All Core Devs - Execution #58

Transcript

[Music] [Music]

[Music] [Music] you [Music]

all right good morning everybody welcome to Cornell's meeting number 58 I just updated the agenda like a second ago with Martin's EIP and I think that's it so go ahead and refresh on the agenda if you haven't already let's start with the roadmap I don't think there's any updates on that but I did add a link to the etherium wiki so you all can keep up with some of the updates and proposals and the timelines so we can skip over that unless anyone has comment sorry Hudson I thought the roadmap was probably discussion about how often we

want to do releases or we or maybe we shouldn't have a discussion at here but just get some if there was any discussion happening outside of it if somebody can bring up the like some highlights of discussion ah got it so so what we had before was what offer he came up with as far as the timeline goes that we all agreed to at the time but I'm happy to change it with suggestions right now what we have is May 17th a hard deadline to accept proposals such as a IPS and then we have a soft deadline of the 19th of August no wait July to have major client implementations it's under the timelines

section on the link on there what were your or your thoughts on that Aleksey so I guess there is a discussion going on at the moment and magicians for example and it looks like people were talking about so that the discussion is about a smaller and I guess more frequent smartheart Forks whether it is achievable or not with a with our sort of setup and resources and I think one of the crucial bit about it is the the testing is that at the moment the testing seems to be a bigger bottleneck because the testing kind of in most cases is something which happens after the IP has been accepted and included into the hard fork and that means that if we keep doing things this way we

would not be able to do smaller hard Forks because for the smaller hard Forks it there's needs to be some kind of pipeline going where there's multiple kind of hard works in the in the process of being prepared and that includes the testing so you know I would like to get some comments maybe from Martin because he recently responded in that thread the hard folks we have done historically are very intensive on the test section because the eeap's that are going in they can have into you know related effects so as we change something and

finalize stuff we might have to redo earlier test cases regenerate a lot of test cases and this is yeah so I'm thinking that it might be easier and more spread-out work if the hard forks are smaller and more spread out that means it's not that big of a catch effect on their testers and then I don't have to meet you agrees though it's me tree here yeah I actually agree with a smaller heart for either the question for you Dmitri and Marcin is that so the in terms of the time is spent on preparing a test where is the biggest

time span is it the human labor or is the test regeneration or what is the kind of the biggest component that we would need to kind of paralyzed and push it onto each sort of each parallel line of the our pipeline processing thing I think the biggest time is to actually write down the best case in a JSON format and sometimes it's there's about hundreds of test cases and that would be good if those who propose EAP also propose some test scenarios so I could just so we could have process of converting those scenarios into JSON

test case with many people's and just me and what would be the ideal situation for you would do be that the VIP is been proposed come with the generated test is it feasible for people to actually generate this test themselves do you think in order to do that I will have to teach them how to write actual JSON test and how to generate the test it might be maybe that's what we should do just kind of because if our ultimate goal is to kind of get the more people involved we just start doing some kind of workshops maybe or online workshops and teaching people how to write a test so - totally agree so I think it's unfeasible to have

one person so means we currently have a truck factor of one if Demetri gets run over by a truck we don't have any other knowledge within the system of how to do testing so I think a really good takeaway is let's rally some people to have more humans working on testing actually that's a that's a thing which is already starting to happen and I as far as I know it's mainly on the Python wing under Piper Miriam there's some people on Uche and yeah a few more guys trying to improve the tooling specifically wrong testing I do think that better tooling would make a difference here we did a lot of work

but he wasn't minutes it's tricky there's been there's been a parallel thread in the fellowship forum about just generally setting higher standards for your IP authors and I do think that like putting some of burden on authors or submitters to like include tests or even just the types of tests they want would be a good thing I think better tooling could make that easier yeah one thing though there there's quite big differences between different oops I mean if it's simple shift operations those test cases can be defined pretty well or if it's a gas repricing there's not much those test cases are very quick to write but it was something more

intricate like the s s story pricing where we have this other extra weird edge cases where if you enter back in the same thing and you do it three times in a row and munis and half mode those are really complicated to write and they take a lot of time to write but it's this big variance but I suggest for these particular cases to still basically do it iteratively but with the iteration consisting of the proposer is for writing the first set of tests and then these tests get reviewed by let's say the testing team and then the work comes back to the proposer to just keep improving them or a rather than

basically just offloading this whole work so the testing team would basically just kind of look at the tests and say well I think this should be more of here this kind of test and so using their experience to just guide the process of creation of more tests yes a proposer should definitely be responsible for more testing I think that the way we do that is to change a IP one to make it a requirement to have some test cases for Cori IPS the IP one needs a complete overhaul as it is so maybe setting a deadline for VIP one changes over the next month or whatever and then get a lot of people working on a single PR for

it would maybe be the most beneficial but I'm ready to hear other opinion also could be handled in two three three which is the hard fork meta Alex gave me some feedback on it and it it should be ready to go ideally later today okay great that's something we can also discuss next meeting once it's covered once it's done and then accepted like are there one or two other important like changes that really need to be made in the IP one just so we know what's on your mind yeah there's also a kind of a discussion that needs to happen on how something gets to rejected if we even have a rejected

State or if we just have it withdrawn state what it means for something to be accepted when it comes to an ER see and then eventually maybe setting up kind of the framework to make ERC's into a completely different repos since they're becoming so different from other a IPS a lot of the time there's redundancy with the EIP one template and when there's not redundancy there's just things that don't need to happen at all so it makes it a little bit more confusing for people to write ERC's i i've been told so that's that's kind of my thoughts it strikes me this is important but also like not super technical so i wonder if

this isn't something that one of the cat herders might sort of be comfortable taking ownership of over the next couple of weeks yeah starting the thread and kind of managing it that would be great so teams already working on this we've already talked about it so yes it's happening great okay well we'll we'll communicate and the and the chat like i'm telegram or dinner or something and keep up with that i have that as a note here and i'm there's already a note taker today but i'm taking personal notes for basically to dues for next time and then we will come up with like maybe in the notes they'll be like a you know what's been accomplished what needs to be accomplished to dues for next time kind

of a thing so we can keep up with that that one suggestion from Gregg thank you Greg I have one more question about roadmap before you move on I don't if anyone else has any open questions in this topic you could go ahead the question is just so this has come up a few times now in conversation this May 17 dad line I'm just reading here it says hard deadlines to accept proposals for Istanbul did it like is there any requirement on how mature those GIPS are at that point in time like could someone hypothetically like just file a placeholder AIP it'd be good to have some consensus on this good question I don't have the answer to that well I mean I can give my perspective I didn't give my perspective on this because I sort of you know when we did

the workshop in Sanford earlier this year I'd been asking everybody like what gives you specify by by mainly 17th and I was kind of anxious about myself being able to specify something which I think it might be not very likely that I would be able to specify everything but then I kind of really decided to relax about it because I think if we rather introduce the sort of the freak more frequent hard works pipeline things so that people will be more inclined to actually do proper work on a piece before proposing them and not trying to rush it in and just to make sure that they get like to get someplace holdren because i don't think it's a good idea to just put the placeholder so I think the eyepiece

needs to be worked out pretty well before and and then the problem with this deadline that now with the plan of having our hard works every nine months and these deadlines is that now obviously if I miss the deadline I have to wait another year or something to pour it something to happen so but if I know it's only gonna be three months and I say okay I'm gonna take another three months to make sure my app is properly defined everything is done tests generated and then I will go in and definitely will make a big impact think another piece of this is it depends on what you mean by placeholder merging an EEP to get it into draft mode that just fits the

template is not very hard and people are welcome to leave text saying you know I'm still working on this so the bar of getting your EEP merged should be a hard format well here I don't I'm not sure about what to go into the specifics of ERP process because I'm not very well I know I haven't thought it very well but for me it's not really like yeah it's something which you propose to be definitely proposed to be included into the hard work whether it's a IP or like fully specified or I don't know how what is it so one of the reasons I published

my statement proposed was not as an IP but free IP which I just want to emphasize that I haven't got to the stage with the IP yet but I still want to conversation going and yeah IP or not I mean it doesn't really matter how I call it but I don't know so it's it it does unless we're now proposing that we don't use the e process at all and we don't need to have an excited I eat number and we don't need to track process then if I'm gonna drop fully-formed things in without using the e process no it's fine to have it proposed like written and proposed and merged and that's that that's fine but the eeap's that are kind of slated for the inclusion into the heart worksheet

have a certain and very high requirements in terms of readiness so we can create hundreds of each and then get the merged but it doesn't mean that we're gonna find them in for the hard floors absolutely all battling to e2 3/3 which basically has a proposed session so that people know that they're signalling that they're gonna try and get this EEP in but it has to move through other sections and that may include tests and everything else to actually get in but I think that the sooner you get in a merge deep that isn't just filled with to dues that's a useful step to signal that you're serious about engaged with the process so I would I would propose that we changed the process insofar as not have it hard for

eccentric but have it EEP centric so that say the first step is that they need get accepted and the second step is that it's implemented in just parity and the third step is that there are tests that fully cover it and everyone's helped with those tests and the clients pass it and they step up with that is okay it passes all these three or four or five steps so when can we include it okay a month from now and the client great idea totally in favor of more pipelining I think kind of drafted a process and I'm trying to propose that I'm from Canada so I'm doing it from that perspective what we can do is light

curve we can suggest EEP order to send a man open an issue each saying that each should be considered for next hard food for this hard work if we are going for Istanbul we can do that and I have created a tracker at github of ezh where we are listing all the eats it's it states that these eats are at proposed status and then it goes for the then we bring it to the coda discussion if it is accepted we will change the status and and eventually we will take it to client implementation test nets and finally included in there and I'm also referring to the tracker

that Boris have created about the client implementation we can go ahead and do that so I'm just yeah please sorry so what I my alternative is that would be to not focus on like xanthium is next one and which one are much enough but to instead focus on one by one like oh this week it looks like this EEP is now finished it's implemented everywhere this has tests for everything it is well written everything so we'll just schedule that EEP for one month from that time yeah I think that I like poojas plan if it goes with the

current status quo of having like unnamed hard fork many months in advance if we do these smaller hard Forks we would have to adjust it so well we'll just look at all the different I think we should just go to etham additions and look at all of the different proposals we have here and come to consensus on undoing this I'm hearing a lot of consensus around doing smaller hard Forks though and I think that's a great idea personally so we should get to talking on that and we can probably get that set up pretty soon especially with some of the e IPS that we know are maybe gonna go in to another two and you know the next hard fork that have tests done and things like that

any other comments yes one quick one before we move on from this we haven't really concluded like what we meant by the May 17th deadline and what's the state of the Eve's who won there are we expecting each to be in a state where clients can implement them by then to already have implementations yeah we might throw that whole deadline out the window if we change to smaller hard Forks I think is the answer if someone has a different answer feel free yeah yeah I think this is what I kind of implied but you know if we do switch to a different model than we'd and this is what I meant when I said that I stopped I kind of relaxed about said my may 17 hour deadline because I want to work

towards the quality of proposal rather than to the deadline yeah doing smaller hard Forks does that I believe yeah I very strongly agree with Martin and I like to say here that smaller hard Forks in the nation with a more AIP centric process I think it's pretty clear would produce higher quality VIPs I'm gonna move that we table this conversation and take it to the fellowship of athira magicians forum and like work out some more concrete proposals there and not take up more awkward that's bandwidth on this right now sounds good what's next we have some AI peas we can start talking about is Vitalik here cuz the next one is actually one of his let me check I don't think so if he comes later we'll just have him go

back to this but the one that I included here was the IP 1599 if the market change for the eath 1.0 chain there's a link to that if people want to check it out the next one is one that I think Danno did which is early which is the etherium network upgrade windows and i think that's kind of related to what we just talked about but just a little bit more defined is that right Dan oh yeah my goal in defining this just provide a little more certainty to when we can expect the networks to launch the network upgrades to launch and I've built some of the stuff around it with the idea that we can use it for either large infrequent or shorter more frequent artworks the gist of it is that we would set aside four weeks out of the year that we would target launching our

network upgrades in and Friz make a discussion I picked the third week of January April July and October to start the discussion and they're right there each four weeks apart they dodge August and December and fourth of July for the US you know big vacation times people want to take so on that subject one of the big advantages is it provides engineers a calendar that they can take a vacation I'll have to worry about a hard fork being plucked in the middle of it you know when they plan their stuff in advance and also for network operators node operators when they know they need to do upgrades they also don't need at the same time worried about and network upgrade unless it's in the window so some of the details of it I split the network upgrades into three broad categories just for the sake of convenience

first ones roadmap which is once we've always done I created a new category priority which is something more important than a road map upgrade but less important than the third category which is a critical network upgrade that's when something's broken and has to be fixed and we need to do whatever we have to do so the three hard Forks we've done before homestead Byzantium Constantinople movies roadmap upgrades and like the DAO fork spurious monkey and tangerine muscle would be the critical upgrade examples and you know there's some standards around priorities you know do those you know like the third week of the month and critical do it needs to be done and again the motivation for this is to provide a certain amount of scheduling certainty it's kind of like you know Patch Tuesday for never got

grades cool any comments on that beyond what we talked about earlier yeah I'm completely supporting this it would be very good in combination with what we just discussed before because it sort of looks at the problem on from the other side from the side of people who actually have to run the nodes and they need to certainty about when these and grades can actually happen great any anyone else okay actually quite different opinion about that I mean I don't mean I don't like the idea what I think we try to put a lot of constraints and decisions how the process should work at the same time and

we haven't proven yet we can keep the deadlines the skill for Istanbul for example so I would like white with applying this one until we we actually have a smooth one for the first time so one of the things that this damn not is it Constantinople is this priority because we didn't have a deadline to work towards the first time have deadlines and that's why I wouldn't like add additional let's say requirements to that so they don't process work studies in the absence of deadlines tends to result in a longer work cycle with the same or lower quality so you know putting some sort of

certainty and it's not you know if we miss a window it's only three months and that gives us you know there's less haggling about well can we do in January do we do it February what if we could sneak an image in Christmas a new year it it removes some of the arguments of the process so there is there isn't a theorem thread on this since that's the theme because we're taking a lot of time on these I would encourage everyone with an opinion on it to come in express it especially if you want to change the dates it's still not in draft I haven't I need to do some upgrades on it I think one of the authors put some comments in there I'm going to upgrade it but yeah so there's there's new three magicians threads mention them so if I'm understanding public correctly he's not

saying that it's a bad idea I also strongly support this idea I think Pablo is kind of suggesting let's stick with the plan we have the deadlines we have for the next hard fork and then like make this something that happens after that assuming that hard work happens smoothly and I think that might be an okay idea yeah I think that right now Istanbul is targeting the third week of October so what you're saying it's compatible these template was compatible with the old proposal right yes that's why I picked those three months okay cool the next yeah $15.99 oh well is that that it's yep you 1515 59 sorry is not in sorry 15 $15.99 I'm not sure which what it is is not in draft yet so

if someone needs to do something about that if we want this to be proposed for Istanbul oh the merge bot might have caught it okay yeah someone go in there and fix that I don't know who wrote it no no it's not currently a correctly formatted EEP it's a random thread that the talaq started ah I see okay sounds good thanks Danny next up we have an EIP proposal from Martin it has really pretty graphs on it oh the link doesn't work let me try to find the right link and post it in here but like the graphs like and stuff like that put PowerPoint graphs to shame so everyone should check it out and it says page not found Martin did it go

private no it's I can post it in the okra duck channel yeah go ahead but it's about specifically pricing for yeah try size dependent opcodes exactly so briefly yeah I did a lot of measurements doing full sync with the guts instance which was customized like do some better aggregation of where the time is going and yeah sorry doing job but yeah definitely linked on his work but it was working a couple of hours ago through and forth having blood so it try drink easier so wait it's working for you Trent yeah I posted it in the chat

now that's a 404 maybe there's some consistency issue and github right now yeah yeah but the thing is this focus is on trying to make the operations which have deteriorated over time put them back in balance again and those specific ones that I've in for our s load and balance and there are two versions to this in both of them the s load would increase by a factor of four from 200 to 800 and in version a the balance would be repriced to 700 from 400 so they imperative with Xcode sighs in Xcode - in the other version I also introduced

off code called self balance which would be priced very very cheaply at like five because that is theta the balance of the executor contract something which the EBM guarantee is guaranteed to have already in a cache there are other things that could be done to make all ol op codes more equivalent equivalent to heavy but that's not a really concern for this eat because the the main goal is to make sure that the things which are deteriorating and might deteriorate in a bad manner that we fix that problem for

now and maybe how to fix it again later but that all depends on how how would how this state size what happens with the state size in the future with all the one of the X initiatives if you have any suggestions yeah there's a please tell me we don't have to discuss it everything now but I'll be glad to get some feedback on this before I before kind of finalizes okay Thank You Martin and yeah the link is working now if you all want to click on that and the agenda it should pop up maybe you need to refresh the agenda I'm not positive how it's gonna work there so for working group updates there is a pretty thorough

post by Aleksey near the bottom of the agenda and I'll let Alexi go ahead if he wants to go over that otherwise we can just read the update on there what would you rather do Alexi I mean I could you know go over some moments that said that that might provoke some questions so I would like to get any initial questions there yeah great so so basically you probably saw some of you saw that yesterday publish the stateless client prototype which is also outlining so essentially collecting data how big these proofs would be and what are the next steps and in short the next steps could be either to try to bring those sort of to bring the the bandwidth requirement for the

by using some more statefulness or to use darks so that but the thing is I really wanted to get some questions about is the there's something that come out of very sort of some time ago it's called probabilistic estimation of contract sizes essentially one of the issues that in the real estate rent and now is in the in the thinking protocols is that how to to get to figure out the size of the contract meaning how many items just the contract storage have and so for the state rent proposals the idea was to introduce it into the states like over two stages using some two hard Forks so but I thought about it some

time ago and I realized that you might be able to actually estimate not precisely but approximately the size of the contract by using the assumption that the hash function K Chuck suite 256 has some kind of on random Oracle property and I will probably publish it in in a couple of days because I have some code to do this estimation already but essentially for the state fees proposal that would mean if we could bring forward so this is in number four on the in my post my agenda so there's this change J in the latest state fee proposal which is the fixed repayment for contract storage and this is the basically this is the requirement from my point of view to increase the block

size limit because it will stop kind of acceleration of state growth so I'm hoping that this probabilistic estimation of contract size could help bring this change forward to the first hard work so we can actually do the sort of harmless block us limit increase earlier and so basically that that is the main thing everything else has already been out of the out there and yes and after the number six as well it's I'm also planning to do some more research on linear storage because it could be integrated better with you as an engine but I'm kind of I'm still waiting for some further updates from he

wasn't team before I kind of launched into this that's it so everything else is Hudson said you can read out and then just come back maybe next time if you have questions or come back to me personally okay thank you Alexi does anyone have any comments okay awesome so there's still the Istanbul and ëthe 1x roadmap planning meeting which is gonna be quite different if we do like upgrades that are more frequent but we'll still go over the EIP s that we're wanting and whatever the next hard fork is that's April 17th and 18th Boris is there any updates on that that are worth mentioning

we are confirmed for full node thank you gnosis team for hosting more and more people are I believe using that event and the deadlines we have in place to propose oops I just saw one from Virgil a ton of them are related to pre compiles and we can also use that live format to I think discuss what a more continuously process looks like there's lots and lots of meaty stuff out of there which is why I wanted to put a deadline that was even earlier than the hard deadline for free proposals so going well 25 people or so have confirmed that they're coming now just to be clear there is going to be like a

remote participation option for folks who can't be there in person right there is Trent is helping me out with the live stream so we'll capture all of the formal stuff so we'll have things to refer to I know that Virgil is gonna be in Korea and Vitalik wants to dial in so Trent and I are gonna work on timing to make sure that we don't have to make people get up in the middle of the night if they want to present great and I wanted to say that I really would encourage any kind of client developers from gas and priority to come around because this really makes these meetups much richer in terms of like the content so yeah so if you are you if you can come I mean if you somewhere close

by so please do sounds good anybody else have comments okay yeah just one more quick thing if we do end up adopting this kind of like quarterly release cycle or doing more often during hard Forks more often and having a more concrete fixed schedule for it I think it would be much easier to begin planning events like this way ahead of time to get people time to plan to be there which I think would be really awesome just another nice kind of knock-on benefit of doing that as long as we don't overdo with the travel was that is there any place where they're gathered like where I can go over them yes so that the etherium wiki that's in

the schedule right at the top roadmap I'm I and others are keeping that up to date and with two three three that'll ideally be in github really the deadline is bringing people out of the woodwork so I'm basically constantly curating this and it should be relatively up to date but ideally it will move to github so you can actually track a feed okay anybody else all right next we're not gonna actually discuss Prague POW today as like itself but we will go over an audit update and a carbon vote update first of all we're shutting down the carbon vote in four weeks lane what is two weeks two weeks minus one day so 13

days I'm gonna actually hard code that to a particular block number so I'll just um go ahead and do that and then just like post the numbers of people know specifically when it will be cut off so if people want their coins voted definitely go in there take them out of your vault if you want to you're like all your safety precautions for your cold storage if you feel really strongly about this yeah just to remind people I think the way it works is the the tokens the ether needs to be in that address up and to up to and including that final block number but after that they can be moved and the boat account will not change yes good point and as far as the problem how audit update goes for some reason people don't want to give me money for stuff or the

cat herders money I should say because we're we've been going around talking to people and it's not really working very well because people are kind of skeptical of the cost I will say Atlantic crypto gave 10,000 died which was really nice of them we do have a get coin set up that I'll link to and we're just continuing to fundraise that's the only update I'm confident that I'll have a lot more info the next time that I am on the court have call and then I'm gonna post the get coin in the YouTube chat and and Oh someone put 5,000 more and and the zoom if anyone's interested in sharing that link around so there's

the link and I think that it am I missing anything cat herders anybody okay sounds good then that's all there is for that next we'll go through testing updates Dmitry just a second I'm totally confused sure is is the audit something we would like to have or is it a showstopper yet funded so I consider those two separate questions so it's something we would like to have oh I see what you're saying Oh or is it a showstopper okay I don't

believe it would be a show stop personally but that's something for us to decide after we figure out if we can find it or not that just leaves everybody hanging all of this argument just going on and on and on did we did we make a decision on this already or what I mean technically we have the community there are some loud voices in the community who don't want it anymore but I don't care about loud voices in the community what when we decided to do an audit did we decide that we needed an audit or that if we could get an audit we would use its results as part of our decision if we could get an audit we would use the results as part of our decision the

reality of the situation is that it may not get funded and if that's the case then it's not happening okay so if we cannot fund it it does not affect our decision that is my understanding right now that could change if people speak up within the core Deb's or loud enough within the community that the core devs start to listen because as it is right now the core devs are the ones who are coding this into their clients that is my understanding I however am NOT this isn't my decision I'm just saying this is my understanding of what's going on okay I'm more confused I mean like I don't make the decisions Greg well I know I'm trying to understand what decision we already made it's I mean did you see my

reddit post no I never go on reddit I'll send it to you that should clear some things up and similarly with the carbon vote it's being done it's going to be finished what are we supposed to make of it whatever you want I yeah I'm sorry to be to be rude here and butt in but like I think we've sort of reached consensus and I think you may agree with this that like these are not technical questions and that this awkward is is probably not the right channel for this stuff so are you okay with our taking this back to git er or femm or one of the other other characters or something not because all the discussions I look at it appears that somehow the outcome of this vote is supposed to affect the core devs

decision whether to put it in when I thought we'd already agreed as to the parameters of that decision and if we haven't this decision just is is never-ending I guess the the answer that I have in my opinion is that if the carbon vote says something wildly different than it does now or if for some reason there's enough people within this call or outside of this call who raise enough us decisions can change it's not like we make the decision and we're going to be stubborn and never change it I would say that right now we're still we would still go ahead with our previous decision okay this is a

mess but this isn't the time to discuss it so we'll move on just decisions when we get more information so I think it's not unreasonable that if we get more information over time that it is no change obviously because a very technical reason like if there's some massive bug and we came in forward everything at CBC ideas Dalton rapport with them but there's also like a lot of folks in the community that like a change of decision that's not unreasonable that we keep our ears to the ground yeah what was that Amin yeah I mean here I was just trying to get clarity on what you said because I didn't quite understand it when you said if you at

the risk of stating the obvious it's like you won't implement Prague without not it right hmm I don't know that for sure we're just kind my next step is getting the audit funded so I don't really have an answer what I mean is like is there a scenario where the audit doesn't get funded and we still implement Prague bail there is a scenario where that could happen yes okay but I'm gonna hold off and we'll talk about other channels yeah take it to the Fellowship of a theory of magicians that's probably the best place can someone post a link to the best threat on that I know there's a really good one that someone probably has pulled up and if we can post that then we can continue discussion there and thanks everyone for being so understanding of the fact that

we don't want to get into a 50-minute discussion on prog pal get in this chat when we don't even have the audit funded yet and it's not something where we even have a date that prog pal would go in or anything like that so I didn't want to waste that much time on the call today on it okay so let's go ahead with testing updates that be Dimitry and maybe martin so hi everybody I'm glad to announce to you that I finished stabilizing retested application on blockchain tests so now it could run over the blockchain tests including the general state test in a blushing form on any client that support RPC interface

which is described on retested wiki page and I also have instructions for you to use docker build so you could come when retested and run it on your client if you support this interface see how it works your client also I was thinking about this Jenna Jenna this general state channel is for Mark II IP numbers that you have we have it fixed right so he could fix the for claims as well and I opened the EAP proposal about that discussion about how convenient it works

so I think it should be like a eyepiece for claims must accept it should not be changed because currently I got proposals from some people to rename the Constantino both weeks work during the test development work name is being changed and it requires me to change all the tests and source offer shouldn't be client and stuff like that so once we come up how it remains a port we could talk about Genesis format for every button for every client and the bare minimum of Genesis is described to be 50 to 80 page and this pair minimum

is required to run blockchain state test so we're ready to switch to mp3 test execution this is a CBP client not test generation yet but at least test execution is already possible then you could think about debugging through the RPC and I feel no more blackbox testing they use Excel you could use metals to debug the particular transaction particular tests and I have a test role to work so please take a look at 3:30 anybody else have comments questions or updates ok usually by now we do client

updates and then he was them in research updates but I wanted to kind of get a quick feel of the room on if people think these are valuable these used to be the main focus of the core dev meetings because there was a lot of movement on implementing the IPS and some of the clients were earlier on and needed to talk about their progress what I don't know if we have a full purpose for them now and I've heard some feedback that it'll give a chance for people to kind of shell their client not that that's the end of the world but I just wanted to hear people's thoughts on that if we should continue to have this in the call personally I don't need to report my progress but it can be on-demand so I

mean people can still put to put an an agenda item if they want to give update up for something that happened because it at least happens to me that there are some some meetings when I don't have anything to report like most of the time I don't know what detailed and a client update and it's usually just like oh yeah we released this version and I don't think that's useful for anyone really unless it's like you know this is this is a version that you have to upgrade to and I want to highlight that oh we're doing like a million things on

the client otherwise and I'd like there's not enough space or time to really dig into those details either so I like that that doesn't make it in so I think it might be good and having an on-demand might be good but like me personally I don't know what your report BOCES are I agree cool I'll start having a comment in there that just says please add a comment to the agenda if you have any updates for now we'll keep the testing updates because I feel like that's pretty important personally but if someone wants to get rid of that talk to me privately I only think it's important because testing is something that happens in every client and we need to keep up with what's happening there go ahead know which clients are attending

the calls it gives us off form of an attendance report you know so which ones are involved in the process hmm that's true that could be taken care of just in the notes to saying which client each color represents that there's one way around that and actually we already do that so yeah that's that's something we already do in the transcript / notes so that that's something we can continue to do I was gonna say and then same thing with awasum we'll just if they have an update go in there who feels like we need research updates in here Danny what's your opinion um I mean leaving space for them is probably reasonable I don't know it just

kind of depends on a week on whether we have interests thank you sure yeah I think that for the a lot of the people in this room don't attend your calls for the ëthe 2.0 implementers calls so in my opinion it's valuable and so much as people get an idea what's happening on that side because this is more than 1.0 in one point X side-eyed something we've seen this worked really well in the e azam team is that people post their updates to the agenda in text format and then only we only like on the actual call we only discuss questions that people have ideally everyone would have a chance to get their updates in like let's say 24-48 hours before the call and everyone would have a chance to

review them before the call that definitely saves time okay I'll put that as everything including research in that case and then Danny or Vitalik they have anything can put it in as an agenda item that sounds good to me Oh Danny did you have a comment so I just realized that I forgot to say something about the working groups and it's also related to research updates I think by research we normally mean something from zero in 2.0 I think usually and so I wants to I want to create I don't know how I'm gonna do it but I'm welcoming ideas a new working group within mysterium one point next with the

goal to create a prototype and then later proper implementation of the finality gadget for the big user in the beacon change the finality gadget and I know that there are some proposals from italic you can see the link in the in updater that posted an agenda but I would like to explore kind of various ways of doing it like because vitalik was suggesting to do it by let's say BLS pre-compile but maybe it's a practical maybe it could be done in native code and things like this but I wanted to get this work started somehow I mean I don't necessarily mean that I will be leading it but I want to kind of launch it so I wanted to put it in yeah I don't want to leave but I'm

happy to be a part of that kind of advice and part of discussion cool you know I know not suggesting that you abandon it but we'll find some we'll find somebody to lead it and then we just got so is that an action item calling for a lead for this initiative yes if somebody what really wants to lead such an initiative yes we would kind of welcome and talk to them and see if you know I don't know if we make interviews for this world let's have a conversation yeah and this item at least from my understanding seems to be a high-value community and people want it so

something we should probably I mean is except that we're driving forward kind of in parallel and making sure that the processes when the API side are such that they have with usable path the embedded entity one step but having somebody kind of drive it a little bit more quickly works well okay any other comments all right do you have a research update today Danny that you want to discuss are you good a lot of simplifications on the phase one custody games for data availability and that the phase one specifications are very being done very

actively where's the phase two conversations continue kind of emerge and happen the design space there they have said before is pretty large so we need to narrow things down a lot of of these clients that are working on the tube you can chain are getting near pressed net data so I'm excuse me interesting stuff when I probably awesome was that Justin who did the custody game stuff yeah I think it I think it's sweet okay I think we finished a half-hour early does anyone else have things they want to discuss any items that were missed

there was an PLAs research anything here we had a beta release this week or actually last week that introduced a new algorithm for gas estimation that is I'm gonna say we call it gas exact nation it's accurate down to two one gas even for the IP 114 opcodes and it doesn't use a binary search so it only requires one pass through the transaction out codes to get the estimation well we're planning on writing a detailed blog post on the algorithm and how it works but I

think it's something that a lot of other clients could connect use of for performance reasons cool hey looking forward to it yeah I posted the release notes link in the agenda for this week so if you're interested anybody else yeah I have one item I would like to mention so in olive we preparing to implement this new discovery protocol for the FPTP that Jeff what I understand already have so I just wanted to mention that that's that's I think it's already ongoing effort and if you have some

comment or you would like to change our mind about that let us know okay anybody else all right thanks everyone for coming I won't be on the call in two weeks because I'm going to a giant star wars convention that should be pretty awesome and so someone else will have to take over I'll find someone in the cat herders too or maybe a few people to group up and do it or if anyone wants to volunteer from this group we have right now feel free volunteer to come to the Star Wars which absolutely not because they're sold out and have been for months all right thanks everybody

[Music] [Music] [Music]

[Music] [Music] [Music]

[Music]