GridCoin Idea: More Whitelist Project Data for New GridCoinians

in #beyondbitcoin8 years ago

There's always a little part of me that wants to be selfish, and mine as many coins as I can before offering up any ideas that could cause GridCoin adoption to explode.

Then I remember: that's not what GridCoin is about. We want to make distributed computing mainstream, so that any scientific endeavor that needs free CPU/GPU time can contact BOINC and BOINC will always have a plethora of willing crunchers.

gridcoin.png

What's the biggest hurdle for new GridCoin Crunchers/Miners?

I was asked this on Reddit today and I didn't have a great answer. Some people want to say it's the setup of both GridCoin and BOINC. I think that's a perfectly fine gripe, but I wouldn't call it a major hurdle. It's just going to take a little bit more of your Friday night than you were probably hoping for. We all sit behind the computer on Friday night anyway, right?

Right?

Actually, I don't want to know the answer to that.

After some thought, the biggest hurdle is after you get the GridCoin wallet going and you've got BOINC running, there's no clear-cut way to know which projects you should be crunching. You can get an indicator if you happen to find the Projects page on GridCoinStats.eu -- just pick the projects with the fewest number of users and you're set, right? Not really. It's a bit more complicated than that, and one could waste weeks trying out each project one-by-one and waiting for a few days to see what kind of Magnitude they're getting.

Would it be possible for us to build a whitelist table that would show a few pieces of additional data such as...

  1. Which projects are primarily GPU projects and which are primarily CPU projects?
  2. An estimate of how much Magnitude is earned per 1,000 units of RAC.
  3. An estimate of how much GridCoin is earned per 1,000 units of RAC.
  4. Perhaps a simple algorithm that takes the available data (GRC users, GRC users' average RAC, available workunits, etc) to rate each project on a scale of 1-10 on competitiveness.

I understand there are equipment considerations when trying to decide which project(s) a user should consider, but I think a handful of data points as simple as those above would prove invaluable for new users, especially considering GridCoin has no functioning profitability calculators at the moment. This would prevent new users who are looking to maximize profits from spinning their tires too much by arming them with enough data to at least get started, and it would take some heat off of us on the GridCoin subreddit who answer this question seemingly on a daily basis.

What do you folks think? Perhaps this could be something that could be set up on GRCGo.com or even on the GridCoinStats.eu website.

~XaqFields

Sort:  

In my case, I was using BOINC many years before Gridcoin came about and the standalone seti@home before that so I already have my favorite projects. I love being able to earn Gridcoin while doing this but I generally stop short of leaving/joining projects just to maximize profit.

Kudos to you! I was hoping most people who convert from BOINC would stick with the projects they're most passionate about rather than profit chasing... even though that's kinda what I do. To be fair, my theory is that once the team requirement is lifted, we'll have so many people profit-chasing that it should serve to equalize the value of each project to the point where there's not so much variation from project-to-project.

That wins you a follower :)

Though I admit I get frustrated when my magnitude goes to 0 for no discernible reason like it did this morning...

Its pretty rare for this stuff to happen, I just let my machines run and do what they need to do

Mostly it has been that way for me as well. Except a couple of months ago I apparently did not upgrade from one particular version of Gridcoin to another in a timely enough fashion and my keys somehow disappeared (the backup I had was not valid...I think this was during a hard fork). I had to send a request to delete my beacon and had problems with that. I was out of commission as far as Gridcoin was concerned for a few weeks. It finally got resolved.

Found what the current issue is all about though: https://steemit.com/gridcoin/@jringo/gridcoin-0-magnitude-updates-here

Yeah just watch the channels and you should see any mandatory upgrades, they do try to keep them to a minimum, be on a lookout for mandatory to solve this latest issue

@scalextrix i should introduce you to spintax ;p ;p

I'm in agreement.

I spent two weeks trying to figure out the RAC and which projects would give me the best mag; at the end of the day I walked away seeing it as a challenge when it was done. But I've read the frustration people have here and there about the unknown.
I think it'll be people that get excited about the projects that will convert to it, they'll overlook the difficulties easily. The average user... Might not. A universal number would work, One number to look at that tells you which project will give you more bang for your buck.

The main gridcoin.us website currently doesn't run any scripts which scrape data from external websites, gridcoinstats.eu does a really good job at presenting this kind of information you're requesting.

Yeah I was thinking a 3rd party site would be best for something like this anyway. I don't really have the programming skills, myself. I could set something like this up on a Microsoft Excel workbook, calculated manually but not on a website. Perhaps the best way to approach this would be for me to calculate it manually and post it up here on Steemit once a week or so, and then if someone with the programming know-how likes it, they might choose to recreate it on a 3rd party site.

What I'm thinking of is something like a calculator that really spoon-feeds a new user the bottom-line data they need to pick a project. Each 1,000 RAC on "X" project is currently worth 15 Magnitude, which translates into 5.52 GRC per day.

From there it would be super simple to browse through the userlist on the GridCoin team and find someone with similar hardware to see what their RAC is to estimate what you might be able to earn.

PrimeGrid is a great example. There are actually quite a few users on that project so when you look at the Project list on GridCoinStats.eu, it's not the first project you're going to pick. It turns out the competition in that project really isn't very high, though, and it's pretty easy for a single GTX 1070 to yield you a 200 Mag or so (currently).

I'll think about how I could set something like that up to spit out here on Steemit. That way when we get that question on Reddit, we can just link people to the most recent article on this site.

Primegrid is a special case. PPS Sieve is awarding more credits than other subprojects because you can't actually discover new primes when running PPS Sieve (you are doing some sort of "preparation work", check here for more details).

When running other GPU Primegrid subprojects (GFN), you can actually discover new prime-numbers pretty quickly and you get some recognition for that, so Primegrid admins decided that's already a big enough incentive and reduced the credits awarded for GFN (while increasing them for PPS Sieve).

The point is, any "rewards calculator" can be quite tricky to implement, since it will require in-depth knowledge of every BOINC project and they are bound to get more complex with time. In this case, it wouldn't be enough to point users to Primegrid, they would crunch GFN, earn mediocre credits and get disappointed. They must be pointed specifically to PPS Sieve, if you are promising big rewards to them.

P.S. I am a long-time Primegrid cruncher, got all this info by following their message boards for years, it isn't readily available anywhere else, as far as I know. BOINC admins can fine-tune credit reward system as they see fit, for every project and subproject. Perhaps WUProp can provide necessary correlation.

Gridcoinstats are working on a way to present datapoints like the ones you describe. We wish to make people able to figure out what you're possible to earn.

Stay tuned.

Very cool! I'm glad to hear it! This is the one item that GridCoin lacks that most other coins have. I know it's more complex, but even if we could get to a ballpark range of revenue for each project, that would be super helpful.

This is a great idea, would be great for new users.

  1. You can check in applications of every project if it has GPU tasks or not( gpu tasks has values like opencl or cuda in version column) . In case it has do not use your CPU for mine that project.

  2. I do not know how hard it will be to implement that algorithm, the stats can be taken from wuprop project or sites like boincstats.

I agree totally, bookmarked this article in my ideas folder. We need to integrate the information on how to choose projects right into the onboarding process for new users