On the Gridcoin Whitelist, and Where to Draw the Line
The whole community has been actively discussing the whitelist for a few months now, specifically focusing on what types of projects should be supported. While many valid points were fielded in support of all opinions, I would like to take the time to address a boundary that we seem to miss quite often - projects should have the potential to be useful.
But what does that mean, and how do we draw the line?
The delisting of Moo! Wrapper currently looks like it will pass, with around 60% of vote weight in favour of removal. I support this decision as detailed here, because Moo! Wrapper inherently does not do anything useful. The entire aim of the project is to decode a known message through brute force, meaning that by the end of it all (in an expected half a century) nothing will have been achieved. This project is, therefore, useless.
Parallels were quickly drawn to some of the following projects, with calls that they be axed alongside Moo! Wrapper. However, these projects can be useful:
Enigma@home works to decode Enigma messages sent by the Nazis during the Second World War. The project does so cleverly - using a hill climbing algorithm to reduce the number of decryption attempts needed. This stands in contrast to Moo! Wrapper, which blindly covers the entire key space. Further, the Enigma messages have historical significance, whereas the Moo! Wrapper message does not.
Collatz Conjecture - works to test the conjecture it is named after. This conjecture hypothesises that any number may be reduced to 1 by following two simple rules that lead to the next number in the sequence. If your number is even, half it. If your number is odd, multiply it by 3 and add 1. Either a proof or disproof of this conjecture has huge implications in the field of number theory.
SETI@home is looking for signals from outer space, hoping to pick up messages or activity from extra terrestrials. While it is true that to date they have not heard such a thing, a single finding by this project has colossal ramifications for everyone, so it makes sense to support the effort.
Based on the above examples, I think a concise way to check if a project should be on the whitelist (ignoring work availability) would be a sanity check:
- Does the project contribute to a field of research?
- Does the project have any cultural, historical or other quantifiable value?
If the answer to both of these is no, then it should remain unsupported. This is critical to prevent the spin-up of many useless projects with the aim of gaming the rewards system. I see no issue with any of you encrypting a random string, but let's not waste everyone else's time finding the plain text.
Imagine if they had thought Einstein@home wasn't a valid project, because it still hasn't found what it's looking for.
What would it look for if it was already found 😋
Great views from @dutch
Precis and to the point. Valid views and good explanations
I think each of the BOINC users have their own motivation to participate in one project or not, that’s why there is an ample variety of BOINC projects in the first place. There for the GRIDCOIN community will and shall never agree on rules how to whitelist projects on “scientific” merits of a project.
Some of the projects you list do not have any “scientific” value to me, but in some I am participating because they guaranty a steady workflow and/or have a nice community.
Out of the last two whitelist process we should draw other lessons, and establish simpler, basic rules on technical and ethical criterias:
SOURCEFINDER: Did not have the back-end to support the demand of work after being whitelisted, this distorted the “profitability” of the other projects and the community.
ODLK or ODLK: a complete disaster! Your self have written an article about double projects with the same aim. The title of the poll was:” Poll: Whitelist Poll: Odlk Round #2” and the correspond ending question: “ODLK now has an English page and is SSL. Are we ready to add it to our whitelist?” And quess what, which Project get’s whitelisted? ODLK1, the question was never ODLK1 it was ODLK. So a few have benefited heavily from a head start.
If we are not able to prevent such “technical and ethical” shortcomings first, how will be able to judge “scientific” merit of a project to white list? The GRIDCOIN community or at least some of its users are already optimizing their profits, whitelisting small projects with dubious or no track record at all, to get the first mover advantage, on the back of users supporting old, big projects, like SETI, PRIMEGRID, CLIMATEPREDICTION to name a few.
Using useful hill climbing algorithm (mixed with brute force method, BTW) doesn't make the project useful. Using brute force method is not a valid argument for de-listing either, if no better method is known.
I struggle to find usefulness in just being interesting from historical point of view. Also, there are thousands WWII mysteries nobody tries to solve due high cost and no usefulness.
While hidden message itself most likely has no historical significance, some may argue Moo! Wrapper project has historical significance in computer and encryption history.
Does (enigma@home) decoded message you describe here add anything significant to the knowledge already gathered in the books? No.
On the side note - a strong argument for why to de-list Moo! Wrapper I have described here
I don't understand where the issue is. Enigma being "solved" was one of the reasons often cited as a contributing factor to fall of the Nazis. However little significance it has to you personally.
Historical data -if aggregated- could become a valuable contribution to a research project -or a project of different nature- in the future. Knowing more about the past can't be a bad thing in my book.
Does it add anything?
Actually it does: It adds to the exact account of events that happened on that day- in this case a troop/ship movement and the way it's crew communicated their intention of reacting to a new situation.
Maybe I am missing your point.
I agree with this - there is historical value in the messages, which helps build a picture of events during the war.
If you look at it that way, every method is brute force. I consider a brute force decryption to be a full search of the key space, not a guided search that cleverly skips irrelevant parts.
I fully agree with this.
In the article you talked about usefulness and I did not agree with the argument. Being an interesting from historical point of view and useful is two different things.
It's not about personal. I'm arguing above about reasoning.
Can you explain how does it constitute usefulness?
Also I wrote:
Being an interesting from historical point of view and useful is two different things. I don't really stand against enigma, what did I show is that the same reasoning defends Moo!Wrapper too.
Can you explain what useful or at least historically significant is in the message @dutch wrote a very interesting article about? (Linked in my previous comment). While I find decoded message is historically interesting, I don't see why it is historically significant.
Are we still fighting Nazis?
Polish and later British mathematicians did great job. And they had no computers.
The lesson is - usefulness as a criterion for listing and de-listing projects is a very tricky one.
I haven't join gridcoin yet but I have a small comment. For cryptographers, it makes sense to have the actual cost of bruteforcing something, specially it gives you an idea of how fast/expensive you could run such a computation in a decentralized way. It is because people tackled the RSA challenge that we still believe RSA2048 to be usable.
That being said, it doesn't mean the project is useful or not.
Actually I think you just defined the utility.
Yes, I believe he did- and very good at that.
I would like to add though: However the original RSA challenge is over by now and this thing was kept alive artificially if I understood this debate correctly.
Correct. The original challenge was discontinued in 2007, and has effectively been replicated by a new entity.
While I see where you are coming from, the cost of brute forcing something can be readily calculated without even beginning to attempt the decryption.
You make a good point about decentralised compute power too, but that could be measured while put to use somewhere more useful.
Exactly I feel like they could make a better use of the computation power. I could make a project about running through all the private key space for the elliptic curve used by almost all the cryptocurrencies but there is not much scientific wisdom to get from that :-D
I really don't get how the collatz conjecture has any real use :/
It has immense significance in terms of digital security, which relies on prime factorisation. The Collatz Conjecture effectively explores in what ways the prime factorization of a affects the prime factorization of (a+1).
Well. I fail to see the merit of almost anything comparable to that. Maybe that's the point: I am not a mathematician -let alone a theoretical mathematician at that.
As far as I am concerned: this is an unsolved problem of the field of mathematics. Maybe it's the completionist in me, but I would like to solve it based just on that. :D
We should all thank SETI@Home for one very important development: the creation of BOINC.
Although SETI@Home may never find proof of extra terrestrial intelligence the project has contributed massively to the field of publicly distributed computing. The research being done is as much related to data science and computing technology as it is to astronomy.
There is often a great deal of value in the process of conducting research even if the immediate outcome seems fruitless.
Collateral research is a result of scientific inquiry. Searching for a better way to do A often results in methods to also do B, C, D, and R.
hey that's my line! xD
Great post. Disclosure: I did indeed vote to delist Moo! Wrapper. However, for those of you worried about an all-out witch hunt on other BOINC projects, let it henceforth be known:
I will staunchly support even the so-called "less-useful" projects like Collatz and SETI. Although pretty dang speculative, a positive result on one of these projects would have huge implications. Not only that, but negative results from these projects would be useful too -- not detecting E.T. despite decades-long searches would be an important empirical observation.
High-risk, high-return research gets little to no funding from governments and corporations. That is why Gridcoin and BOINC are needed to support these kinds of projects.
Great point. The ongoing running of these projects i meaningful even if they are never brought to a satisfying conclusion.
Seems you guys forgot bitcoin Utopia. it got delisted for not being a science project but it helps funding like h202 said "high risk high return little to no funding projects".
so where to draw the line?
I would argue that Bitcoin Utopia actually had more utility than Moo! Wrapper. Neither should be whitelisted though...
It is hard to believe there are people who want to delist SETI@home, which is the father of all BOINC. This was the project that BOINC was originally developed for. My own opinion is that the search for extraterrestrial life is an important endeavor, even if the probability of finding something is low. I see there are other opinions out there.
Quite frankly this is devolving into sort of a priesthood. It is almost a mob mentality on which projects should or shouldn't be part of the whitelist.
I think we need to go back to a simple standard where a project is whitelisted based on
Its ability to supply steady workunits
Process credits for submitted workunits ina timely manner
Remain up for computation/reporting of statistics for research payments.
Does not participate in ASIC-based mining of other cryptocurrencies.
We should vote on the standards for availability of workunits, credit granting latency, and uptime requirements for the project servers for statistics inquiries, and then implement an algorithm to make the decisions.
If all projects were explained as they are here the community could choose them better. I approve your choosing methods, thanks !