!WMZPgpQpdMulhwFOdV:matrix.org

Ethereum Classic Improvement Proposals

65 Members
Discussions for Ethereum Classic Improvement Proposals4 Servers

Load older messages


Timestamp Message
15 Nov 2019
07:46:41@_discord_242032644232183808:t2bot.iosorpaas Everyone, can we get review on those two PRs in ECIP repo? I believe both of them are non-contentious at this moment:

https://github.com/ethereumclassic/ECIPs/pull/176
https://github.com/ethereumclassic/ECIPs/pull/184
07:55:41@_discord_557201004060213248:t2bot.iosoc1c I am not opposing this approach, and if the community believes this is the best, I'm happy to facilitate this process.
07:56:35@_discord_557201004060213248:t2bot.iosoc1c I would however, recommend to keep working on one document because the goal of all 5 competing ECIPs is the same: bring istanbul in whatever way to ETC
07:57:15@_discord_557201004060213248:t2bot.iosoc1c So what about having ECIP-1061 as base document and add 4-5 alternate specs to it, as ECIP-1061-{a,b,c,d,e...}
08:54:51@_discord_242032644232183808:t2bot.iosorpaas soc1c The whole issue is there because bobsummerwill and you have expressed that you strongly want to get #176 merged. Having separate ECIPs describing different options would be the only way forward for #176, IMO, because of its potentially controversial nature. Otherwise, if it's a non-goal to merge #176, then we do have other approaches for this meta hard fork ECIP.
09:17:32@_discord_403937835956174848:t2bot.iobobsummerwill I see no issue with merging all four of them, sorpaas.
09:18:55@_discord_403937835956174848:t2bot.iobobsummerwill And https://github.com/ethereumclassic/ECIPs/pull/183
09:19:35@_discord_403937835956174848:t2bot.iobobsummerwill This one still seems to be under review, eh?
https://github.com/ethereumclassic/ECIPs/pull/181
09:29:56@_discord_557201004060213248:t2bot.iosoc1c I have and will never express that I want anything strongly merged. My work is writing and proposing specifications.
09:30:10@_discord_557201004060213248:t2bot.iosoc1c There is a lot of talk about nothing in this channel.
09:31:04@_discord_557201004060213248:t2bot.iosoc1c If a proposal is too controversial to be even considered as a draft, this would mean we are not even allowed to propose it. if #176 is too controversial to be a draft, than we shouldn't even bother having a call
09:32:11@_discord_557201004060213248:t2bot.iosoc1c I'm not a process nazi, but denying the merge of a draft means actively censoring proposals by rejecting them before they are even proposed.
09:32:16@_discord_536270718980063233:t2bot.iowolf_li soc1c I was wondering why you thought that a sha3 asic would be very easy, whereas an ethash asic would be very hard. How do you come to this conclusion?
09:33:43@_discord_557201004060213248:t2bot.iosoc1c this is the ethereum's EIP policy by the way: merge anything that adheres to the process and basic formatting rules and let the governance happen elsewhere.
09:34:09@_discord_557201004060213248:t2bot.iosoc1cRedacted or Malformed Event
09:34:23@_discord_557201004060213248:t2bot.iosoc1c saying a PR is "controversial" moves the governance away from the public calls to the repository
09:34:49@_discord_557201004060213248:t2bot.iosoc1c and that should never happen under any circumstances because this will put the governance in the hands of a few editors
09:35:52@_discord_557201004060213248:t2bot.iosoc1c wolf_li I will not engage in an asic discussion with you as I will only be the loser for whatever I say. but to give you a short answer: you told me 🙂
09:36:30@_discord_557201004060213248:t2bot.iosoc1c you told me anything can be built and optimized with hardware. ethash is a monster of an algorithm compared with sha3.
09:36:42@_discord_536270718980063233:t2bot.iowolf_li that's true
09:37:11@_discord_536270718980063233:t2bot.iowolf_li btw is never about losing, it's a discovery process... did you see the sha-256 article we posted yesterday?
09:37:35@_discord_557201004060213248:t2bot.iosoc1c if you truely believe it's more difficult to design an sha3 asic, than I am happy to hear your thoughts. otherwise, I will try to avoid any discussion because this is outside of my expert domain.
09:42:58@_discord_242032644232183808:t2bot.iosorpaas soc1c No one's rejecting merging your PRs. We have said clearly yesterday that to proceed like what you said we'd just need to define the scope of your ECIP clearly, and avoid making it anything "official". This is the reason why we have https://github.com/ethereumclassic/ECIPs/pull/184
09:43:18@_discord_536270718980063233:t2bot.iowolf_li a PoW chip, any algorithm, has very unusual (for other chips) optimizations. Maybe you can think about this for a moment. We took a lot of time to document this for sha-256. dynamic circuits, power cascading, extremely low voltage, custom thermal packaging, and so on. Dozens of features. I'm quite proud of this article 🙂 https://medium.com/@Linzhi/history-of-bitcoin-mining-hardware-60be773e5f5d
09:44:50@_discord_536270718980063233:t2bot.iowolf_li a sha3 PoW (!) asic is much more "elitist" than an ethash/progpow/randomx asic, and more expensive. I add one more real-life example. In 2017 a Japanese company, GMO, thought that a sha-256 asic would be easy, coming from a non-PoW background. After having lost 200 mio USD (two hundred million USD), they finally gave up. https://siliconangle.com/2018/12/27/gmo-internet-quits-bitcoin-mining-business-following-321m-loss/
09:46:32@_discord_536270718980063233:t2bot.iowolf_li our ethash asic costs 4 mio USD, really cheap for asics. We are delayed, maybe we are not "elite" enough 🙂 but we are smart enough to not touch the "easy" sha3, that's for sure.
09:47:11@_discord_536270718980063233:t2bot.iowolf_li Encourage you to read the 'bitcoin history' article, I think it's great. These are the same reasons that make a (competitive) sha3 PoW chip very hard, very elitist.
09:48:29@_discord_242032644232183808:t2bot.iosorpaas bobsummerwill
> I will create a longer Rationale section in here based on prior discussions, on ETC Summit content from the individual talks and from the mining panel, and based on Discord / Twitter / WeChat / other debate around this ECIP, as that takes shape.
Would you mind to finish that? The ECIP is currently just one sentence
> "ASIC resistance" is a myth. The ETC ecosystem rejects this myth.
It's quite difficult to justify that it fulfills ECIP-1000 standards:
> Reasons for rejecting ECIPs include duplication of effort, disregard for formatting rules, being too unfocused or too broad, being technically unsound, not providing proper motivation or addressing backwards compatibility, or not in keeping with the Ethereum Classic philosophy.
At least, "providing proper motivation" would be a good first step.
12:23:32@_discord_242032644232183808:t2bot.iosorpaas multi-geth v1.9.5 is released! https://github.com/multi-geth/multi-geth/releases/tag/v1.9.5
13:45:01@_discord_349885060297785345:t2bot.ioyaz wolf_li just feedback but I don’t like reading articles with 0 references. If you’re claiming something is a “history of” then one would hope the things are referenced otherwise it reads more like an opinion piece. Op-eds are fine, but this isn’t marketed as an op-ed. I have no way of verifying any claims in that article if it’s not cited

There are no newer messages yet.


Back to Room List