Moo Wrapper is a BOINC proxy for the distributed.net projects and is distributing work to participate in the RC5-72 challenge, a challenge ran by RSA Security from 1997 to 2007. The goal was to find the cryptographic key given a cipher text and Moo is currently working on solving an RC5 encrypted text with a 72-bit key using brute force. Do note that RSA have already ended the contest and this is a continuation of it run by distributed.net.

The project is whitelisted in Gridcoin as of right now but is up for delisting via community voting. This is why I vote for removing it.

While writing this post I noticed that @dutch posted pretty much the same thing in Should Moo! Wrapper be Removed from the Whitelist?. His post goes into more detail regarding RC5-72 so go read that first then come back here.

Disclaimer

It is very likely that I have misunderstood the point of the RC5 challenge. This post is also going to be aimed from my personal perspective, given my personal views of Gridcoin and what we should strive for. Also, when I write "Moo" I am referring to the RC5-72 subproject, not the entire Moo project itself. Given that the project is capable of solving other problems but just focus on the RC5 one, my statements still hold true.

Note that we have intentionally left Distributed.net OGR-project out of scope since you can already join that project at Yoyo@Home.

Project Outcome

In the Gridcoin Hangout #48 we discussed the poll and why the project should or should not be delisted. I tried to compare this to Bitcoin mining which I still think holds true. You are given the answer and try to find the question. At least with Bitcoin mining the problem solving would serve a purpose whereas with Moo we are merely trying to solve an arbitrary problem with no real outcome. I would be very surprised if we whitelisted a SHA256 solver.

One could argue that this is the same as the Enigma@Home project where the goal is to break the remaining Enigma machine ciphers. That is true, they both try to do the exact same thing. The difference is that cracking enigma codes holds historical value and the codes were not published for pointless bruteforcing.

Note that in the hangout I made a statement where we should only be including research projects. After giving it some more thought I have realized that this is a dumb requirement. Not only is it a stepping stone to endless discussions whether a project is research or not, but it also shuts the door for projects which do work with a usable outcome but is definitely not research. One example of this is BURP which is a video and image rendering farm.

Magnitude Distribution

In Gridcoin there is a set total magnitude to distribute among all the projects in each superblock. This magnitude pool is divided equally, regardless of the project's intent or participant size. This means that including one project reduces the rewards of all its sibling.

Now, if magnitude instead was based on crunching performance including small, or "toy" projects if you will, would not be an issue. Magnitude would go where crunchers go even if it's just to a project checking if today is Tuesday. However, given the limited magnitude pool I feel it does not make sense to allocate reward resources to a project which will not have a concrete, useful outcome.