Poll to De-list Moo! Wrapper?

in gridcoin •  last year

After an exhaustive debate on Reddit, another stakeholder and I agreed that we should to have a poll on whether or not to continue to support Moo! Wrapper on the Gridcoin blockchain because of the non-knowledge producing nature of the project. @dutch (I think) had expressed will and ability in putting it to a vote. We wanted it to be democratic so please make it a participant poll. In addition, you might want to make sure that individual can actually vote before making it.

I am going to vote keep it for these reasons:

  1. It provides magnitude from old/aging equipment that would otherwise become E-waste. It runs nicely on decommissioned smart phones. It is also gentle on other equipment.

  2. It could be used as a hedge for workunits in the event that all other BOINC projects fail (unlikely) or do not have workunits. This is due to the fact it should take something like 150 years+ at current computational resource allocation levels to test all combinations. Also, it is producing a steady stream of workunits.

Reasons for de-listing:

  1. Fails to pass mission-critical test of benefiting science in a constructive way.

  2. Runs a brute force command that has been proven to work already.

If anyone has anything else to add please do so below.

General Questions

  1. What are the requirements to fund a poll?

  2. Is it 100,000 GRC?

  3. Is the GRC consumed when the poll ends?

Edit as of 01-16-2018 1824 UTC -8:

Link to Reddit discussion regarding this topic;

https://www.reddit.com/r/gridcoin/comments/7qre20/moo_wrapper_is_this_a_scientific_project/

Edit as of 01-16-2018 2024 UTC -8:

I am dogecountant. Prescott was taken.

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!
Sort Order:  

Regarding your questions:

It costs 100,000 Gridcoins to create a poll, however these funds are not consumed at all during the poll creation or at its conclusion. The 100k poll creation limit serves to limit spam and congestion in the poll system.

The actual cost to create a poll is the same cost as sending a beacon, roughly 0.1 GRC or so, depending on data sizes involved in the poll.

·

Is there a way to add a "write your own response" choice for polls? Of course the creator would pay the data transmission fee.

·
·

At this time, no. The poll creator must specify all options that can be voted on.

·
·
·

I've a mixed view on Moo-wrapper. RC-72 calculations are a wasteful use of CPU/GPU cycles on the other hand the Golomb ruler (OGR) calculations are not.

Golomb rulers are for example used in:

  • Information theory and error correction
  • Radio frequency selection
  • Radio antenna placement
  • Current transformers

Furthermore distributed.net did find the Golomb rulers 24 to 27 (the highest four found at the moment).

Anyway I do support a vote.

·

Thank you for your contribution.

As much I have defended Moo! Wrapper, I only want a knowledge producing ecosystem. Anything other than that would be un-useful.

·

The highest four? So 24, 25, 26, and 27?

·
·

Correct, info can be found here: https://en.wikipedia.org/wiki/Golomb_ruler

Participant polls have no history of being accepted for action on large item proposals. They are not as secure as mag+stake.

Large action polls (development, funds, and whitelist), while having no official structure, parameters, or processes, generally require a large % of participation (15% 20% 25%? It is difficult to say due to the scattered information out there).

I doubt a participation poll will be supported by those with control over the whitelist. You may wish to confirm this before making the poll.

We are currently having some amazing discussions regarding what makes a poll "actionable". This proposal will undoubtedly be part of that conversation.

·

I doubt a participation poll will be supported by those with control over the whitelist. You may wish to confirm this before making the poll.

It would not be. Only mag+balance polls are actionable at this time.

I must say I find such a poll a total waste of time. Why vote on a specific project that at the time you feel is not scientific enough? Just make a proposition on adding extra conditions on the whitelist and in that proposition define exactly what makes a project "scientific" enough to be whitelisted. Otherwise people will just keep making polls like this one without a clear goal and at least personally I find it both annoying and also I think it just splits the community every time such a poll is made.

Just make a propositon on changing the whitelist conditions.

·

I agree with you on that point about potentially changing the whitelist conditions. The poll is still not back on that yet though. It will expire January 30th.

The idea behind it not being considered "scientific" enough was that it is basically wasting electricity like Bitcoin does but in an even worse way because once the project completes all of those resources will have been spent and nothing (besides a small monetary reward) will come of it.

I disagree that the poll is a complete waste of time. Any group of individuals making a choice together is a worthwhile endeavor and will hopefully leave us in a better state of consensus.

The community is already split on this project (atleast a small portion), which is why the poll was created.

I am encouraging people to keep it. I believe that it would be exclusive of us to exclude it, therefore degrading our community's brand and potentially impeding our mission. Gridcoin needs to be inclusive and defend any project that is willing to comply with our current whitelisting parameters.

Someone posted good reasons to keep the project and to not have had the poll in the first place here:
https://cryptocurrencytalk.com/topic/94283-poll-to-de-list-moo-wrapper/?tab=comments#comment-429204

If there were different tiers of white listed projects I would vote to keep it. However, since the probability of obtaining some type of publishable result from this project is very low, I think it should be de-listed.

·

I have to agree on that point.

Maybe we could keep it until we find a muti-tiered system for allocating Gridcoin Rewards?

·
·

Yeah, maybe it is good not to remove it immediately since some new users are benefiting from Moo being less competitive.

I would love to see whitelist tiers resembling something like this:

Tier 1 - High quality publications in reputable journals and + ample work units. Examples: Rosetta, World Community Grid, Einstein, Milkyway, etc.
Tier 2 - A few publications in reputable journals + ample work units. OR High quality publications in reputable journals + low work units. Examples: GPUGrid (sporadic work units), climate prediction
Tier 3 - No publications/low probability of future publications and/or very dubious scientific or historical benefit. Example: Moo

Scientific publications are the currency of academic science so if we want to follow the scientific world we will have to encourage crunching on those projects that have the highest chance of publishing high quality research.

·
·
·

Clarification: climate prediction is not whitelisted and has not been for close to a year due to them not publishing daily stats information (which is required for our superblock system to work correctly). They have stated they are working on this for the future.

·
·
·
·

I was wondering why they were not included.

·
·
·
·
·

Proposed TCD system does not require daily stats update. If accepted and implemented, CPdN can be white-listed again.

·
·
·

In a perfect world ..

"SCIENTIA HUMANA LIBERTAS"

boinc
Courtesy of @joshoeah

·
·
·

Sounds like it would be functional. This might also help with integration of other Distributed Network Communities (DNC) into Gridcoin. I envisage Gridcoin being the base for credit (computational power output) records. All other coins could be awarded based off of Gridcoin’s records.

Please submit the poll draft on Gridcoin-Tasks community repo, or corresponding CCT thread, if you have not already. I will create the poll when I get time. Of course I will verify if the reasoning is sufficient to vote.

·

Ok. Thanks!

·