GRCStarter@Home Proposal

in #gridcoin7 years ago (edited)

GRCHeaderBlur - Green.png

GRCStarter@home

This proposal describes a project called GRCStarter@home. GRCStarter's operation is similar to that of WCG -- it is a single project that works on many projects. The difference between GRCStarter and WCG is in their priorities and how they distribute magnitude to their crunchers.

Priorities

  • GRCStarter's overall computing objective is to support researchers or project heads with limited or sporadic Work Units (WU), or few volunteers working on their project.
  • GRCStarter's second priority is to introduce new users to the different projects and how to choose the right one for their hardware.
  • GRCStarter's third priority is to help new users build a base magnitude.
  • GRCStarter's fourth priority is to educate new users on the inner workings of grid computing, BOINC, PoS, and DPoR.

Initiation and Operation

To initiate GRCStarter, a poll is started which contains proposals from researchers, project heads, and community members regarding which project's WU, called a Work Unit Set (WUS), to accept. This poll can be for any length of time. The top X number of WUS, let's say 10, get added to a WUS Ledger. With a complete WUS ledger, a month long poll is created with the options to vote for the 10 WUS in the ledger.

Crunch Units

GRCStarter splits the month -- the duration of the operating WUS Ledger -- into units of crunch time. It distributes these "Crunch Units" (CU) to a WUS based on the WUS's weight in the running poll. A WUS uses these CU to reserve crunch time on GRCStarter. When half of the month's CU are distributed, a poll is created to choose the WUS for the next month's ledger. This is similar to project rain, but has an economy contained within a single project instead of occupying the entire BOINC economy.

Magnitude Distribution

GRCStarter distributes equal magnitude to all hosts working on GRCStarter, regardless of a hosts relative RAC. This magnitude value is determined by giving all hosts a superblock RAC equal to the average RAC of GRCStarter. Hosts with high mags will be helping new users build their own magnitude. They will also be supporting projects that need help getting started or finishing a small amount of work.

User Experience

A new user will download, install, and set up the Gridcoin client. The user will download and install BOINC and sign up for the clearly marked GRCStarter@home project. GRCStarter's website is a hub of information for new users. Whitelisted projects are listed and described, and there is accessible, up to date information for helping people find the right project for their hardware. Further, explaining how GRCStarter works to new users will introduce them to RAC, Mag, WU, DPoR, and superblocks at the same time and in the same place. From GRCStarter, users can launch to other resources if they wish to learn more about the inner workings of Gridcoin.

Thank You For Your Feedback!

What do you think? What pros and cons to this proposal do you see? What risks? How does this compare to other proposals seeking the same goal - helping projects that need help and helping new users?


The idea behind this proposal comes from @nightshift1134's post A New Project Called Rain.

This specific proposal stems from a conversation between @Vortac and myself which was sparked from a conversation on @dutch's post, Don't Give In to BTC FOMO - Why I am Buying GRC Instead

and ties into the continuing conversation from these posts:

  1. Long Term Goal of Gridcoin by @applepiee
  2. Response to Appliepiee's Post "Long term goal of gridcoin" by @jringo
  3. My Thoughts on the Value of Gridcoin by @decayingzombie
  4. The Value of Gridcoin - A Continuation... by @jringo

NOTE: Read @applepiee's comment for discussion regarding the distribution protocol

Sort:  

I like the idea of a Gridcoin Umbrella project but why do you have to remove the competition? It is the thing that makes BOINC successful.

Giving each host the same mag in my opinion would make the project a none starter.

You will end up with only the worst hardware running on the project (unless that is the aim). Why would any science projects sign up if the hardware is not good enough or the project does not attract enough researchers?

Well why cant just one project not be competitve?

And like i said below make it a small/mobile/arm only so that people without big pc or farm have a chance to contribute. My samsung galaxy tab did aboit 1000 credits a day but took forever to do so and only with moowrap

A huge benefit to a proposal like this, if we can work out the protocols, is that it can be replicated. What does this mean?

You can have a projects like @nightshift1134 and you and theearl on Slack have suggested. GRColdhardware@home, GRCmobile@home, whatever. This makes it so that people can still earn GRC based on the work they complete, as Viggers619 suggested on Reddit.


I like the idea of a Gridcoin Umbrella project but why do you have to remove the competition? It is the thing that makes BOINC successful.

I'm not sure what you mean by remove the competition? It replaces one type of competition with another, it doesn't remove competition -- it's very difficult to remove competition from anything people do.

It also does not remove gamification -- the thing that makes BOINC successful. I would argue it helps it because it is a single project among many and its intent is to help new users learn about projects, BOINC, and Gridcoin while also strengthening the BOINC project ecosystem by supporting new researchers and projects.

It's like a handicap in smash brothers -- new users vs long time users = new users get a handicap

I'm not sure what you mean by remove the competition? It replaces one type of competition with another, it doesn't remove competition -- it's very difficult to remove competition from anything people do.

By changing the magnitude calculation to a socialist style everyone gets the same. It removes the competition from the project. Why would someone BOINC with their most powerful hardware when some of the mag will be distributed to users using low powered hardware?

They would be taking their powerful hardware off other projects. Unless I'm misinterpreting your meaning.

Big guys would just temporarily crunch to help a new/small cruncher and resume whatever project they were on.

To keep myself from going insane, can we not label things with terms that have very complex histories and meanings, like socialism -- socialism is absolutely nothing like "everyone gets the same." It's actually very far from it. It's not even always viewed as an economic model (because it's not). It's often seen as the process of transition away from Capitalism and toward something else -- anything with more social priorities instead of individual priorities: the socialization of economic priorities. Richard Wolfe is a brilliant and reasonable Marxist based pragmatist who combines Marxism and Capitalism.

Anyway,

They would be taking their powerful hardware off other projects.

You're absolutely right. So ask yourself.

Why do people donate to the faucet? Why do people give away GRC when new users come and ask for starter coins? Why do people use the tipbot in IRC? Why do people donate to people who produce things? Because it feels good, is one reason. Because it helps build a user base which benefits everyone, is a second. Because it shows off their philanthropy is a third, and this philanthropy benefits them in some way. So why would someone donate some of their BOINC credit production to GRCStarter?

A scenario:

If 500 people with 500 RAC put 1% of that RAC toward GRCStarter@home you're talking about 2500 RAC. If someone with 1 RAC needs to charge their mag so they might stake some starter coins, they can join up with this group of 500 and, using the current flawed distribution method, get a RAC of 4.99201 instead of 1 ((2500+1)/501). This is nearly five times what they would be getting if they were solo crunching. What a great help for charging their mag! This also means that the 500 heavy magsters will be receiving the 4.99201 RAC. This is 99.8403% of the 5 RAC they each contribute. Since they're getting %99.8 of the value of their invested RAC, their loss in this scenario can be considered negligible. Not bad. I just pulled these numbers out of thin air. I'd love to see what happens in different scenarios. Prove me wrong. Just remember, the distribution protocol will be changed if this proposal moves into production.

So that takes care of the first two reasons people donate. The last has to do with where the competition has been moved.

Those who help new users generally gain more respect from the community. So even without an incentive protocol, there is reason for heavy magsters to throw a percent or 2 or whatever at GRCStarter@home. Who donates the most! It's an open ledger so everyone can follow a CPID and how much RAC that CPID is sacrificing (or gaining) to the GRCStarter average RAC. It is possible to add an incentive structure on top of GRCStarter@home, but not necessary, though worth exploration.

This process creates a GRCStarter economy which depends on RAC, the WUS being crunched, the average RAC of GRCStarter, the number of underaverage (new) users, the number of overaverage (donators) users, and anything else we put into the protocols. So the cost/benefit of GRCStarter will change based on many factors. It may be more beneficial for someone to donate at one time instead of another. This means there is a profitability factor involved. The profit simply isn't GRC. Instead, it's respect, trust, visibility, and a well educated and active user base.

You bring up some valid reasons that I had not thought about, however, I don't ever see this complexity being added to the current system.

Maybe it could be something you modify in the BOINC credit system itself when you setup your BOINC project. But remember it has to be crunching something worthy or it will never make the whitelist.

If the aim of this is to make it easier for New users to stake I think there are better alternatives. One might be to use delegated proof of stake similar to STEEM and BitShares and have the rewards paid out regularly using a transaction so that staking is no longer required.

However, I'm not sure what would happen to the price if people no longer required a balance to stake research rewards.

Meh -- who cares about the price. We're here for science and that is what is going to bring monetary value.

The goals of GRCStarter@home are as mentioned above:

helping projects that need help and helping new users

Its priorities are:

  • GRCStarter's overall computing objective is to support researchers or project heads with limited or sporadic Work Units (WU), or few volunteers working on their project.
  • GRCStarter's second priority is to introduce new users to the different projects and how to choose the right one for their hardware.
  • GRCStarter's third priority is to help new users build a base magnitude.
  • GRCStarter's fourth priority is to educate new users on the inner workings of grid computing, BOINC, PoS, and DPoR.

This means that all actions taken by GRCStarter will seek to fulfill a top priority before those below it. If it can achieve multiple priorities at the same time, all the better.

So really, GRCStarter is about helping researchers and projects - BOINC - while education new users about BOINC and Gridcoin.

Delegated PoS, or SuperNodes, is something completely unrelated to this project = ), but something I support looking into for GRC.

In relation to magnitude, the point is to reduce the Mag charge time from a few weeks to a few days. We all know how frustrating it is to start the client and not get rewarded as your mag charged up (because your RAC has to build).

I think the best option is to completely rewrite BOINC and GRC as a unified platform (a volunteer oriented grid computing incentive based blockchain), but who here has the money to fund that?!

So is it a pool that redistributes other project's work? Or a project which distributes the work which the public submits for computation? Or simply a faucet for new users?


GRCStarter distributes equal magnitude to all hosts working on GRCStarter, regardless of a hosts relative RAC.

What's stopping me running many user accounts to earn more than my fair share? I have no doubt that this will be abused by many.


Regarding the projects name, perhaps including GRC in the name will put off those outwith team Gridcoin? At least until the team req is removed that is..

Plus, it'll appear to many as an 'official' gridcoin project which may reflect badly upon the grc network if the project closes down in the future..

Were there any alternative names that came to mind before 'grcstarter' ?

First question, could be all three or the latter two.

Second question, yeah it could be abused but with @bullsharks suggestion make it only for small/mobile/arm devices

Third question, i knew "rain" was taken @jringo came up with grcstarter@home but if we aim for small/mobile/arm it could be mobile@home or arm@home or something like that.

arm@home sounds good :-)

I still don't like the idea of modifying the way magnitude is given out though. If it is modified I would assume that you would end up with only ARM devices anyway because why would someone run a powerful machine just to give a portion of their mag to someone else?

anyone else looking to an answer for "Why donate," read my response to @bullshark's main comment.

The most practical way to look at it is as a clone of WCG.

It is a blockchain ledger, beginning as a centralized project, which distributes processing power to projects which might not otherwise have enough work or participants.

The centralized version determines the distribution of this work based on a running poll in the Gridcoin client.

What's stopping me running many user accounts to earn more than my fair share? I have no doubt that this will be abused by many.

See @applepiie's comment

Were there any alternative names that came to mind before 'grcstarter' ?

This is only a placeholder name.

It is a blockchain ledger, beginning as a centralized project

What? So it's a BOINC project at first but then it becomes its own cryptocurrency/blockchain?

I will be sure to elaborate on the potential of integrating a blockchain into a BOINC project in v2 of the proposal.

Dude first you took my idea 0 to 100, now yiu done went from 100 to 1000! Lol cant wait to see this idea of yours.

haha I hope it works! xD

How about creating a GridcoinARM@home project which is an Umbrella project that only supports ARM devices. It could target Grid computing with even lower energy usage and it would be easier for new users to compete because it's cheaper to purchase the most powerful ARM device.

It would also be great for the Raspberry Pi Foundation :-)

Could do that, would also get mobile phone users to boinc especially if thats all they have.

I think the most we could offer here is a WCG-like umbrella for little projects, that can't get added to WCG (because IBM is probably strict with their criteria) and sporadic work makes them unsuitable for full Gridcoin whitelisting. Magnitude boost would always be a target for potential abuse and could end doing more harm than good.

Right that seems like the biggest two problems were running into atm 😢 cant never tru to the right thing without somebody trying to ruin it.

This sounds great and I would have been thankful for such an "introductory project" that helps me understand the background of Gridcoin.

While reading

GRCStarter distributes equal magnitude to all hosts working on GRCStarter, regardless of a hosts relative RAC.

I thought to myself: "I have a Raspberry Pi Zero W laying around. I will throw that into the project and get out more GRC than I ever would letting it crunch on its own."

I know this sounds pretty egoistic and it was just a thought, probably nothing I would do. The point I want to make is, that I would not be the only one having this thought and the project might end up as a collection of RaspBerry Pi Zero Ws doing almost nothing and making it impossible for new users to make actually use of the really great idea this project is.

Good point! It appears that if we want to move forward with this proposal we will need to work out that issue.

forgive me if this is stupid, I'm new to this. But instead of the mag being the same, can every CPID just get the same amount of WUs? That way the people with enough computing power will complete all of their WUs and get the same RAC thus, the same mag. But if you're running a raspberry pi, you won't complete all your WUs and you'll get a lower mag.

So the protocol would be some combination of Project Average RAC and WU completed... this definitely holds potential!

If the slower machine gets the same amount of WUs like the faster machine, you are basically forcing the faster machine to run out of work, so that they are all 'equal' in the end. It goes against 'the faster, the better' principle which is common sense in distributed computing.

Well, if someone ran out of WUs for GRCStarter@Home, then they'll probably start crunching some other project. Science still gets done.

The point of distributed computing is to welcome more computing power, not to turn it away by artificially restricting the number of available workunits. Projects which don't have enough workunits are usually removed from the whitelist, makes very little sense to create such a handicapped project intentionally.

Yeah, in its current proposed state I wouldn't vote to whitelist this project.

of course not! this is a very minimal proposal version put out there to get feedback from the community and to see if anyone wants to help make it a reality.

The point of distributed computing is to welcome more computing power, not to turn it away by artificially restricting the number of available workunits.

Absolutely agreed! However keep in mind this is just a single project out of dozens. The point of the project is explicitly stated as a project which provides a mag boost for new users while helping new projects complete work and gain visibility. This means that anyone with a lot of processing can offer some up to the GRCStarter@home project with the intention of helping new users instead of gaining Mag. Once new users learn enough from GRCStarter -- or perhaps there is a time limit for low mag CPIDs or some other mechanism -- they are encouraged to move on to a project or projects they wish to crunch.

There are many details to work out, but a handicap project does not mean that WU will run out. It will constantly be crunching. What @personthingman2 is describing holds potential as a protocol. Hard to say right now if it will work, but it's definitely worth exploring as we build protocols.

My understand of the rewards is that its an equal share of the total rac/mag right? So that way there isnt a beast like you have with bitcoin, yes some people my abuse this project just like some people spam steemit lol
But for the ones that are legitimate this project can do good.

sure, but i think if there is a mathematical/algorithmic way to prevent or discourage people from taking advantage of the system we should take that route. I'm sure there must be a way somewhere out there.

Remove trust from the equation.

Rasberry pi's would be pointed to wu's that they are good in and most noobs prolly dont know how to set up a pi and if they they prolly would be doing something else with it.....

I could make a hundred CPIDs, keep their RAC low, and direct them to GRCStarter@home to boost their Mag outputs. This is the problem

Wouldnt that require a hundred emails and wallets? And to do all that just to take advantage of this system is against the whole idea, there has to to be a way to detect and deter that.

i agree there must! we will find it

proportional rac for proportional computation sounds fair to me :P

= ) this project is about helping new users so it cannot be proportional.

Then it'll be abused by users creating many accounts to siphon the rewards from the system.

You're right! We will have to work out the protocols.

Loading...

I like this idea actually. The learning curve can be a bit steep, and when you just get in to it it can all be a little daunting.

Since this is supposed to help newcomers, how will this project keep well established "miners" in search for highest possible magnitude out?

Boinc was created to "volunteer" a users pc for the betterment of science, this project isnt for well established miners to increase their mag, its for said "miners" to continue reasearch while helping new people learn and earn.

Most of the projects already have high rac/competition which could discourage a new person but this project could welcome new BOINC crunchers while giving them a good RSA to make earning worthwhile while learning.

I get that, but not everyone is this altruistic. There are people that simply hunt the project with lowest possible competition and point a ton of hardware at it. These people would probably swarm this project, making it harder for newcomers to make a dent and compete for a reasonable magnitude.

Also, remember the priority is to help build new user magnitude, so any high magnitude cruncher will always lower their mag if they point everything at GRCStarter. It is not a matter of a person being altruistic if it is built into the code.

you're right -- the distribution protocol described in this proposal does not account for greed. it will have to be changed. Any suggestions? = )

It could be managed through the voting system? If someone is racking up too much they're not considered "newcomers" and can't participate in the project. Then the community as a whole would have a say in the project.

What do you mean by

If someone is racking up too much

Sorry I'm unclear. If for example someone decided to point a server farm at the project, they'd probably destroy any competition from newcomers pointing a single or two home desktops at the projects, thus destroying their chance to get any meaningful magnitude.

They would actually lower (by a lot)their own magnitude and help many new users. Their mag will be reduced to the average of the GRCStarter mag after their huge RAC was added.

Get banned -> Split CPID -> Evade ban.

GRCStarter@home wants miners with huge mags to direct some processing power its way. If what you're talking about is having a miner with large mag directing a single CPU or whatever to GRCStarter to boost their mag with the distribution protocol, i think you're right. That is probably not the best way to distribute magnitude. =)

This post has received a 0.63 % upvote from @drotto thanks to: @banjo.

maybe iam not seeing exactly what this is aiming at. What portions of getting to know Gridcoin do you want to adress with that ? The entry into the cryptocoin world is anyway something that needs a bit of time. the same counts for the entry into Gridcoin ... For sure there are things that really could be much more user friendly and much more self explanatory. But i dont see how a umbrella project like this could help here ...
sorry i just dont get it :D

Accessibility and education are only the second and fourth priority of the project. In the simplest terms, these are improved through this proposal by having a single website with the introductory information of the ~20 white-listed projects instead of 20 websites each containing sparse and technical information on their own work.

A bus station instead of a bus stop.

Coin Marketplace

STEEM 0.30
TRX 0.12
JST 0.034
BTC 63688.35
ETH 3125.30
USDT 1.00
SBD 3.97