All Core Devs - Execution #51

Transcript

[Music] [Music]

[Music] [Music] you

you okay let's get started hello everyone we're gonna be doing the core developer meeting number 51 we didn't make a big deal out of celebrating 50 last time so we'll celebrate 51 this time yay yay attack yes exactly we're gonna joke like that and then one day it'll happen and I'll blame you okay so the first item is testing so we're gonna have Dmitry talk about that but

before we do that I wanted to introduce Brian he's new to the testing team Brian if you want to give a quick introduction and just kind of explain what you're gonna be helping with yeah hi everybody so I'm Brian nice to meet you not quite a chord F I thought it'd be nice to have eating anyway I was recently hired to work full-time on making testing better that's a pretty like I'm also not sure if that means yet I guess ideas make it easy to generate tests easy to like use the tests I think a big piece of that yeah I've talked to Trinity a bit I have an idea of like what they want but I love Ben can reach out to me and tell me about the things that like they would like to see better see improving the past so I'm be easier for like you can't

be written to integrate with them cool thanks so much Brian and then we're gonna have Dimitri just go over if there's any testing news along with Martin because I know Martin's been working on hype so let's start with Dimitri so finishing Xcode Heche implementation along with I discourage that every Xcode hash test could be used to check X code size X code copy and call to such contracts so those deaths are not related to a constant you know

about every test that we have for exposure could be enhanced like that I received a lot of help actually on writing as confirmation again and Hyuga and it's mostly follows expert Josh was at core ideas that we had are covered now I finish incurring the same ideas that we still have asked in our list while still waiting for any contribution 4-bit shift figure and I'm waiting for hyper adults and in asking Martine herself has some maintenance well I'm

waiting for client results on high to see you there are any issues and we could we need to fix it or get maybe you might be some new tests after that well create two tests already a test or test ready Egypt s exist what I think we need a review on those constants it's been written by Y here half a year ago and exporters mostly all covered still adding a few of them like tests every day ok thanks so much Demetri Martin did you have anything updated for hive that

you want to discuss yes so yeah we're in the process of rewriting and it will also see some additional tests the peer to peer stuff that I mentioned previously regarding the tests so as least would get we updated all the tests I think about less than two weeks ago and all the tests pass and I kind of expect even though we still have hive I kind of expected client implementers to actually import the tests and run them

in their native setting once in a while but I'm hoping to get hive up and running as it should again within the next couple of days we've also had no new failures on the first testers that will be running continuously for the last I don't know at least a couple of weeks without them a need to reboot them so I'm very confident in the consensus of the clients right now at least get some parity yes thanks Martin okay we're gonna go through clients now we'll start

with Harmony etherium J there is no major updates or a mess and we've been mostly working on beacon chain implementation there's been a lot of recent changed in this back and it's not yet settled down so a lot of things to do catching up with those changes at the moment and we'll probably do that next couple of weeks that's it mamas okay parity we released - to - or a new version with quite a few improvements

some nice exciting stuff there I think so it improves sync performance a little bit I think big improvement is pairing stability we fixed a bunch of hearing issues over the past couple of months where it now finds a lot of peers but if there was so some fluctuation in the pair counts we put in the block propagation fix that has been talked about so once people more people are out in this version we can see if it starts having any effect on the network overall and yeah they're like minor burns I think one other mentionable thing is we made took the decision to move Jason our PC api's that are not yet accepted in an

AI key but that have been implemented as a proof of concept of an EIP or for some reason has made their way into the eighth personal or web three namespaces those are now disabled by default and you need to enable them by running with the flying json-rpc experimental so mmm I think good release lots of improvements cool thanks Fred guess nothing mostly trying to do database optimizations and yeah that's about it so boring just trying out ideas I'm looking at charts you see what

sticks okay turbo gasps hello so a couple of sides first of all in terms of development I have those rewind bugs that I found with Robson sync so it kind of sinks to Robson now so then I started to work on the test Suites so I just Martin mentioned that he would encourage people to run it once in a while so I've never actually tested them so I currently have some issues but I'm fixing them and I in order to make it work I had to modify both GB which might be interesting for some people that now I can run both DB in the in memory mode which means that

it doesn't have anything on disk something I wanted to do for a long time but just never got to it so now I have that option so the client is still not ready for Constantinople because there's still a create to revival was still not addressed but I will address it at some point and so in terms of the actual usage I found myself using the trooper gets quite a lot recently because I don't know it for other clients but it turned out that it's actually quite useful for extracting the data I would it I used for state grant proposal and so I am still using it quite a lot so I've got this little like as I said 300

gigs database in my extra noise as deal and I just created another copy of it and another one so I'm just kind of going around running all this data analysis things that was actually quite good you know that's it for me thank you a theory I'm Jay s okay do we have anyone from a less yes so we had some fixes and improvements network code nothing to mention in particular but we'll have probably today a new kind of tool that will work as a boot node so

it can be used similarly to get footnote and I hope you can put it somewhere on the minute and business as an alternative implementation of that that's all cool death areum okay Nimbus yeah so lately we've been focusing on on the parts of the VM which relate to developer experience debugging and tracing and such so if you'd like if you have any good ideas in that I have feel

free to mention them to us something we were working actively on right now that's a good time and also call attention since or testing edition of that all coordinates call there is the first it to point out testings of framework and proof of concept things are landing right now and there's a discussion going on Gators so I will post linked all quarters in case you're interested in that part it's a good time to sort of influence the direction of these two testing there we go great Pantheon I think I'm the only one on from Pantheon think we're planning on

a minor release next week Oh III haven't gone through the release notes yet mostly bug fixes it does include contributions from external contributors Anton submitted flags for the curling test net or the pretest net besides that release adrian was looking into some of the Lexi's statement prototypes inside Pantheon and our R&D team is also working on a be contained implementation great Trinity ok XT reham yeah this is from the manna project oh man oh sorry oh no problem

we're still making good progress on a number of fronts right now we're synching Constantinople blocks on Rob Stinson dissipating a full sync shortly great oh I know I'm missing a few who else am I missing I don't know swarm we're gonna start including swarm Adam who is working with the Geth and swarm teams a bit is going to start giving some swarm updates so thanks so much Adam and go ahead save a little back in the context swarm is a part of Gath it's in the same repository

with its own semantic versioning so the last guest release contains warm improvements such as access control and feeds which will be announced in the 4 version with the blog post which is coming with the next major release of Gath and an update for the last two weeks we introduced user perception tests and benchmarks this is 8 okay great thank you um I think Thomas is Thomas is usually here and I forgot which client he is does anyone remember okay then if there's no other clients

will go to awasum yeah this is Lane I don't know that we really have an update we we have milestones that we're working towards so the most recent one I think I think we hit milestone two we're working towards a milestone three you know more work going in like we released a proposal for eighth one X I guess that was kind of the main thrusts recently and just kind of trying to pull design together to get ready for that table will do anything else to add okay great research yeah a lot of our efforts have

been put in revisions of the phase zero which is the beacon chain specification for serenity although heavy revisions these are more in the readability organization you know polishing up data structures that kind of stuff rather than doing overhauls of like the core algorithm is going on so as people have mentioned it's been tough to keep up with these changes but again we're really honing in on what we would call like a first version of a really release candidate beyond that the phase one stuff is discharge data chain algorithms

and general data structures are kind of forming there I mean the general algorithms are known whereas they are being specified in a systematic way the beginnings of that and then in terms of the state execution account structures etc that's kind of where a lot of the thought is going right now and there are a handful of proposals on each research that metallic dose is that we're all kind of looking for feedback on right now beyond that I think with respect to Justin vdf Alliance there's just there's an increasing number of block chains that are seriously interested in incorporating PDFs in various ways which

is pretty validating at least from that perspective that's the the major front and the research team has begun to do beacon chain implementation within IBM great thanks for the update as far as vdf schoo I know Justin had a really good mainstage talk at Def Con about vdf that should still be on slides live and the other DEFCON 4 video should be released today or Monday I just got done uploading all of them and manually putting in every title so they're all in there and I just need to write up the blog post and get that going and get approval

for a few things so that should be hopefully up today I'll try to make that happen am I missing any other clients or research teams or anything like that okay we're gonna go to working group updates make it pretty quick but if you have anything to show about something that's happened in the magician's forms or a meetings that are coming up or ways to get more involved feel free and we start with Zach yes hello so we're putting together specs for a test net we've been working with One X and the simulation group I think this is

probably something that's going to be relevant for the test group as well so rather than taking up a lot of time here on the call I just like to say that we're going to schedule a separate call for today that I'll post in the issue that I opened so if anybody could join and provide some feedback we could start getting the specs for this test net and building it out and it's going to be powered by it the white block framework all right awesome do we have any updates on just who wants to go next maybe Alexi yes written something up for the state

could code state rent it's one of the things so the first version has been published for the person of the proposal has been published and since then it received a lot of feedback I'm really grateful for that and so the most interesting one was that Vitalik was actually suggesting to use create to up code with some other programming things to implement something which I was proposing to implement with linear cross cross contract storage and I started doing that and I have managed to implement the RC 20 contract based on those ideas and I haven't completed tested it yet but it does actually work in a way that it can transfer control so it can mint tokens and can transfer

tokens and that was an interesting to have a get a first feel of how would you use the create - and that is why I've created the well I've commented on issuing solidity and also created some thread on these magicians for people who are interested in looking into solidity code and try to even like on a prayer on their own branch to create some kind of new primitives to make it easier to work with create - because at the moment it's a bit of a pain because you have to like copy paste the bubble the bytecode of the contract in the source code anyway so um but I will still although so basically based on that the linear cost control storage were most likely going to be dropped from the proposal in a

next version and also priority queue and a step two will probably be removed but I haven't kind of done this second proposal yet so then but I will still be looking into linear storage elsewhere for example it might be quite useful for you as an integration so but at the moment ago that might basically must probably go out of the scope so now as I mentioned by Donna I have asked Pegasus team whether they would help me to do the proof of concept and the greets of Adrian Sutton has been working on that and he's created the the the very first for the side kind of moving through the proposal which was really good

I'm I'm reviewing it and I'm I hope to get my my prototype running as well so it's based on Pantheon at the moment so and then also I've done a lot of data analysis some of which was not presented in the said grant proposal so the latest one was I was trying to write a hit and run heuristics to identify all where C 20 tokens that exist in the state and so I was the usually successful token transfer so and it turns out that it was about seventy one thousand a-year C 20 tokens in existence and they take about thirty fifty three percent of the contract storage and actually

interesting proof chickadees is also air C 22 a contract I didn't realize that because it's got both a or C 20 and our C seven to one and considering that it's very important class of the contract so that's why I've done the sample implementation of it using the italics ideas and the next step would be to look into the on chain order books is I also from the italics post and research so I put some ideas about how to identify them i I'm going to do some analysis about token axis contracts which basically where all the tokens go in and go out so these might be very likely gonna be some kind of online sorry on

chain or the books and also there's another idea based in this analysis that we might be able to compress this state by kind of using some sort of generation so probably if you need to write down all this but essentially the I'm looking into the ways how we can have a very very compact in representation of the current state which was based on the concrete things that are for example like all the contracts that create so a gas talkin is creating could actually be generated very easily by using a very minimum amount of seed data so essentially I guess talking in this case will not take a lot of space in the snapshots when you try to sync it up and for this

particular that I'm not sure how well suited the other clients are for this data analysis but if somebody wants to look at the Google aetherium dataset like I haven't looked at it myself because I didn't have that time I apparently have but they have a good data set and is it possible to let's say to reproduce that the data analysis I have done before show which is some of it is in in the state grant proposal or and then these are C 20 analysis if somebody can try to do that and then if they get the kind of the hang of it so maybe there could it be a good way to sort of split up this work because obviously there's so much so many things

to do in this stream and I won't be able to do this all along so and that's why I I'm trying to enlist some people to do proof of concept together and analysis data analysis and maybe modifications and solidity invite or to make all these things easier so that's it for me great sounds like an awesome update next I don't know if there's much except for the magician's thread for the he wasn't proposal because it was pretty clear-cut before and he was a proposal update or working group brother not to my knowledge I think the proposal

represents the sort of snapshot of our work and understanding and there's also in fellowship with their magician's forum thread open which I don't think has been super active up to now but we could take the conversation there people have questions okay wonderful and then finally Peters working group on logs and stuff I don't even know if that has a thread or not or if it needs one did you have anything on that Peter yeah so Peters working group is Peter yeah so I do have an experimentation thread I a really long github gist about everything way I think it was about two weeks ago or

maybe like one week oh yeah I published it seven days ago anyway I don't have much to report on it a few things that I'm currently trying to do is that in the blog post I mentioned that in my opinion the two viable approaches that we need to somehow experiment we don't play with is either ipfs or big trend but honestly both of them kind of suck a bit and since then I was actually considering experimenting a bit with Felix's eonar work so essentially if anyone hears isn't familiar with it Felix is working on on a tiny extension to discovery protocol which is really elegant where notes can

advertise certain capabilities and we already have most of that most of that infrastructure in place and what I would like to do is play around with it for notes start advertising for example that they are light server or they are full node with certain data sets available and if we can start advertising those kinds of things then I'm trying to figure out whether we could we could do the whole historical state retrieval in protocol without having to rely on without having to actually hack another word decentralized protocol for it I'm not entirely sure it's possible but just wanted to share that this is a third by direction which in my opinion is it's a valuable direction and we're going to

need it anyway for light server so might as well try to see is try to push it a bit and see where it goes well no so what discovery v5 it depends on how you define it so the way current light servers work that discovery v5 was complicit Bend because it was just a horrible hack so that that is not it's a different thing but I guess Felix wants this eonar to become the new discovery be five so in from that perspective you yes you could call it a v5 so the top memory was that it was just kind of hacked into the protocol but it wasn't really scalable or working properly and then this was a

really clear DNR approach is kind of clear felix has a few e IP is open so it so there are quite a lot of material already on it published on the on the yaki repo yes answered we're discussing in serenity to use the you know whatever the mature version of discovery b5 is for advertising the share cards that you're a part of yeah so exactly that's that's something that I I would also like to to experiment with putting the historical state announcements or just telling the world I'm sort of a state have on top of that exact same mechanism can you put a few links in perhaps yes I

will think of the eyepiece and post them appreciated so I guess that's mostly the update from my part I mean I can't see at least for now it's an exact solution that is definitely going to work so I'm probably going to sit on it awhile until I figure something out okay great Greg you're raising your hand sure I tweeted it out I was supposed to get on a fellowship of ethereal magicians but I

will paste it on the cord of calls in one second I just pasted it there I'm pasting it here to face it it everywhere thanks Peter hey Shawn we just went through the working group updates is there anything on your side that you wanted to discuss we just kind of went through everyone every working group and Zach talked about the meeting he's gonna have on simulation in a couple hours or

an hour and a half nothing for me to add Thanks great okay that's all the working groups Constantinople hard Fork is next on the agenda I don't think there's much to talk about except for two things stir be and what off reposted in the agenda about the possible block numbers and calculated based on time now that we have block time now we know more about block x so let's talk about that first with offeree I think the gist of it is we're still fine and there's a plus minus two day variant right or variance yeah exactly

so last time we decided to talk about a block number this meeting and yeah I just put down some numbers and if we go for like middle of the week or Wednesday I did some calculations how much it could marry under different scenarios and if we decide on a block number today it most likely be one a week day cool so we need to agree on a block time today still right I forgot if we did that yes good okay if we agree on seven zero

eight zero zero four to the hard fork will happen between Monday and Friday cool who wants to suggest a block time or did you suggest seven zero eight zero zero four two can we do seven zero eight zero seven zero seven zero eight zero eight zero seven let's make it Ballinger you know they are I don't care let's see how much would that purse down

is that okay our free yeah that's tolerance yeah that's out half an hour or one hour or something Oh at least forty two is also fine with me so honestly I don't care so what does everyone else think we can go coin one nope I support the palindrome made - has there been a pattern in the past like have we used even numbers in the past I just I just don't want to like if it's always been like zero zero zero zero zero like suddenly change now well appreciated even in the past yeah I think we I don't

think it's always been zero zero zero if I if I'm remembering correctly we can what is the chance of somebody's trying like they did it and what is it was Rob's turn I think somebody has to manage to to push it closer on Rob's turn when we agreed in a block number in the end it wasn't we wanted is it possible to do it in the main net but more expensive okay so we reasonably happy that it's not going to be like somebody's gonna start super mining in and just for that purpose and I'm just gonna push it one month in front or

something or one week ahead of what we planned that's all because of the high difficulty of the magnet but what could happen is said it's like a first case scenarios if the price of Issa continuous dropping very fast that a lot of - turn turn off the mining works that could slow us down but yeah we have top of it I don't know I mean we have to agree at some point - for block number and I think it makes sense I think it's it's very stable yes

so it's then tiny exposed to the link so yeah I don't think it's a concern yes so just to make this clear because we're thinking of putting in the block number into the next release possibly with a switch to say so you can delay it from on the command line but I just want to make sure so we're we're not planning on changing the block number just because it will fluctuate on a weekend right the only reason to change the block number would be to find another consensus right yes yes yeah that's my understanding as well and yeah all the previous ones have been triple zero but I mean we don't

have to keep with that unless someone thinks that we should well it makes it easier to remember when you because I'm sure there's a lot of people who remember all the numbers by heart I don't know what's the use of it but you know so far I managed to remember most of it okay is anyone there is anyone against having I guess quad zeros at the end or we still pro palindrome palindrome is cool but keeping with the zeroes is good - would we commit palindromes going forward or this view

went off good question yeah palindromes can be a bit hard so currently it's kind of easy because it's the number is a bit boring but maybe I could have imagined it being a bit harder to pull off the palindrome yeah that'd be hard to do in the future and people are kind of probably a little bit like OCD about this stuff wait I guess as long as we have we rolled our block numbers are only in the millions parind rooms are fairly easy because it's still just maximum 1000 blocks or yeah maximum 1000 block drift so that's

that's still within a day if we enter the six digits then the drift might be ten thousand locks which is two days oh yeah so boring only three numbers to remember for a palindrome - okay we're starting we're starting to get to the zeroes being what most people want it sounds like I'll keep the rinkeby that's not working interesting I second it I

like that idea I like that so seven zero eight zero zero zero zero it sounds like that's what we've come to and it looks like the troll box is liking for zeros as well not that they matter sorry troll box okay so we'll just go ahead and do that unless there's any one super opposed which yeah and then stir B is there any updates on that I think it's just been going smoothly but I'm not sure I think generally I think it was quite useful I think not just

because we we kind of transition that of course somebody tricked us in terms of for mining so I guess for me the transition happened during the night when I was asleep which was unfortunate because I wanted to see it but I think Martin has discovered some of the kind of conversion to bugs or something like that for Genesis was there any kind of problem with the other clients we discovered something we or - Martin did you discover something weird not this time around no in the in the previous Robson screw-up where there were some weird things happening and I discovered a new

kind of attack which I don't really have time to go into in this call but I'll try to make it good right up and post it somewhere so you discovered it when we were during the Derby no that was during brokering I haven't really seen anything particularly interesting in Derby but there's some working guests but I'm not sure okay yeah so I mean of course the the fail of Derby was that we were felt filled prey of that miner who was GPU mining Derby and because the spec was publicly known so we could not control that at all so it turned out to be but I think the plan was to sort of have kind of a bit more mining during

that transition but I don't think it happened so I don't know I'm not gonna push for another rerun of it with the secret specification but I think I'm still I'd like to think that was useful yeah and at some point we maybe should spin up new test nuts there's been a suggestion for the name Gangnam a new perk for work last night which can have a bit more in Genesis allocation so we can also force it and stuff and then defecate Robson what for well where adapts can play

i I think this so yeah maybe this is an interesting time to bring up some discussions that have been floating around here and there that we kind of agreed that Rob stone is kind of too big of a monster to continue but an interesting question is what do we want to relaunch of proof-of-work that's not or whether we want to launch proof of authority because every and company has been working quite quite a lot on the girli test net and that might also be a promising alternative and of course I do acknowledge that that would mean that we don't have a proof-of-work test not a public proof of work test net

but we could always do so to say public proof of work test that Forks just to verify the proof of works and difficulty is - ragini to the hard fork so there's there's not really too much point to run a proof of work replacement for upstone but it's just my opinion we would we would drop off Trinity I love all of those I haven't yet implemented actually but I guess they can do that with relative ease working progress so this

can be done and with all the money in developers out there you can do this together if we want us and I agree with Peter that going forward was proof of work not feasible and just as lonely GPU miner on Derby just showed how unfeasible chestnuts are the end especially when you talk about having a test net for applications not for consensus needs I think Paul and Mikhail had a comment yeah I was trying to say that fresh test networks in not good in terms of testing difficult difficult to bomb delays because they

have to low block numbers and difficulty bomb was not yet activated on those numbers that was on thousand Sterba yep I would suggest you guys during this call because we can create a femoral test nets as well that are purpose-built with relative ease that can be provisioned and controlled yeah so essentially that was my suggestion to you that whenever we want to do a hard fork we can create a purpose-built test not to specifically test the difficulty adjustment or or something like that but

that's the only thing that we kind of need a proof-of-work that's not for for everything else proof Authority is possibly a saner approach at least from an app developer perspective yeah you're probably gonna want different consensus algorithms within different test nets for different reasons but you would want them to be purpose-built of course but the issue is provisioning them and ensuring adequate activity within the network I think that's something that we can address and that's what that's what we'll discuss on this yeah but my point really is that if we want the purpose-built test not to

test difficult the adjustment then we don't need any activity I mean empty blocks are perfectly fine whereas if you want to test any other modifications then a proof of authority test that is perfect for that because that can actually test everything else and we could probably configure at the configuration for difficulty Bom Bom and for delays I mean to actually test them to make them enabled anyway we definitely don't need to make a decision about this here and it's it's more like just a suggestion that if we

want to replace Rob's done it might be a good idea to maybe replace it liar proof of authority that's not but everyone please think about it and if somebody has a really good such good really good reason why that's a horrible idea I think everyone can agree do you consider it yeah I think that activity and behavior like background traffic is going to affect consensus and various metrics within the network Miami that's the kind of the point of often having a public that's not because then you have everything all the traffic or whatnot so

I suggest that we go to you know explore the option of purpose-built POV Network powa Network exact yes and then we come back maybe on the next go and just see whether this is feasible to just have P proof-of-work Network for specific things and then as you said feature and proof of authority for other things for most of other things you know so these purpose-built test nuts can be ephemeral so rather than being entirely public they'll be permissioned and then the activity and the behavior at all that can be provisioned and automated with the tools we've developed awesome yeah this is a good topic for

the next meeting for sure so I biked oh I had one other question about Gourley offeree I think there was something more aetherium classic was going to be compatible with it would that make it bad to be a public aetherium tests net if it also has a theory and classic on there or is that compatibility not a problem no what will not have classic compatibility but we work on something similar for the classic community that's what it is okay so it's actually not the same test that got it yeah okay anything else on Constantinople we want to talk about it

sounds like we got a block number down and that's great anything else okay next up openness for meetings I sent out a tweet that I know not everyone's on Twitter basically there were enough people who said over my dead body when it came to having any type of closed meetings that we're just gonna have open meetings for now because there wasn't as strong of a inclination to have the closed meetings at least on here I don't know about in-person meetings yet just because the techno the technical feasibility of having those as open as possible is different but as far as these meetings go and because we're not

going to really have any more what we're previously called a theory m1x or aetherium main that improvement calls we don't really have any more of those the working groups are created now that I think that just us doing the same thing we've been doing is gonna be just fine if anyone has a differing opinion you can speak now or you can speak to me privately if you don't want to speak now since this is being streamed and everything just a bit of confusion I mean there are working groups and we will talk within working groups is that like frowned upon to talk work oh no yeah that should be private if you want

it to be private it can be as private or public as you want because that's not a core dev meeting and that's something that your people are coming up with on their own so there was some confusion around that on Twitter but I think it was just people not understanding how these meetings work so did that answer your question yeah I get though okay Greg what's up Oh Greg you're out you're muted there was useful discussion on on how to make the judgment of you know where do you open it up and how much and it's a matter of judgment what can you

do yeah you know I just threw in chat that you know we worry about speaking freely in a public forum we worry about the press and what they might say and when I was organizing anti-war demonstrations we had FBI infiltrators we had CIA infiltrators we had you know we had oppressed and would accuse us of violence after the police had finished beating us up so you know I just sort of shrug and go this is important but really we don't have that much on the line so it's just isn't that hard okay

thanks for the comment on there Greg cool that you were in anti-war demonstrations so yeah anyone else have comments yeah well no actually not but there was the thing that I thought was not on the agenda and I think maybe we should have oh you're right sorry I didn't add that on there go right ahead yeah so yeah update well from last two weeks there is one one issue left that was reported I think at the beginning of this week that has to be

I resolved we know the solution is quite decided how to do it but there are some small changes to the implementation required to mitigate some a small flow in in the in the algorithm and so that's mostly that's mostly the average I think yeah I don't know if there's anything else to say unless you have some questions or maybe I think that seems obvious to me but it's not yeah so a couple of weeks ago there was so there was implementations in cpp in

it hash in verifies for cash integrity and they wrote and sync and then the spec changed bit and guess and CBP updated and now there are and I think as far as I know apparently heard you have not yet implemented the latest changes and oh yeah that's correct couple of some more discussion about attention to changing this specification so yeah right now we have two clients kind of in saying but we don't know if there might be hurt [Music]

okay sounds good Paul urine needed because yeah I might might as well take it on the public channel so we've had some bit of discussions with various people in crypto communities one of them is David Warwick the creator of SIA and also one has studied a lot of assets and his take on the whole frog power and I'm quoting a bit I gotten his approval to do so but I might not you know I might

make some mistake but my takeaway from what he says is that from how if we were to replace hash motifs at hash with Howie - that would probably you know keep Essex bias about a year a year and a half foreign asacs to take any cannonade threshold he does have a point that if we increase the resistance towards I guess ethics then that could itself be a centralization kind of something which increases and rotations

because only a really advanced assic manufacturer would be able to manage to spend the resources and the researched and build such an asset which could gain get efficiency gains so his take is that although this would bias by the year or two another approach would be maybe to use switch an algorithm which is extremely as a Kremlin and that is still an option that we would have so one speculation that one might do is that we could adopt Prague how now and if from a

year from now we still don't have Google State we might consider switching to something which is extremely acid friendly and which any acid manufacturer can produce no matter if they're large or small just to remove this kind of centralization regarding the actual prevalence of asacs on the network today it's really hard to find any numbers I mean there have been numbers reported from various sources but it's none of them can be taken as fact because there are no official records of the number of

units sold or the number of units manufactured so it's really hard this is hard to take that into account those kinds of numbers yeah and and some people think that the fact that we're incrementing power that's it's already decided and that's not the case we're hoping just to do the implementations and get the options out there on the table and then make decision later on all right I think that's about it from

me can I say something so I think I was I started to think about Pro power a bit more about because of two reasons first of all is because we started to do this proposals for improvements of East 100 and obviously one of the elephant in the room was that okay so what are the other things that are slated for for the theorem 1.0 and then obviously there's profile which will take some amount of work and then we'll definitely be taking some resources away from other things and secondly the Justin Drake's interview for the century Bitcoin or whatever it's called Center right now so when he is saying

that the reason why Foundation and the others wants to develop the PDF ASIC is because they want to basically develop as best as they can do so that another potential person up another potential they cannot get any like super advantage by by creating something much much better because it's simply not possible to create something much much better so and I'm really glad that Martin Lee mentioned that as well that there is a downside so there potentially there there is there is another side of the coin so I I wanted to be able to detect people to explore both of them and just don't be an eccentric and one thing yeah thank you

yeah thanks for the thorough explanation Martin and just to make it clear for those listening like Martin said Prague POW has not been decided is something we are definitely putting in or doing same with any of the working groups we discussed today besides the simulation one and probably the II was someone just because those are non-controversial things that are more than likely gonna go in state rent and any kind of log related working group that Peters doing it's not final yet as far as a decision being made to pursue that as part of kind of the one point of roadmap just for per se I have a quick question

Hudson yeah sure just tying together two threads we spoke about launching a new Testament earlier and whether it should be proof of authority or proof of work and Pavel and Martin just mentioned that they may want to may be ready to spin up a proof-of-work test net using Prague how soon this may be way out there but would it make any sense to for that new test net that new app test net to be a prompt our proof of work test net actually I don't think so because you know the switch from hash motive pro quo is really just the envelope of the blocks and it doesn't matter as far as I

can see whether the blocks are fuller so I would say not really yeah moreover we want to actually have a Testament that has a suite from one algorithm to the other in the middle so it's not pure progbar it will be like if half first and then prog file or maybe even like switching back and forth many times on the way if if client configurations would allow that so it's I think it's bad idea to mix like tested for doctors think with this one actually then it

might be an idea to have a first tested with losses I've got five scrap that have another doesn't with launch that's at the second epoch can start and try different configurations like that with the serious distance that makes sense thank you great all right anything else on that okay I have a quick announcement myself Lane Danny Piper an offer II have been getting a group together a an internal group that is bringing together people who are interested and coordination and project

management related roles in the etherium ecosystem some of them are going to be coordinating hard Forks with people like myself and offeree some of them are going to be helping out maybe with AI peas and some of them are going to be helping out with the core dev calls for weeks that I can't be there and Lane can't be there and stuff like that so hopefully we'll be introducing those people over the next few months mostly starting in January which is maybe when the next coordinate of call will be anyway but that you're going to start seeing more of that and the main goal of that is to number one lighten the load for myself and Lane and other an offering and others who are

a lot of the coordination around our specific teams and our specific initiatives like the cornet calls and another one is to train up some of these people who are really interested in volunteering or eventually getting hired by teams so teams that need a project manager can kind of be introduced to these people and maybe reach out to them too so if they want a job so I think it's a good initiative and I think it's hopefully gonna be really successful and we're gonna work really hard to get them more integrated and to these calls and do it in a way that supports the core developers and doesn't apply any more pressure the other main goal is to kind

of see what the needs are for these calls and four core dev coordination in general and make it as easy as possible and as stress-free as possible all right the date for the next meeting the holidays are coming up in a couple of weeks it'll be the 21st and stuff like Christmas is on the 25th I don't think there's gonna be anything but new releases for the fork number between now and then so unless something radical happens do we want to just skip that because people will be on vacation and stuff yes please yeah I might turn the question around

and say would anyone be opposed to January 4th being the next call that's a month from now I think that's reasonable okay great in that case yeah we'll just do January 4th and yeah everyone enjoy their break and yeah if there's anything else have it in the coordinates Channel have it on the working group chats and The Magicians and we're gonna give back about half an hour today so thanks everybody have a great holidays you

[Music] [Music]