.  You worked very hard and released a version that did not work.

I paid you 100BTC before trying it assuming you did it, but your valiant efforts did not pan out as I hoped it would.  I agreed to give you access to a server on Amazon and 1.5 GHs of my miners, I have them currently set the domain name you gave me as the primary pool and they have not mined anything from your test domain.  You messaged me once asking where the miners where and I responded telling you they are still waiting for your pool to come alive and I have not heard from you since.

With that I am not sure where you stand with regards to incorporating MMP into PSJ, but here is where I stand... First off I agreed to pay you 100BTC to complete it and you did not.  You stated in the IRC that you would role it back and drop MMP code should I not continue helping you so I offered my assistance and I considered give you 1.5% of the revenue of nmcbit.com once the code was completed.

Well, this is my final offer for you, complete a releasable version with MMP incorporated by the next NMC or BTC difficulty change and I will pay you 200BTC that's on top of the 100 I gave you OR you can take 1.5% of the revenue generated from nmcbit.com until it's sold to someone else.  I have worked very hard to get this pool running and I have gotten up to 167GHs of miners partly thanks to PSJ, but just remember that no other pool has offered you anything for your efforts, I at least gave you 10BTC as a donation [before I offered and paid 100BTC for MMP] witch you told me it increased your BTC account by thousands of percents.

Finally, after the deadline has been reached I will find someone else to help me or figure it out on my own.  With that you should you not be able to complete it in time, it is my opinion that you should complete the incorporation of MMP into PSJ at your leisure for the community to use as I did pay you 100BTC to do so, however that's up to you and your conscience.

Best Regards

Davinci

This is just badass and props to both of you. I have merged mining going on my pool now. My only request now is it separate the winning shares between chains.

No need to ask permission to modify psj.  It's GPL.  Earlier versions were no-derivs creative commons license but that's long gone.  However, I don't think you'll need to. Take a look at the very end of the sample properties file.  There is actually an option to set difficulty though no one's ever used it in production as far as I know.  I would recommend after setting you do a manual getwork and inspect the target to ensure the endianess is correct.

I don't know anything about litecoin but psj has been successfully used on several alt chains.  The only real requirement is that the alt chain uses the same hashing algorithm as bitcoin, if so poolserverj won't know the difference and will happily do it's thing.

One this to watch with alt chains though is whether or not they've got the duplicate work bug.  Have search back in this thread for an explanation of it...

Quick question on poolserverj - is there by chance a config setting that will let me change the difficulty for 'accepted work'?  Haven't had a chance to take a closer look at the source yet, figured somebody else might know.

The reason I'm asking is I'm considering porting it to make a pool for litecoins, but of course would ask permission if I had to modify any of the source (I did manage to take a brief look at license.txt while perusing a bit of the poolserverj code ).

I'd prefer this over pushpool because I like Java.  

As far as I can tell anything other than share difficulty would have an easy workaround to get litecoind and poolserverj to work together (in terms of, if I run into an issue, I can make another program of my own to translate between the two, or modify litecoind).

Hi Chris,

I'm not a bastard like most pool operators and coin developers I will tell what I know.  I am using this fork...
https://github.com/ArtForz/namecoin/commit/127deb4aff13965741130dba7304073330a4adea

When it compiles it will give you an error you need to comment out the code as it's not needed for mining.

Good luck

Davinci

Thanks DavinciJ15, this solution works fine for me.

May i ask if you found a 4diff patch which can be compiled with the current bitcoind version vom vinced?

Best regards,

Chris

The upstream result should be true for both chains for now and you just check your wallet to see witch block chain got  it.  In the future shadders can add column in the database or a new table (preferable) that indicates witch block chain found a block.

So what happens to the shares table with merged mining option enabled?

Can you confirm, that upstream result from each row refers to the parentchain (btc) only? So the upstream results from the merged chain is currently not in the database right?

100 BTC sent.
Any tips on compiling the code? Or is there a binary JAR I can download?  I have 600 MH for testing so I can see some results faster.

just the address in my sig is fine...

Just uploading rev .02 atm.  Found a deadlock in the work submit thread.  There's some awfully convoluted syncing that has to go on there.

about 6 mins to go on my lightning internet connection.

also, recommend that you set 'forceAllSubmitsUpstream' until it's proven that it detects winning shares.  There will be a log file created in yr launch directory called 'real-share-submits.log' that tracks the progress of a suspected winning share through the server and gives the info we need to prove it's going to work without the force option switched on.

Awesome dude send me your address and I will send the first 100 right away.
Nodemaster has his own test net with the difficulty at crazy low value for namecoins
add this to yoru bitcoin.conf with noirc=1

addnode=78.47.40.55:18334

However I am not sure if he as taken it down.  Create a VM and check I'm busy trying testing your code.

Committed...

Well we may have to wait a while to decide if that counts as working.  Of course I can actually test if winning blocks are accepted without winning one on my 50MH rig.

sent http://blockexplorer.com/tx/9717e933e0e2cfde85cf4af930f89afeaeebc96a231673ed13c2e18e2c1a36b7
btw. can't connect to balanceserver1 and 3...

been going at this damn thing for 6 hrs straight now.  I remember why I've been putting it off for so long.

BTW I told Davinci that I wouldn't release this publicly for at least a week after he got it which I think is fair given the hefty bounty he's posted.  However he generously declined the offer and requested I make it available to everyone immediately so if you do use it please consider making a donation to Davinci because I certainly wouldn't have gotten off my lazy arse to finish if he hadn't dangled the carrot in front of me.