Should a GPU-based miner be integrated in standard client?

[amincd]

It was once possible to mine for bitcoins on a standard client, but with GPU mining and mining pools becoming necessary, the CPU-based miner in the standard client became obsolete and was removed. I think having a GPU-based miner that can connect to pools integrated right in the standard client would be good as it would give people a greater sense of being part of the network, and allow them to earn their first bitcoins more easily. What do you think are the pros and cons of this addition?

Edit* I'd like to also submit this as a possible way to implement this change without hurting any parties:

For you miners who have invested considerable amounts in your rig on the expectation of the difficulty not increasing significantly, and consider this change as potentially creating an unexpected increase in difficulty and hurting profitability of mining, would you think it a fair compromise that the miner be integrated into the client, but only after a six month delay, to give you more time to recoup and profit from your investment?

I'd say there's a good case to be made that more miners, which this change would create, would increase the demand and price for bitcoin, since it would help increase adoption and bitcoin activity, so this six month window would give you time to save the coins you're mining, so when the price goes up, you stand to profit off of the price hike.

[ThePok]

+1    Maybe the p2pool pool. (No servercosts).  And more miner make the network saver. So make it as easy as possible to mine

[DeathAndTaxes]

No because over time mining will become a more and more specialized industry.  Simply having a GPU is insufficient.  You need the right GPU, an efficient rig, proper cooling, understanding of thermal stress and effects of 24/7 operation at high load, and a better than average electrical rate.

As Bitcoin grows and more money is put into mining the margins will continue to shrink. It is inevitable that FPGA will replace GPU and as they do drive the price:difficulty ratio so low as to make GPU mining uneconomical (much like GPU did to CPU).

Much like most consumers never operating a VISA processing server in the future most consumers will never perform blockchain hashing.

[gmaxwell]

No because over time mining will become a more and more specialized industry.  Simply having a GPU is insufficient.  You need the right GPU, an efficient rig, proper cooling, understanding of thermal stress and effects of 24/7 operation at high load, and a better than average electrical rate.

As Bitcoin grows and more money is put into mining the margins will continue to shrink. It is inevitable that FPGA will replace GPU and as they do drive the price:difficulty ratio so low as to make GPU mining uneconomical (much like GPU did to CPU).

Much like most consumers never operating a VISA processing server in the future most consumers will never perform blockchain hashing.

Today is not the grand future.  Casual mining is still a reality and still probably will be for at least another year.

It's true that you're not going to run some big serious profitable mining operation without some ninja mining skills... but there are plenty of reasons besides that. If you just want a little bitcoin to play around with or to offset various fees w/ bitcoin services then a little mining can provide that easily (even on a CPU!).

The fact that your setup might not be technically "profitable" on a strict electric power vs bitcoin exchange rate is moot— sending money to an exchange or some OTC trading partner is quite costly in terms of time (and sometimes fees)... while doing a little mining and paying for it in your power bill is painless and even if your operation isn't the most efficient the "fee" you pay vs buying the coin is fairly modest.

[amincd]

+1    Maybe the p2pool pool. (No servercosts).  And more miner make the network saver. So make it as easy as possible to mine

That would be ideal, but p2pools are pretty small, and therefore the payouts are irregular. Given the large number installed clients relative to the max size of a p2pool, there would need to be multiple p2pools so it would also need to be determined how the client finds other clients to create a p2pool with.

No because over time mining will become a more and more specialized industry.  Simply having a GPU is insufficient.  You need the right GPU, an efficient rig, proper cooling, understanding of thermal stress and effects of 24/7 operation at high load, and a better than average electrical rate.

I agree with gmaxwell's point that trading a slightly higher electricity bill for some bitcoins would in many cases be a better way for a casual newcomer to acquire bitcoins than having to sign up to an exchange and make a bank transfer.

More casual miners would mean a greater share of people who download the client actually getting to play with bitcoins.

[o]

Adding mining ability to the client will encourage more people to use it. It will increase the number of nodes in the network and also more copy of the block chains, which will make the network more strengthen.

[cbeast]

Someday yes, but not at the moment. It will upset people to get high electric bills if they don't understand about mining. When Bitcoin is used for multiple applications where is behooves someone to leave the client running 24/7, then the option of making transaction fees will be a plus.

[ovidiusoft]

No. Most users will not mine (will not even understand what mining is or how the bitcoin network works). No need to bloat the client.

[localhost]

I think it does no harm to include it as an unobtrusive option, so that:
- it doesn't bother a casual user that doesn't have a clue (do we really have those right now anyway?)
- it is practical for a user who doesn't want to bother grabbing an independent client. It would be nice to have at least a few options, though, such as configuring where to mine. A bit like *cough* SC actually.

Clearly not a priority feature, though.

[amincd]

The interface could be something like this:

pool website or IP:

port:

bitcoin address to receive payout:

Where only pools that don't require registration, like Eligius, can be used. That would keep it simple.

[Cdecker]

No, the client get's bloated for the benefit of a minority that will use specialized software anyway.

[Gabi]

Dunno.

Bitcoin mining is an interesting concept, even if you don't become a professional miner, it help you getting a grasp of how the thing work

But as others said, putting it in the standard client may confuse new users, also since it use the GPU it would create problems if they don't know how GPGPU computing work (aka, start GPU mining and start a game, the perfect way to freeze the computer and being forced to press the reset key).

[Luke-Jr]

I think it would be nice to get OPTIONAL support in Bitcoin-Qt for embedding an external source tree of cgminer into the client. With Eligius, it doesn't even need configuration: just check the "Generate Bitcoins" box and watch them come in

[unabridged]

the more widely bitcoins are distributed, the better it will do. there should be a way for a casual user to select a menu option in the standard client to start gpu mining and see bitcoins flowing into their account without having to edit config files, use cmd line options, or download more software. just popup a small warning the first time they start it "Mining Bitcoins may slow down your computer and increase power usage". by default just use an embedded cgminer low intensity connected to p2p pool and have the user select an address to payout to

[Transisto]

Yes it is important, because it will help many people understand what's at the CORE of Bitcoin, How distribution is FAIR and not some ponzi scheme.

But warning, no mining code should be added to the client.
 
The implementation I suggest is limited to a drop down menu with choice of top miners weblink, documentation/download page.

That remind me of when LiteCoin mining began,  The miner provided in the windows client was utterly inefficient compared to others.  There was no linux binaries for these, they had to be compiled and required loads of dependencies, I never been able to do that and cancelled the project. 

[westkybitcoins]

I see one possible reason mining should be somewhat simple to activate for the casual user: network strengthening.

Not only will it bump up the strength of the network slightly (even if the average casual user only leaves it on for a few hours a week) without requiring a price increase, it'll also allow those who follow bitcoin news to be able to actively contribute to defending against a network attack. If there's a sudden noteworthy increase in hashing power, but the new player seems malicious, the ability to quickly mobilize a little extra hashing power might be useful.

As far as how to best implement that? Meh, I probably wouldn't bother if it wasn't just some menu option in the client itself, but I can also see the downside of affixing another rarely-used chunk of code to it.

[coin_toss]

Definitely add a mining option - even if serious miners don't use it. It would be a great selling point to bring in mainstream users.

[HostFat]

I think that a link to a well organized Bitcoin-wiki page with all the information about mining and miners will be a good solution.