Gridcoin Technical Whitepaper ready for Peer Review
Over the last three months I worked several evenings on the Gridcoin Technical Whitepaper. I worked mostly alone, but the result is still a community work because I sourced many pieces from all around the web:
The whitepaper was compiled with MikTex 2.9, and the result is visible here:
https://github.com/gridcoin-community/Whitepaper/blob/master/gridcoin-research-whitepaper.pdf
@thecharlatan created the Whitepaper github repository:
https://github.com/gridcoin-community/Whitepaper
There were 100 commits until now.
I sourced the structure from this anonymous Google Document:
https://docs.google.com/document/d/1PUmLpL3jb-Dhf9-60MPpCYpCvBukUIWeKO8Pww0bhII/edit#heading=h.1gqiga6cbzja
I took several sections from the Gridcoin Wiki for the chapter Proof of Research:
http://wiki.Gridcoin.us/Proof-of-Research
The introduction is of @jringo. Mathematical details about BOINC Recent Average Credit (RAC) are taken from this steemit article by @jefpatat:
https://steemit.com/gridcoin/@jefpatat/the-math-behind-boinc-rac
The violet pictures were done by @joshoeah and @barton26 did a grammar revision of the whole text as my mother tongue is Italian, through later I added and reworked some sections.
If you have corrections or extensions, you can either checkout the github repository and do a pull request or ask to write permission for the github repository or you can simply write in the comments below. I will implement changes evening by evening.
I heard there is a bounty pending on this task, so if this whitepaper should get the bounty I will do an effort to divide the Gridcoins among contributors/reviewers based on how critical the contribution is. If you recognize sections done by yourself let me know.
today i increased bitcoin miner efficiency from 1W per gigahash/s to 0.5W per gigahash/s to make the estimate more inline with the official estimate and added a chapter on transaction efficiency as per @thecharlatan request
@brod pointed in another post at this whitepaper:
https://www.usenix.org/system/files/conference/woot17/woot17-paper-grothe.pdf
i am thinking on how to integrate it as it is well written. Fortunately the mentioned security issue was solved long ago.
@brod explained me in detail how PoS works and gave me this wiki page to know exactly what was done after the "Breaking and Fixing Gridcoin" whitepaper referenced in the previous comment: https://github.com/gridcoin/Gridcoin-Research/wiki/Stake-V8
today i added a reference to @brod's Gridcoin Kernel Stake v8 in the security chapter
Where is the resteem button? Wanted to reestem your post! Great post!
I'd be happy to do some proofreading -- not just for grammar but also for wording and conciseness. How much time are we setting aside for community review?
ciao caleb23, i would say 2-3 weeks from now on, would that be enough time for you?
Sure! What's the best way to send my comments? I'm not set up on Github, although if that's the best way I'm sure I can figure it out.
excuse me, i saw your post only now: i gave you write access to github (user @caleb23) github is the best way, join also teamgridcoin #whitepaper channel at https://teamgridcoin.signup.team/
Hold on. I got a little confused. I didn't have a github account to start with. I just created one (username gridcoin23). Can you give write access to that account? I plan on printing the whitepaper today and setting aside some time tonight and tomorrow to proofread. Should have suggested changes ready by then!
Just started some edits on a new branch. Hopefully those can be integrated OK.
ciao Caleb, i gave you write access to the github repository but you need to answer the invitation email
Ok awesome. I went ahead and committed my changes. I hope I did it correctly. I made some fairly substantial changes to wording in some places, so review and criticism are welcome!
Ok cool. I'll check it out.
i implemented following changes requested by @s4mmy: more page breaks and statement in chapter 4.2 about 51% attack in conflict with chapter 7.4
You've put awesome amount of work into this!
As for RAC formula, as I have already explained under @jefpatat ' s article,
Instead of an overcomplicated form:
decay = exp( (-t/7) x ln(2) )
you can use:
decay = (1/2)^(t/7)
as:
exp( (-t/th) x ln(2) ) = e^( (-t/th) x ln(2) ) = e^( ln(2^(-t/th)) ) = 2^(-t/th) = (1/2)^(t/th)
For some reason this non-simplified form of equation is copied all over Internet. Maybe because it looks more 'scientific' with exponents and logarithms than simple 1/2 to power t/th? The latter one is easier to understand and you can easy see that if t = 7 days RAC =1/2 or is halved.
thanks hotbit, i will look at it this evening and put the simplified version then
change implemented, hotbit can you check it?
I would completely remove eq. (1), as in my opinion it makes no sense here. In eq. (3) looks like only 1 is to power t/7 - some formatting needed, maybe (1/2) instead of 1/2 will help.
7 is 7 days, arbitrary chosen half life.
I have 15 nightshifts left in a row, so sadly little time to engage :(
hey hotbit, 15 nightshifts in a row! may i ask what is your job? i implemented the brackets and th instead of 7, however i left eq. (1) just to give the info that in literature this is often seen.
The work put into this draft is amazing, however I personally think there is still a lot more to be done including several important fundamental corrections and significant cuts to content. I have no intention of pushing for a vote on this paper.
ok, no problem, let's see what others think about it and how many corrections, extensions and cuts we implement.
Congratulations @dangermouse77! You have completed some achievement on Steemit and have been rewarded with new badge(s) :
Click on any badge to view your own Board of Honor on SteemitBoard.
For more information about SteemitBoard, click here
If you no longer want to receive notifications, reply to this comment with the word
STOPCongratulations @dangermouse77! You have completed the following achievement on Steemit and have been rewarded with new badge(s) :
Click on the badge to view your Board of Honor.
If you no longer want to receive notifications, reply to this comment with the word
STOPPower consumption note.
According to numbers in the whitepaper, setup i5 + gtx1060 delivers 4 895 598 GFLOPS. This would be an equivalent to 134 126 i5 processors and only 635 GPUs. As in this setup GTX delivers 98% GFLOPS, it might be more reasonable to assume that 98% of Gridcoin network GFLOPS is delivered by GPUs. Now the network would be equivalent to 1246 GTX 1060 cards and 5365 i5 processors (it comes surprisingly close to my guess-timations. With above assumptions energy usage would go substantially lower.
thank you hotbit i change it this evening, i will also reference your steemit article which is btw very interesting, maybe i'll source some sections :-)
change implemented, now gridcoin consumes much less! the reference to your steemit article is [21]