RE: Burning The Value Into Gridcoin
I support the idea of burning Coins for private projects, if not as well of all voluntary projects. I thought of it also. As the supply of GRC is unlimited, we have to think of a mechanism to restrict / subtract Coins from the supply to gain some value of the Coin.
I do see Vortsac’s argument under 1) and 2) in his comment in the other thread, but I am not able to see his point 3) as I do not see it for Sparc, Golem etc. either.
Why would a private company take all the hassle to implement a BOINC project in the first place? As several projects demonstrate, it is quite complicated to implement a project in a cross platform environment: different OS used, different Hardware (CPUs and GPUs), age of the hardware and so on. I think it is much easier to go to a company with homogeneous Soft- and Hardware to set up a project, than investing in the set-up of BOINC, buying GRIDCOIN, convincing the community that the project is trustworthy to get it whitelisted and so on.
I think it might be much easier to convince a Hardware manufacturer or an Online Retailer to accept GRC as direct payment option (See for example NEWEGG accepting BTC), as a “marketing gag” or another form of supporting public/volunteer computer science as IBM does with World Community Grid. There you create real demand of the GRC. We cruncher might directly buy hardware with the Coins offsetting the electric bills to pay for, and the others “cryptocurrency” aficionado indirectly over the exchange.
What I do not see, if the company will come to the FIAT currency to pay the wages and so on… but this is a whole other story.
It is actually incredibly easy to implement a project on a full BOINC server. You can get a BOINC server up and running, generating work units, sending them out, receiving the results back and validating them all in the space of a single afternoon. The only real work at this stage is populating your work units, but most of the heavy lifting for that would need to be done regardless of what compute source you are using to solve your problem. If we are talking about hardware optimisation, then yes you are correct and we run into a whole lot more work.
The reason that a company or research institute would want to use BOINC is twofold:
First and foremost, Gridcoin already rewards BOINC crunchers for projects on its whitelist, for free. There is no cost to the project at all. Getting a project whitelisted is trivial if you are a little active in the community and have a solid research proposal with a steady supply of WUs.
Using BOINC requires practically zero hardware investment, and would be far cheaper than renting hardware for your compute cycles. Not only is the upfront cost far higher to invest in hardware, but ongoing costs too. If I get a research project onto the GRC whitelist and double its GRC mint with rain out of my own pocket, you can safely bet cruncher will flock to the work. It would still only set me back some USD$30/day, which is incredibly cheap.
I'm not just making all this up, for what it's worth. I'm a PhD student working on HPC models of the brain, and spend a lot of time tackling endless compute problems. I have also set up my own BOINC server. BOINC is unparalleled in potential for most research applications that are not timestepping simulations (but that is another discussion).
Exactly! @dutch @limacoin. Sure a private organization/company could set a BOINC server up if they wanted to, but that doesn't guarantee that people would devote processing power to it. Now if they paid to get into the in GRC whitelist, that's a different story. Plenty of people would run the project, especially if you rained GRC to the people who work on it. I still say we add the option for less "noble" projects, if they pay, and the community deems that they're good to do business with.
By less "nobel" I mean private orgs/companies that need the computing power for their own reasons. It would help get Gridcoin's name out their and increase the value of the coin, since we'd be offering a service on top of the currency.
This will never happen, by the very purpose of the whitelist. That being said, it is already possible to rain GRC on crunchers who are doing work on a project outside the whitelist. The only benefit a whitelisted project has is that the blockchain will pay crunchers by minting new GRC.
Using the cross platform nature of BOINC is an optional benefit, not a necessity. You can target whatever platforms you want (Windows only, Mac only, Linux only, Nvidia/Windows only...whatever you want). Also, it isn't necessarily difficult to write code that is easy to recompile for another platform. Depends on the nature of what you are doing. I agree though that I'm not sure Gridcoin is providing any motivation for private BOINC projects. They can already be implemented for free after all and I'm not aware of a lot of them out there. Adding the ability to pay for whitelisting with Gridcoin doesn't seem likely to attract more.