On Mon, 01 Jan 2018 at 17:39:47 +0100 Anselm Lingnau <ans...@tuxcademy.org> wrote:
> Running the LPI on blockchain technology would presumably > only be useful if enough people can be incentivised to waste CPU cyles (and > boost their electricity bills) signing transactions on the blockchain such > that a suitably large “mining pool” can be guaranteed. I'm afraid there's a misunderstanding over what a blockchain is and how it works. A blockchain oes not itself require high computational resources and no mining, electronic currencies have both to serve two distinct needs of theirs: creating a decentalized ledger everyone can trust and no one can tamper on one hand, creating the currency itself on the other, since there is not a central issuer like a Bitcoin Central Bank. Bitcoin mining addresses the second issue, but LPI does not have anything to mine (knowledge and experience perhaps, but we do that on a cerebral infrastruture, not a digital one). > Bitcoin does that > by effectively giving successful blockchain signers free money, No, it does not: everyone can have the full ~150 GiB of the Bitcoin blockchain on their PC and keep it up to date, but this alone is not going to prouce a milionth of a bitcoin in a hundred years. Mining will, stochastically speaking, > but that > may not be an option for LPI. And operating a “public blockchain” only makes > sense if LPI (or some cabal of outside interests) doesn't control the > majority of CPU cycles in the mining pool because no one else can be > bothered. There is no need to "control the majority of CPU cycles in the mining pool", because LPI does not need any mining at all. Distributing the blockchain is enough, which only entails handling TCP/IP traffic and storage on the part of the participating parties. The only computing is performed when a new record is added to a block and when that block is verified and read, but that's no more computation intensive than GPG encrypting and signing/decrypting and verifying a 1 MiB file. > Finally, since all of the voting is going to be on LPI matters, > anyway, I personally have no issues with the LPI running an efficient > centralised voting infrastructure to facilitate this; on the contrary, I > probably would prefer that. This is where the blockchain technology matters: it enables having a distributed mechanism of creating and maintaining consensus just as strong as central infrastructures, but it avoids the issues related to a compromise in that centralised infrastructure. And it also relieves with the duties of maintaining a secure centralised infrastructure, which today most of the times means it llows organizations to avoid relying on a third party cloud services. > If you want to figure out how to do distributed Condorcet voting on the > Internet with – comparatively – very little resources, the Debian project > has been doing that literally for decades. It needs a GnuPG keyring for the > people who are entitled to vote, but we may want that anyway, on general > principles. (In comparison to random public-blockchain implementations, > GnuPG is very well-understood and widely established. We even require > people to know about it in LPIC-1.) This can be done on a blockchain, Ethereum is an example of a blockchain inrastructure that was designed to be able to keep track of an abstract concept of transaction, not just a monetary one. On Ethereum people and organizations can signa and entrust a generic contract, for instance. Alessandro -- Alessandro Selli http://alessandro.route-add.net VOIP SIP: dhatarat...@ekiga.net _______________________________________________ lpi-examdev mailing list lpi-examdev@lpi.org http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev
