Pooled mining for official bitcoin client?

[bitcoinex]

Who else think that the pooled mining must be embedded into the official bitcoin client (as option in GUI menu) to allow for any user could start again to generate 0.01 bitcoin per day?

Bitcoins generator in fact is an excellent advertisement for bitcoin!

[bober182]

This wont work as pooled mining needs a server. central server. we need a p2p miner done.

[kiba]

This wont work as pooled mining needs a server. central server. we need a p2p miner done.

And why it won't work?

[bober182]

As part of the official Fully p2p anonymous client.
Who will run the server what if the server goes down.
Until a p2p pooled miner is out this is asking for trouble.

[kiba]

As part of the official Fully p2p anonymous client.
Who will run the server what if the server goes down.
Until a p2p pooled miner is out this is asking for trouble.

Just ask another person to run one.

[bober182]

So the IP/domain is coded into the client and government takes out the domain and IP. What do until next update?
before you say opensource. How many people edit there actual opensource apps like firefox and gimp.

[bitcoinex]

So the IP/domain is coded into the client and government takes out the domain and IP. What do until next update?
before you say opensource. How many people edit there actual opensource apps like firefox and gimp.

who? user will set IP manually in the GUI

I just do not understand how much secure pooling protocol

[MrFlibble]

Who else think that the pooled mining must be embedded into the official bitcoin client (as option in GUI menu) to allow for any user could start again to generate 0.01 bitcoin per day?

I think^* it's the Way Forward, but perhaps should mature before being shipped with the GUI.  Can you make this thread into a poll?  (e.g. Yes Now / Yes Later / No)


^*  I could blow some money on GPUs, but I don't play *any* big 3D games so I can only justify that if I start writing my own OpenCL code, or purely to support bitcoin. 

I lack enough CPU (spare, on machines in my home DMZ) to expect to mint a block ever, unless difficulty flattens off soon.  I could buy some coins, but have no use for them yet...  I'm left with 0.05 from charity, so I am a (not disgruntled) outsider.

Bitcoins generator in fact is an excellent advertisement for bitcoin!

I didn't understand this bit, sorry.

[BitLex]

i agree that it should be implemented.....when it's ready.

it's way to early now, we only got a few users running the first and single (at least public) pool and it hasnt even generated 1 block (to see if it works at all).

so give it some time, join the pool, keep miners connected and up to date to help testing.
maybe we get some GPU-versions to work with it soon, that'll help "outsiders" a lot.

[RHorning]

Here is an interesting thing that could be added to the Bitcoin protocol:

Right now there is an 8-byte (64 bit) identifier which can be used to identify which "services" that a particular bitcoin node is participating with.  (see http://www.bitcoin.org/wiki/doku.php?id=bitcoins_draft_spec_0_0_1#version for details) The only field currently used in the client is the "NODE_NETWORK" flag, with a value of 1.

Presumably, if somebody was running a node that would act as a "server" for concentrating mining activity and forming a core for pooled mining activity, that could be "broadcast" as a separate "service" that the node provides.  If none of the nodes you are currently connected to has that service, you could keep searching for other nodes that might be offering the service you are looking for.

The trick here is to define the protocol that would be used for such a pooled mining client/server and perhaps more significant would be setting up some sort of trust metric that could also be applied as well.  The purpose of the trust metric would be to know who to trust if you perform work for a particular node.  Some nodes would be "greedy" and keep all of the coins themselves and others would be "generous" where it can be confirmed that coins have indeed been distributed due to mining activity.

Special "messages" could be defined for those nodes that implement this particular "service", which would take care of actually sending and receiving the handshaking data necessary for such distributed mining activity.

Note that none of this really needs to be put into the official client right away, but if it was implemented in this manner with coordination through Satoshi in terms of implementing this as another service, it seems likely that it might be "blessed" as official in a future update.

Another advantage of doing it this way is that it would be done through the "peer-to-peer" network itself, so you could have some "part-time" mining servers that could go up and down from time to time.  If one server goes down, you can switch to another server as it is merely another nodes that is acting as a pool collector.  The switch could be automated, with priorities going to nodes that have a high trust value (and a trust metric that can be tweaked "manually" by the person running the client software).

This sounds like a fun project to try out.

[Anonymous]

 Would be cool if you could find out on the fly which server is going to give you the best result. 

[jimbobway]

I think implementing pooled mining in the main client is necessary for bitcoin's future.  Right now less and less people are not generating because of GPU domination where a handful of users are controlling the network.  Pooled mining will also give newbies a chance to generate.  The more people that are interested the more valuable and important bitcoin is.

[Timo Y]

I'm a purist and think that the official client should remain as simple as possible.

I don't think it should be bloated with mere nice-to-haves.

There are many different ways of doing pooled mining and many independent projects. Which project should then be chosen as the "official" one?  Too much centralisation IMO.

Also, even pooled mining with lose its appeal to Joe Average 1-2 years from now because by then the real specialists will dominate the mining market (with specialised hardware and such). 



 

[RHorning]

I'm a purist and think that the official client should remain as simple as possible.

I don't think it should be bloated with mere nice-to-haves.

There are many different ways of doing pooled mining and many independent projects. Which project should then be chosen as the "official" one?  Too much centralisation IMO.

Also, even pooled mining with lose its appeal to Joe Average 1-2 years from now because by then the real specialists will dominate the mining market (with specialised hardware and such). 
 

Fine, don't make it an official client, but you certainly can have other clients that implement the main protocol and put in extensions.  That, BTW, is one of the reasons I'm trying to document the protocol in the first place because I also agree that a project of this nature shouldn't have everything pushed into a central repository that can be attacked.  Gnutella survived precisely because its specification was published and many variation of the client were created.

Of course it took shutting down the distribution of the "official client" to make the variety of clients pop up, but it still happened.

[FreeMoney]

I'm a purist and think that the official client should remain as simple as possible.

I don't think it should be bloated with mere nice-to-haves.

There are many different ways of doing pooled mining and many independent projects. Which project should then be chosen as the "official" one?  Too much centralisation IMO.

Also, even pooled mining with lose its appeal to Joe Average 1-2 years from now because by then the real specialists will dominate the mining market (with specialised hardware and such). 

 

I agree it should not be in the main client.

I think that even if it becomes wildly unprofitable new bitcoiners will still find it very cool to contribute for a little while in order to get a few fractions of a coin to play with before accepting them or buying some. Kind of like a faucet where you pay your own way.

[teknohog]

I'm a purist and think that the official client should remain as simple as possible.

I don't think it should be bloated with mere nice-to-haves.

Agreed. I think private mining pools are OK, but the idea of p2p pooled mining sounds like an attempt/desire to reinvent the entire protocol. The bitcoin network is already a distributed p2p miner, it just happens to create money in blocks of 50 for sound technical reasons.

[FreeMoney]

It's probably quicker and easier to set up an mtgox account and download a miner than to get the client going anyway.