EIPIP Meeting 3

Transcript

okay welcome to the third EW IP meeting I am hosting it as people are getting ready for if Denver which starts today / tomorrow or at least the community activities are these meetings are every other week every other Wednesday alternating all in between the awkward ends calls at least so far discussing progress we can make to the EIP process thank you all for joining today

we have a couple things on the agenda I'll just go over the agenda first there I made a PR to EIP one that has some consolidate some information the number two fleshing out requirements for EIP is related to when something is more requires a little more work than like a the normal or the typical EIP that is seen and how we can communicate that effectively a code of conduct process documentation which I'm not sure which that one means is that one that you

added puja okay and in progress on the IP we clean up any other business and the next call is there anything also they would like to be brought to the agenda so I don't know that site encompassed in any of the points but this proposal you had a caller to ago around like this sort of DMV office style section on a IPS yeah is I just be curious to get an update that but I don't know it's do any of

those because I don't think the P the first PR you made the EIP one doesn't cover it but yeah it doesn't but I can I can give an update on it which you still need to get someone to look at what it would take to change the bot to make it possible okay and otherwise I've generally heard that it's a good idea it's a matter of how do we do it and in in like an effective way yeah like implemented technically it still is still I got unknown thing and I didn't follow up with Hudson about who in the EF kit would have the technical

resources to make changes to the bots cool yeah we should make that as an as a like an owl sighting outstanding item to follow up on good and the IP one I don't know if you have if all you had a chance to look at it I don't have an easy way to I'm on my mobile so I don't have an easy way to link it to the chat someone could do that yeah Tim and I have gone back and forth a little bit on it yeah I was looking into it and I had a couple

of questions there I'm not sure this is the appropriate time of it because I was there in the previous call when we discussed it but at that time I did not come to my mind so I'm just sharing it I like the response of Tim about the upgrade usage of upgrade instead of hard folks so in my mind I was thinking that like if we can since we are making so many changes to a IP one not so many but yes we are making some will it be a good idea to update that as well like I see it in your line - 9 1 / 3 - 4 in the link that I have shared there we are writing aetherium

hard folks and updating what hard folks started on what block number so even go ahead and update it over there as instead of aetherium hard folks make it an idiom upgrade so that would maybe all futures condition any time I almost wonder so because this is like a controversial topic like people have strong feelings on that I almost wonder if it's maybe just for a separate PR you know like basically take everything that's an EIP one right now switch hard Forks for network upgrades see what people feel about that yeah yeah I like the idea it will give more visibility to what major change that we

are proposing and will get more feedbacks on it so yeah I like the idea yeah the other thing I was wondering about was like in your PR you mentioned that we are referring to more on two three three and two three seven eight both of these VIPs seems to be in the status of draft so is it the right time to add it into the EIP one I mean like when it is not fully accepted so should we first move towards working them to be accepted first and then work on this PR to be updated in EAP one what is the general thought over here I I thought about it the other way as when I was

talking with Alex bergüzar see the he walked he he wanted the governance information about the process to be in he IP one not in the EIP so he it we moved it to draft so that people could link to it and it was a little a little more easy to use but he would have rather had it not have that information and but and so my was the word compromise was to make it to draft and then when a IP one is updated we can remove all of the things that doesn't need and make it final it's just a registry there are a couple of things associated

to it I remember sometimes back the discussion of you know not having major changes to EAP 1 but to the separate AIP the way we are following like EAP 2 3 3 that defines basically the process of how to create a meta EAP for an upgrade and the EAP - 3 7 8 at this point of time that describes the core EAP eligible for inclusion but I'm not sure about the process that we are going to follow all along like about a year before from now we will for Istanbul and we were considering the model or you know flight model and the bus train

model so that was an experiment that we are trying to perform for this upgrade and I considered this 2 3 7 8 again and similar thing like this is another experiment with the EAP centric model and EFI is just an intermediary for this EAP centric model so I'm not sure that if this is the right time to you know include in EAP 1 when we are I don't know certain or not certain to follow this in the future upgrade so I was just wondering like should we keep it separate for some time till we decide that we are not gonna follow the flight

train model we are gonna follow the EAP centric model or if something else is proposed in future so keeping it separate like EAP one not to be updated again and again and whatever heartful process or a great process we are gonna follow should be kept separately that was just my general thought process around it and I mean like open to discussion yeah I have some fast I didn't want to dominate the discussion so give some the challenge with like having two resources I get your point pooja and I agree like II ip1 so far was very conservative in terms of upgrades right it's like the

process would change a lot and it would have to like have been changed for a long time before a IP one reflects it and so now it's like if we change it and you get to a spot where you have like a IP one referencing something that's kind of in flight the on the other hand the challenge you get is then it's hard to point people towards the process so if it's like we need to point them to IP 2 3 7 8 it's like how the hell does anyone ever like find that the IP so that's that's just my concern there it's like you're trading off like stability versus like making it accessible to the community I get that because of that agenda item is about the documentation of the process

so if we do not want to do it right now it's fine when we get back to that documentation of process then I will bring it forward again but I am fine with the appear that he made and these were just couple of concerns one de IP a hard folk to upgrade which Jim proposed that it should be a separate IP altogether and second the EAP two three three and two three seven eight should be converted to accepted or I mean like because that isn't draft right now so those are those were my two points here yeah so in one sense it's in flight so it was in flight it was much more in flight when we were discussing whether

or not we wanted to try out the EIP I the the EF I model but then when we have the core des call and everyone you had consensus for us to try it I I feel like e IP one representing the current consensus of the core devs and then just if we need to update it a lot then cool people will always know to go to a IP one again another problem would be if we we tell everyone to go look at the IP 20 I can't the number right now I'm not looking at it as soon as that process changes for some reason and there's another PR then that that like knowledge link is broken

and all the people who would go there would go get old information and then be I frakked fragmenting and this kind of goes back to my thoughts for the proposal last call is having one single source of truth for everyone in the community about what is the core devs and what is everything like I just was thinking when I was writing out the definition of the hard Forks and stuff that there isn't really a place to for people to reference to say yeah this is how the core devs are thinking this it's just sort of amorphous everywhere and it just would be nice to get all that

documented yeah I believe that documentation will help a lot like sometimes back about a year back I tried that in AP 1929 which is still again interact which was shelled because we were doing some experiments then so I think that this may be the right time to create a documentation about what process all coordinates are following it's it's every time it's with the devs only and it's not documented because we are still in experimentation phase and have not set something in stone that this is what we are gonna follow so I think we can like shell which here only

and maybe try to pick up this topic after a couple of week and see if we have any different thought or we would want to go ahead with it yes that sounds great and I can read the ones the ones you linked in your proposal to our and later in the agenda as I still have to get my head around what what is all of the context around it any other thoughts on the EIP one pyaare before we move on oh I couldn't find a mute button but I I'll have a

small PR I about the hard fork Network up great thing yeah I just looked and the hard Fork is only mentioned twice in yeah IP one as it is so it's a really small change so up shortly oh great the when I had when I was writing out that section I I'd say if we call it upgrade then what do we do about the Dow hard fork and where do we put it so I think the way I think about it mentally is like the Dow hard fork was a network upgrade that led to a hard fort like Tamiya hard fork is a possible outcome of an upgrade right but the Dow you know like that eeep implementation is still

an upgrade and I so there was a conversation about whether upgrade is the right word I still feel pretty strongly about that just because I know the person who is kind of arguing against or saying upgrade assumes it's better but I don't really think that's true I think it assumes its new right yeah yeah and like the match katalina update is a great example of that where like people it's new but people don't like it so they don't want to upgrade to it so I think yeah just saying we had a network upgrade which led to a hard fork or a network split or whatever but all the nodes that are still on the etherium chain today did upgrade right so so you know it's just one of the network

upgrades but it led to a fork is how I would see it I don't look at the everybody else thinks yeah I agree to it and in fact I find it more of a reason to use the word of upgrade because we had only one heartful which was actually a network upgrade but that led into a fork so we had only one hard folk and rest all their upgrades yeah and there's a lot of like controversial dynamics around like what led to the hard fork like the fact that you know there wasn't really a hard fork for like a week but then some exchange you know started trading et Cie and that kind of revived it so I yeah I I'm still like in favor of calling it a network upgrade you know it's one of the cool you're talking

about calling the dow fork live great yeah it's basically so the the EIP right like that that was patched there was the upgrade and the fork is the process by which like the nodes that are still running it is even if that run that upgrade right so the way I see it it's like the upgrade leads to the fort but they're still you need to have the upgrade to get to the fort right it's it's it's linguistic but with the except this right like if you put the dowel hard for it aside I think everybody

everything else can be called the network upgrade right anything soldier upgrading whereas and the problem with hard Forks I I find is it's kind of a loaded term because of Bitcoin right so Bitcoin has this whole thing around like you can't do hard Forks and it's like a a terrible thing when the theorem has done that ever since the beginning and there's like this one upgrades that's why it's pretty bad or good I guess depending on you know what what you think but yeah tell you you would be a very good practitioner but I agree with you with terminology can is the synergies accept accepted down like the rest is clearly

upgrades and you can or definitely argue that I'm the great doesn't have to be positive it's just a modification of going to a newer version yeah and I think you know it's important to state that because like I you know I still think it's a really important feature of these upgrades that's like they're not certain to happen and that kind of puts the responsibility on core developers that you know not put stuff that's like too controversial to not split the network and that is like it really that is like a really important concern but also think we want to highlight the fact that like a theorem just has this culture where these changes will happen

and and and I think kind of moving away from like the terminology that's using Bitcoin is it's kind of good let's regard I can see how that's also a controversial opinion I don't think it is I mean to be honest we are going to get lag we put we're a crypto Bitcoin doing their jobs and they're trolling and once it's they're going to get over this and it would be done I guess it to be to be honest it's in my opinion what always great instead of out fork was always a very poorly choose chosen name especially when you think between the like people who thought there could be like you know remaining coins from the older chain because of the art for it

could be like a beacon art for between a beacon cache yeah ESV yeah so and that's something I think is like even more important for aetherium is like not having that perception for a couple reasons like a because of the Hulk that's how classic started be because it just creates confusion for current users and see because of a theorem 2.0 right like you don't want to create this perception that like every upgrade is like free coins or like has the risk of having two different types of coins exactly yeah so yeah I agree with that so in that case you you in that case in terminology could definitely define the thou of great as a contentious one versus non contentious one and then ya

have something that is more understandable for a for for for everyday users yeah I think that's like a really good way to put it so they're all network upgrades the Dow was a contentious upgrade which led to a hard fork therefore we kind of want to avoid contentious upgrades because you know they have a pretty significant toll on the network yeah that's brilliant as as you're saying that I'm realizing it or it would fit as its own section really well in the ap1 like something that just described exactly what you said and then people can reference that as the definition of like when people are yelling about it on Twitter I feel very strongly about is I don't

think we should overweight ET C's like thoughts on my query theorem does so I'm sure et Cie will not be happy with that definition but I you know I think we should try to focus on what's best for aetherium if you see well say like oh you know where'd the original chain blah blah blah and that's fine like the EGC repo can say that if they want but yeah I mean we don't have to compete on there at either agree on there it is different and yeah do you recommend dislike like whatever we are discussing here in a IP IP and creating peers and everything I would like to suggest like we should whatever it is our summary or the action

item that comes up in our meeting we just share it with the alcalde meetings one as if they don't think that we are started we have started making decisions on our own and like not yeah yeah yeah for sure and any PR will do to a IP one you know will need to get approved by people and and that's gonna be like a quite a quite a process yeah so and I guess as an action item for this I I don't know James do you want to take a stab at like rewriting some of the parts of your PR like I feel like your PR is a really good start for all this stuff yeah do you think that make sense I'm I'm happy to help how I how I can but

yeah I I think it would be removing the parts that I had done that are relevant and then moving it moving i adding the language you did to yours somewhere else so they can be evaluated separately so does it make sense that maybe yeah so I have my tiny PR that's like two word changes to go from network upgrades see him that's controversial you know if that's not controversial great you have maybe you kind of split your PR into where one of it was just basically format adding you know the list of upgrades and the table of contents and yeah one added like the other part added a bunch of definitions so maybe we get the table of contents merge my small kind of you know

hard Forks the network upgrades merged and then once those two are done you basically haven't changed the IP one much then we kind of take the content from your the rest of the content from your current PR and and kind of push that you know suggestion for a better definition of network upgrades there's hard Forks and and whatnot yeah I like that a lot I the I think so there's network upgrades contentious network upgrades this is just so I can get all my thoughts out into one place for the notes the other kind is a key emergency up by addressing something critical on the network they're like emergency

upgrades emergencies kind of alarmist but I don't know what it is kind of the case because they're also upgrades right like they're literally the same thing but it's just a time component that changes yeah and there and they they do feel sort of different yeah then the normal so there's the contentious upgrade if we just say upgrades like the normal piston bowl whatever then there's contentious upgrades which is the DAO fork like the odd man out is the is the ones that there was like the what's the one when that Beijing when the sauce attacks were happening spurious dragon tangerine whistle those that call

emergency may become your glazier a priority yeah I don't I guess I'm I'm not saying let's figure out exactly what it is but finding finding where that fits in that narrative I think then it's completely full yeah a complete concept [Music] I just really like the I fix yeah they really are hot fixes anyway I feel there must be some software

terminology for that the cloud fix is the actual thing yeah yeah if there's a if there's a bug in a release software and then there's another one that needs to be released right afterwards it's a hot fix and they're usually really small upgrades or deployments and it's just for that one issue using bug fix can get a bit weird though because if you take a look at things like near glacier for instance I mean it sort of implies that the difficulty bomb is a bug which it's not so like I don't know if bug fix would do it but the bug is not the difficulty bomb the difficulty bomb itself is not a bug but it's our failure and I and I was saying hot6 is in like a

hotfix like a fix for flim seeing that hot yeah I think fix makes sense so we'll add that as a thing to reconvene on so then that's we can move on to yeah

and I posted the like in the chat I feel like we kind of have just summarized its network upgrades and then there's like these three subcategories which are probably in the wrong order you know for most commented he's coming yeah we can call the hard fork the one that is people's my I won't get back into it but it's it's it's just it's J its gelling in my mind which I which I like so the Louie this is kind of going back to conversations we've had unless there's someone else who wants to go into that

that portion we can move on I just give a little space for someone to say something so an overview to get people caught up there's an an interesting problem that not yeah not all the IPS really should have the same requirements of due diligence on both the EIP author and the IP and and the Chordettes so the like the the biggest and I kind of see them as sort of three categories that I've observed as we've been going I had they like prod tau is one of those categories where it was where they even required a like a software audit before

it would be considered and that hasn't been a requirement that's put on a on a I forgot what I was saying and and yet are you yes yeah any other eat that has gone in there hasn't been a hey let's make sure every EEP has a has a requirement for a software audit then there's this other level that's kind of existed which is currently the elliptic curve precompile one where we don't have a good way of saying hey you need to do this if you're doing this and I think there's another category of

like a IP 1559 where there's both the technical challenge as well as the economics right so it's like do we you know we want to do this big change it's assumed that it's you know gonna make fees better but how do we actually like how do we evaluate that right so that's yeah that's another question and so far what has kind of come up in conversation is you have having if in in the case of the elliptic curve cryptography one I did the core devs don't have domain

expertise in the very complicated math that is that like that kind of high level cryptography oh they're good well just having other members who have the expertise to review those PRS like explicitly make that as a if you're doing something beyond normal then having experts come in vouch for the work you're doing so yeah so I think go ahead yeah I was gonna say I think that's generally a good solution and and

kind of loosely enough define that it works so if you look at like 1962 versus Prague pal I think frog pal because it touches you know many different parts like there's like the biggest reason you're doing it is to delay a six and that's like a whole kind of can of worms in and of itself with hardware having you know a hardware expert and a software expert kind of sign off and and and say that this makes sense is is kind of need whereas for 1962 you know you don't have this hardware component as much you just have this this kind of specialized crypto so like what Lewis proposed on the previous Stahl having different

crypto specialist sign off on it I also think it's kind of sufficient so maybe it's just about that it's almost like peer review in a way but what peer and what review means will be very dependent on the context I think with the prompt our example one thing to point out is that the algorithm it was replacing each hash also had an audit covering basically the same things so for each hash it was being subjected to the same scrutiny the other thing is replacing was being put under so I think that's one thing we need to consider do we lower their scrutiny standard or maintain his crew D standards easy yetiz so I just want to react on top of all your comment I've know if it's a team of gems would say it but

basically I'm open either indeed three for the VIPs like three sort of VIP that are today not perfectly the vote for which the process is perfectly fine there is prog pal where they're probably cleaned up technically it's it's a political one and and for which I think the if they could have done a better work and doing it even before the all the technical analysis because I believe that the pecan analysis was sort of given as a way to just delay it indefinitely because it's because no one wanted to take the political heat of accepting it on this like they say oh I'm for it and and the reason is and I

do blame I know we had a discussion like on only one core dev recently about miners and and people coming in and defending a IPS in the case of a very Prague pub he should have been said for instance if we wanted it to have accepted put this potentially accept it until technical completion and just having the narrative of explaining this is coming as soon as it's ready sort of soon right narrative we have movie 2.0 today but apply to program so that would allow us to evolve this whole technical auditing biggie but as of today is useless or we don't already where it goes and probably could have helped I say align everyone to say okay this is

where I was gonna happen it's gonna happen soon as long as we have the right on it and everyone is happy with it and but it's going to be extracting and there is no question around it or going to be refused in first place so I think this particular discussions really pushed by one of the stakeholders one of the miners group miners who came in for the previously IPs when weather-wise like the discussion about a reward reduction and keep coming to the to the to the to the core nephew defending and also try to modify the narrative on Twitter and various media to say this is in and you have no choice around that was just restart about a program as of

1962 I think that the view perfectly said the parameter is the technical expertise of the of the court live and every IP the outside expertise are needs to be viewed with a lot of of sorts and a lot of auditing and make sure that the core data actually the core develop developers actually comfortable with the way it sees and or and I in the problem if 1962 although Alex did a terrific job at coming at every meet a brief talk and come in explaining what he was trying to do the problem is that there wasn't enough like coordination or enough like the process wasn't clear enough for him

to gather the sort of natural crowd to come and vouch for his work and I think when we proposed to bring like people like expert to mccraty Kobe go Jodie and various other krypter krypter would help but might we still need to ask people like a user than or or Martin to see what would make them feel adequate with integrating and yes I I agree with all of that I think one thing that's also worth noting is you know it's normal that these things are hard kind of you know to the

degree the of the change you're proposing so on one hand it's like you want to have a smooth process but on the other hand if you're proposing something that's totally new that's never been done on a theorem and this is like I don't know 20 billion something that's work that's live like I I think it's fine to kind of take the time to get to a spot where everybody's comfortable even if that that means that the process we won't be able to define the process for every type of new thing in advance right so like this 1962 is different from Prague pal and that's probably going to be slightly different from something like ungass right so I think we you kind of need a loose high-level

guideline probably to be more explicit than if you're coming in with this huge change the etherium be prepared to spend like you know a year kind of getting you through and just to set expectations because I think it would have been impossible to say like no to Prague now on day one just because there's so many like unknown unknowns right so so yeah I guess it's just like yeah setting the right expectations for the people who will take on this work while also making it clear what they have to do probably better than we do today yeah yeah I just want to come in top of this is there's one of them things I

found mr. disturbing to some extent is to some for some VIP the the the the people pushing it the the the champions are not always on the coding I think because we're talking about the twenty billion dollars were Network it sort of it should be a hard requirement to for some representative of the c---eight of the IP to be available and to come up like at any call yes and we should probably just make it clear to them when they should show up so I can I can see I don't know to take again 1962 as an example if we don't talk about it next week you know it's not super Val it's not a great use of

the person's time to spend two hours on a call being there but I don't they should check the agenda oh I don't care I mean we are talking about a great English system of twenty billion dollar system I mean and for which people are supposed to maintain it for 10 years or more like it should be they should be coming six for six months at like ten meetings it's also yeah yes so maybe that yeah that's kind of part of setting the expectations it's like yeah you should be willing to like iterate on this thing for a year you'll probably needs a non-trivial amount of funding you know to either get an independent implementation and audit or whatever and and like we expect you to be you know

whether it's on every call or to at least check the agenda and if your thing is mentioned to be there yeah I think that's that's reasonable you're probably still find some exceptions like reasons why there might be reasonable cause to not be able to make coatings consistently but yeah sure sure but like that'll just delay their progress right you know yeah and if they're fine with that like that's not the end of the world but a front yeah the expectation of to move it forward you show up people don't need to move ease forward

she just like you know she just been people on papers in mind I'm going to take an example for instance if we on the last course I was a new IP that came in about the thing about like strict requirement for contract to to like require specific amount of gas or something I don't exactly remember the content but the guy that pushed champion was seem very interested but the sort of reason the line reason why he never is pretty not going to get in is because I'm not because I think if the IP is not sound he's just because he wasn't known to the to the to the curve of people I think like if you were to become more often even if he wasn't topic wasn't on table it would have helped gather some

you know some sort of common agreement that this is needing so maybe an action item we can get out of this it's kind of a I don't know like a guide for like VIP champions and I I don't know how we split this between small leaps and big eats you know like there's not like it kind of like you know it when you see it but like I don't know we yeah if we get have like a paragraph or two I don't know if this is an EIP one or just on the eeap's websites like the home page of like you know here a couple examples of very large you know projects that

people have trying to do on the etherium and sears kind of what you should assume you'll have to invest in terms of time in terms of like money or just like technical resources and and timelines you know it's it's it's not going to happen in three months it's more like a you know six to 18 months type of type of thing yeah at least it can set the expectations straight and we can see you know the people I agree or disagree with that are we suggesting this action item is on the APA B team to figure out guideline for EIT champions or is it something that somebody is owning individually

I think it's put into the mentor category that was discussed on the first meeting yeah that would fit in there the so we had wish I the table of contents to have requirements the EIP requirements VIP recommendations and recommendations being just to try and summarize we've gone over so far if there's something outside of the of the expertise of the core devs you need outside peer review from experts if there is if you're if you're implementing something or upgrading something having a parity with the like

the previous implementation like how prog pal got an audit because you've had one that you should expect if the thing you're doing should should maintain the standard of the thing you're replacing I had them lined up in my mind and then they've all fallen out what are the ones I've missed yeah I think just like yeah the timeline attending court F's call peer review attending court have calls and then and then what I would add is it came up like what to do if something goes wrong I what is what is the plan

for action of doing it like in the elliptic curve cryptography one one of the reasons the death team was hesitant was if something happens with that EIP on main net they'll just sit there and look at it and say oh boy which is it isn't a good plan so like having a plan for if you're AIP needs to have a separate plan that is outside of the that needs to be made in advance and I think related to that what was said about 1962 I forget his name but the guy who's been pushing yet says he kind of assumed it was yeah Alex yeah so he said

like it's his responsibility to show up if something goes wrong so I think that should be like an it like if you're the champion on an EP you know you're kind of a person of contact but it should probably be two people so you know I Alex you know seems to take this seriously that's great but you probably want at least one other person in the world and I don't know if they're they're like champions for the eep I'm not sure what their title is or they're like who do you call when there's a problem you know I'm not sure that's gonna be enough though that explicit it's more of they need to have a plan and it needs to be written out and cuz the disconnect that happened was margin was like oh no what are we gonna do and

Alex the whole time was thinking oh I'm just gonna show up and help and if he knew Alex's plan of how to address like a critical something happening on maintenance and that was written out it doesn't have to be I depend each EIP might have a different plan but having that expressed then understood I think would qualify Kunis you need to list who are like the potential failure modes and what to do and those failure modes right so you know what happens say with 1962 if two implementations disagree right like how would you be bug that what happens if like I don't know some other thing goes

wrong and then I can see a bunch of dosatron pow as well right you know what happens if you know there's still a six on the network if there's still whatever so just thinking through like the risks and how on a high level you and mitigate them maybe hmm that's kind of different so there's like the security consideration section which seems very focused on like the implementation in a way but it's almost like the potential risks or yeah I think in my in my in concerning in 1962 I'm I'm one of my concern is actually not victory and Marian is marking is really worried about security my prime was the quality of the specification not because I'm

assuming I like this a bad job it just promised that there when I look at how Bitcoin is pushing like new great so there is Jeremy Rubin who is pushing something called checked on peg verify has been doing like work yes been doing a workshop like in first in February to explain to the community how these things gonna work and for any IP that big and that complex I would expect sort of not only adding someone on spot to react but also just the knowledge to be spread for anyone to you to grab it to understand and to be from from from zero so there's that so there's like a full

comprehensive specification then need to be a requirement because I feel if you get an audit for an e p-- you know you kind of want that specification anyway if you get like peer review you probably want the specification so I'm I'm not sure how to define this but like some like you know yeah full complete specification how does how is that different than or how does that intersect with saying you need to have peer review from experts in the in whatever it is I think it's intersects I think you're right it's probably inter second and and in peer review also mean

people like those peer reviewers coming to the core dev we've their reputation at stake and saying yes if you're good we can do in push for yes so it's completely overlap but maybe one difference so I I'm not sure what I think here but like the peer review I assume there's a specific that another expert in the field can read right but maybe a a virile upper won't be able to like you know grasp the entire respect whereas I feel like I don't know do we want the requirement kind of above and beyond that but like any core developer is like comfortable with this patch and I'm not sure if that's even realistic right yeah so like I guess for 1962 as

an example you know I can see a case where like somebody from a spec looks at the spec and it's like yeah that sounds good to me but I don't know we have way and Dino on the call but like and they don't have like that crypto experience so maybe they're not you know as comfortable with the spec and it's like who should we be aiming for in terms of spec parity is it like the Aztec researcher or is it like way and Daddo and Martin understand every single spec I think in that case I would have aspect to be obviously vouching for it and Dino and Martine and the way to have sort of a basic understanding on what could go wrong and how does it work and you know write sort

of generic like generic idea that no can you when you how would you be comfortable with we should show the AP I mean I can speak to what I am familiar with the engineering tasks but you know like I said to the rock crypto prep knows why is the g1 curve safe and novel by reducing mathematical questions it's not something I'm trained in good at giving answers for so having someone else come in who is trained and things like I'm talking to say oh no this particular krypton thing provides what it claims it's provides I think is something that would be valuable but when I look at these mostly I looked with the engineering perspective and I think people who chime in chime in relative to their strengths of what they're used to evaluating so that's why

some of the crypto stuff you know at least a cursory glance by professional cryptographer is needed even though maybe rock-solid I don't think many of the engineers on the car will qualify to say well yes it's rock solid so it's almost a validate this speck right so it's like the speck is clear enough that any court have engineer can implement it but they don't know when they're plugging you know the sort of constant to the curve what like is that the good one and and you want to be confident that light the sort of assumptions in the spec are are are correct and that's where you get the peer reviews value from right because we will take it if it presents itself as

you know this is good strong crypto and we don't know any better we might take it which is like getting someone who can validate those claims no this isn't good it looks like it but it's not because of NZ's that only a trained cryptography know about those are the report so it's usually not there than that and understand that the permit actually an implementation in any case the bugs doesn't come from the math they usually come from the way you write the code and and I'm not sure that even if the spec was vouch for the math test act of it it would actually save you from any any any trouble so I think why you should actually probably have is like

expert in both the cryptographic and then plantations part of the cryptography I think we're getting a bit trapped in the weeds though if I can try and pull out for a second I mean we're using this example if anything I think all this does is is is proof that we can't codify exactly what each is going to need if anything there should just be a reference that depending on the scope and just how critical an EIP is on a case-by-case basis there may be unique things to eg IP I mean it seems like everybody was more or less on board with the idea that Prague Pound needed an external full audit because it's frog pal I mean it's the guts of the consensus algorithm that aetherium would

run on I mean if anything we don't need to figure out what every kind of change would need right now I would think that only need to do is in that vague way you were talking about before Tim is just say case by case there may be different things needed it may need a more thorough peer review it might need it credible third parties it might need fought it's I don't feel like we need to like really nailed down exactly when we hit a critical mass like when something can be accepted for every kind of EEP right now though no I I agree with that but I think what I still think this conversation was useful because if you go if you if you assume that like people are satisfied with this kind of expert

peer review then what that means is if we had that in the process prog pal probably would not have needed an audit because if they could have found somebody to just come on the call and speak that would have been sufficient right and I'm not saying maybe maybe finding someone to come on the call and speak about prog file basically cost them the same as an audit but it was like bob rao from from intel that did the hardware audit but like if prog pal the prog tell people were kind of just friendly with bob and you know got him to come on the Carl and vouch for it then that would have been like a much more lightweight process that probably would have gotten us the same kind of safety assurances as the audit gives but

in a much more efficient way and like Lois point earlier was like the audit was kind of used as a way to kick the political jam down the road so I do think if we have this requirement for like peer review that's loosely defined it makes it then harder for people to push for something above and beyond that for political reasons and that's a good thing well still obviously maintaining that like this is a high-stakes thing and and you want to be serious about it yeah and the part of the I think the nuance who were getting at is that there's a difference between the general requirement of if you're doing something that's outside of the normal purview of core devs

you should have external peers validate and vouch for it like publicly validate and vouch for it with their reputation like involved and but is there also a need or should there be an expectation of packaging what's going on in such a way that the core devs have a pretty good grasp of it because you could have it all be cymbals or whatever and then they're like okay they have a voucher and as long as you can kind of implement it maybe that's fine but should there also be an expectation and this is an unknown for me of making like little literature or resources or

presentations that are designed to get a core dev up to sufficient understanding whatever that means outside of having some peer someone else come in and say yes this is this is what it says it is yeah that's a good point because that's what Rick was asking you about 1559 a couple calls ago right because what they went ahead and did was like a full implementation as a way to get feedback but you know if they had that type of collateral whatever it is they could have requested feedback on that before going into the full implementation the hard part is defining what that is yeah

and in their case it was it was hard to really get a specification without building out a some kind of implementation which is just sort of the nature of how that it's just sort of how that that was how they chose to approach it I guess is a better way to say they chose to approach it with the create and implementation and then make the spec fitting it I'm learning from the implementation as far as what to do with the spec on the core devs call next week just because we also only have two minutes left and it feels like something where it would be valuable to get like I mentioned that if we have anything that

we can quickly cover I'm not sure if we can take care of code of conduct but I saw some conversation from Brent about a ip3 that he was mentioning we got just a sec if James could just give me a quick rundown but if it takes longer than a minute or two we can do it offline or some let's do it offline the harder part is I have I'm on my mobile phone and that means I can't like show or click on things or great great yeah do it later yep is there stuff you wanted to talk and go over what you're thinking about the code of conduct and and things Danno like just sort of get us prepared to

think about it over the next two weeks sure um yeah as for me to see actually proposes since the Yankees and there's something specific that brought that to mind but I think just in general we need to have some sort of explicit standards as to how we treat each other and these environments like all four devs in the Yippies I'm not concerned about the laundry list protected classes that seem like the government code of conduct I think we're all on board with that but I think we need some sort of standards and boundaries for when certain behavior is unacceptable and how we treat it so that's just to think about next time but you know like I said my concern is not

about the protected classes on your list it's about just generally conduct okay some anyone who is like I've had this on my mind - we're seeing the et Cie like what kinda has happened it would be good to get some of this sorted out in advance kind of that was very amorphous and not very said well I'm not sure but I think some chants from the cat or Osteen had done something on code of conduct for in general like how it should be for a participation and and conversation ongoing conversation is it

something similar to that you are talking about on or is it something specifically to how we proceed with the you know protocols have created and something with coders and EAP it's about basically what Charles was talking about and that the scope would be limited to you know I mean the enforcement is the worst we could do is we could kick people out of all cord outs that's probably the worst we could do it has no impact on the actual runtime which is why I think the API key is the right place rather than they all cord s call to discuss a meta issue like this that hasn't been a change I'm sharing a link off of what we have in like the cat hurdles code of conduct that Charles has created sometimes path it would be or if

we would want to add something else and put it place it somewhere else maybe at EAP github or somewhere else people can give a thought on that is that just the Covenant code of the Covenant contributors anyway that the particular code yeah that's the sort thing I'm want to get bubbled up but above just the cat herders to have it applied to the all core devs calls yelling for dev Skinner and the EIP and reprocess elf be good to have them for the EIP it's something with for that yep you repo we have this guilin and we can think over what all else we would want to add specific to EAP process or kora

then we can create that and maybe place it somewhere on EAP IP or the behavior I think we don't need I think we can just take existing code of conduct like this one or the Covenant contributors and just change the scope and the impact you know what's acceptable or not is pretty well understood amongst this things like this we just need to get nailed down what our scope is who the parties we report to and what actions are can be taken in response to violations okay yeah and what comments I would have I feel like this can open like a huge can of worms and and you know so as we think about the scope thinking how can

we like narrow it to be like efficient but kind of reduce the argument surface because I can really see this kind of becoming a massive dumpster fire if it's if it's too broad so just trying to figure out what's what's really essential what do we want to keep and being very clear and concise about that so that it just doesn't kind of devolve into huge yeah month long argument agreed it needs a full meeting to discuss said I don't think we can hash it out here I know there's discussion in the next meeting okay then I believe that is all so the next meeting we plan two weeks from now

yes I'm get anybody who is not on the calendar invite please drop I mean if they they would want to be on the invite please DM me their email address and I would add to the calendar invite as far but I'll do that and then this this was recorded automatically somehow because Hudson is a genius and so then that can be made posted somewhere at some point yeah and I think that's all thanks for everyone for coming around would just

would be in if Paris I should be there I'll be at EDC but not the hackathon sorry sorry I mean if CC CC yeah I'll be in happy there okay so we could meet in a you know in that French cafe yes perfect and in person the idea yeah that no that's a great idea so yeah I keep a IP IP sorry something like that it's still sort of a name to me but that's alright thank you James

thank you yeah good job man thanks thanks