Protocol Working Group Meeting – March 16, 2023

Captions as YouTube publishes them, usually machine-generated, so product and people names can be misheard.

1. 0:00

   okay Welcome to our fourth protocol working group called believe it or not and we're going to be recording it and posting it later uh so just everything is going to be public uh very excited to kick off this call just on the agenda today we're gonna do an update on developer governance uh do some introductions as we have a few people joining us and then we'll get into the main discussion we

2. 0:28

   have two naps uh 399 flat storage and 455 parameter compute costs there are links to those apps on the invite in case you want to look more closely and then we will open it up to q a and talk about next steps on how to get involved so just a quick update on developer governance so over the last few months we've been engaging with the community over this initiative which is really about bringing people and organizations together to help build a decentralized developer ecosystem

3. 0:59

   and empowering anyone to contribute from idea to code and we've been ducking in most of the developer community activities into four areas so ideation implementation review and support and I just wanted to quickly highlight them to help contextualize our meeting for today so starting with ideation this is about sharing and exploring ideas with other people in the ecosystem so anyone can

4. 1:26

   submit an idea like improve in the protocol or standards and it's important to have a place for people to freely Express and themselves and explore ideas um in a welcoming and permissionless way and one way to do that is in the developer governance gigs board and this is a new uh platform on your social for anyone to share their ideas to get feedback and support from the community and it's still in the early days but there's already a lot of activity on there and anyone who's

5. 1:55

   interested can start posting their idea today now maybe you have an early seed for an idea or a topic that you want to explore with others so another great way to get involved is to join a community group and community group is a group of very passionate self-organized people who are interested in a specific topic such as zero knowledge wallet standards Etc and they offer a great opportunity for people to come together uh and just discuss ideas and and concerns and

6. 2:26

   things they want to uh help improve and many of these groups have already formed and have dedicated communication challenge like telegram as well as regular meetings and you're welcome to join any of those groups or to start from yourself you can find them on the developer governance website and we're also going to be publishing a blog post probably this week that will highlight some recent learnings and insights from

7. 2:50

   a lot of these groups so what happened if you already have submitted an idea and now you want to turn it into code this is where implementation comes in you can also use the dev gov gigs board to submit implementation solutions to your ideas or even if you have an alternative solution in mind for somebody else's idea you can propose that as well and this is just a great way to engage Community interests and supports and to

8. 3:16

   get a variety of perspectives which really can help strengthen our Collective problem solving and innovation now if you have an idea that you want to get integrated across the ecosystem like a new contract standards or protocol Improvement or if you are seeking funding support you will need to get a more official validation from a recognized group of experts and you can do that by submitting an app a near

9. 3:41

   enhancement proposal and this is a rigorous process that is similar to other prominent open source projects like rust and it enables anyone to submit their proposals for review and you know this is actually what we're talking about today on call as and if those submitted so what happens once you submit a nip and how does it move forward so this is where the review comes in which involves working groups like the

10. 4:08

    wallet standards working group that we have here today and basically a working group is a selected Committee of subject matter experts that help review proposals in a fairly untimely Manner and oversee decisions and there are multiple working groups each one focusing on a different ecosystem need we've already kicked off four of them to help bootstrap this process and we've been iterating on it for the last few

11. 4:32

    months to kind of understand what works and what doesn't and we hope to introduce more groups in the future and Empower more people in the community to to run them so before we get into the actual call today I just wanted to highlight the roles that are involved in the next process as well as the the stages so uh on the rolls front we have the author this is the person who writes the proposal and

12. 4:56

    presents it to the working group there can also be a champion who takes that over in case the original author is unable to uh continue there's the moderator the person who facilitates the process and helps validate that the proposal meets the guidelines and progresses forward there's the reviewer the subject matter expert the person who reviews the technical feasibility of the proposal and then the approver this is the working group member who helped make

13. 5:20

    the final decision and in terms of the actual illustration of this process there are three stages draft review and voting and two outcomes approved and rejected and all of those roles help move the next through these stages and towards the outcomes and most of this happens asynchronously on on GitHub on the nip itself but when we do get to the voting stage each working group has a public meeting like the one we're having today to discuss the net with the author and to formalize the

14. 5:51

    decision and this is what I meant to show when I was talking about that we're here to to vote and have make a decision right so uh uh lastly I just wanted to highlight what happens when an app actually gets approved uh and how can the implementer seek funding and that leads us to the last area which is support So once the working group completes their technical due diligence and approves their proposal they can use the developer governance gigs board to post an official attestation and this

15. 6:23

    tells any potential sponsors that there was a formal review done by recognized experts in the ecosystem and that any organizations or individuals can choose to sponsor that project so that was a very quick uh overview of the latest parts to developer governance and as I mentioned there there are more blog posts come in that will help kind of go into further detail and you can already check uh one that explains it on the developer governance website and with that I am super excited to jump

16. 6:54

    into call today and I'd like to do some introductions uh and start with the working group members so Bowen I will hand it over to you to introduce yourself uh yeah excited to discuss this two uh nearly submitting any piece our Metro do you want to connect yes sure uh hi everyone I'm Marcelo I've been involved uh with a new community for a long time uh at early days I was a member of the corps developing a team and after that I was working at Aurora right now I'm supporting the ecosystem

17. 7:38

    mostly by helping move forward protocol uh like Maps related to the protocol um yep happy to help here hi I'm Martin I'm a tech lead for a protocol and yes very excited to get two additional naps and awesome improvements the protocol so let's see nice and I also wanted to represent Alex who's technically a part of the working group as well but I was not able to make it but that's okay because we need basically two-thirds uh majority and look we have we have enough participants

18. 8:12

    so that's that's great all right next I'd love to introduce the net moderators that were involved starting with Vlad hey everybody loves here you can find me um where is Media says at the handle roll and uh I joined here four years ago and now I'm happy to help to facilitate the developer governance needs and orchestrate all the working groups and Empower our community together with Orient Max

19. 8:44

    um worry back to you hello again yes and very excited to to continue to to bootstrap and iterate so yeah now I'd like to hand it over to the various authors that were involved with these two naps and I think we have Alex on the call yes sure so hey my name is Alex I work in the protocol team in forgotten and the last year I was involved into Vegas projects going into storage and yeah

20. 9:15

    storage and here I will represents the neb Philly 99 which is the meal is not here and yeah I'm excited to participate in this discussion nice yeah and I think we have basically there are two authors for each one of the next and we have one author representing so the other one is uh Jacob yep hi my name is Jacob I'm uh engineer at uh runtime team at pakoda been hacking around on your protocol for a bit over a year now um yeah and Andre is also in my team uh

21. 10:00

    joined a few months ago he did most of the work here I kind of gave the initial input and that was already stolen from other Engineers the idea itself so really this is kind of a teamwork that we are now presenting here um and I'm today mostly representing Android's work very cool well thank you for for all the work that all of you have invested into these naps and uh we're very excited to to finally uh get to a decision stage so

22. 10:32

    yeah I just wanted to also call out that if anyone has any questions at any point feel free to use whatever Zoom features you like that are on this call to to communicate we'd love to hear your thoughts as we go along or questions and we'll make sure to address them or get to them also at the end if we need to all right so let's get to the main part the discussion of the naps and we're just going to go uh you know through net

23. 10:57

    number like the first one and talk through it make a decision and then we'll move on to the other ones so let's go ahead in sequential order starting with net 399 flat storage and I would love to invite Alex to uh give a high level summary of the nap and why it's important and and then we'll go into the the benefits and concerns on the next slides so hey again Alex yeah sure I'm happy to present that so in this introduction I assume that

24. 11:29

    you already know what like a blockchain state is and to say it I like At first I say like why um what is this issue we're trying to solve like our state is currently a prefix 3 or try and for scope of this call it means that to get a single value pair you need to iterate over this try Notes From The Roots so the leaf nodes and in total that there can be many notes like in theory up to key length multiplied by two on practice like we see numbers around like 10 20.

25. 12:14

    and the biggest problem is that these are random leads and there is simply no way to like make these kids not random so and I consider this as a performance issue so now instead of just a single storage sheet you need to do like around let's say 10 leads and here for storage comes So the plan is to replace these 10 leads with one leads and this is simply resolved by storing like a flattens key value

26. 12:48

    mapping so instead of looking into try you need to just look into the mapping and so one thing is performance isn't another is much to guess my guess is uh like major pain points for example developers and one of the problems is half incredibility of the costs and it's worth noting that we currently have so-called costs attaching dry nodes which is also listed in the net and this is simply the cost of touching a single

27. 13:30

    node and this cost makes sense we need to have it at least for now because leading each node is some work but currently this course is really hard to predict as well because it really depends on the cliche so some contact developers could be really confused by the cost changing and so they will have challenges you know these costs so instead of that we will just charge them from for one week which isn't obvious benefits

28. 14:07

    so that's it yeah and some upward storage aims to improve performance of leads and also improve instability of Facebook awesome thank you Alex for that overview and yeah as I mentioned there was a lot of I think discussion uh on the net itself we got a you know subject matter experts who who reviewed it as well as the working group members and I'd love to invite one of the worker members to summarize some of the benefits and concerns that were raised during those

29. 14:42

    discussions I can represent it thank you right the tldr is like flat storage gives us like a secondary index over our data so as Alex was saying try structures optimized for an easy way to verify the correctness but not necessarily performance and flat storage is kind of the opposite right here we're talking about efficient access and more deterministic access uh this helps both with as I said with the estimation of the gas cost but that's like a small fraction the bigger fraction is it basically improves

30. 15:20

    overall performance especially for larger contracts that have deeper tries moreover having such secondary index will allow us to do further optimizations in the system in the future because suddenly we can do things like getting the car like suddenly getting the current state of the account is a cheap operation this is just a linear scan over kind of scan over the data rather than a bunch of random lookups that we had in the try

31. 15:49

    so these frequency these three things combined means that yeah password has a bunch of cool benefits but next slide please usually when there are benefits there are also some concerns the big concern here is that it's not easy flat storage adds not only implementation complexity but it's it also introduces a bunch of small risks because suddenly we have two copies of the same data there's always a problem if something goes wrong you know you might not know which one which one is the correct one it also increases

32. 16:30

    these are two copies also increases the storage requirements for the nodes and memory requirements because for these are more technical details but in order to efficiently do these operations we do have to store more data we have to store things called deltas in the in the memory in order to be able to handle Forks in the proper way um this this part that we that is proposed in the snap is actually slowing down

33. 17:01

    over the re modify write workload but because this is just a part of the longer plan the hope is that the future naps will actually then drastically improve their read modifier right but again we didn't I mean that the team didn't want to combine too many things in one in one request because it would be like very complex so you can think about this snap this flat storage as the first step in

34. 17:23

    around three to four step process that would really bring us a lot of benefits when it comes to this Catalyst performance awesome and next slide I have no idea what's on the next slide but let's do the next slide yeah well the next the next slide is uh now we're actually coming to the decision so I'd love to hear from each working group member their opinion and uh how they're leaning and uh yeah basically get to the voting stage

35. 17:51

    so who wants to go uh yeah I said I mentioned on GitHub uh I don't even tours approving this uh uh map uh yeah I think as Martian already summarized uh it's a greatly helped improving the storage uh read performance and also uh there are other benefits like making the customer predictable and making other things related to storage easier to do um yeah I mean there are some downsides especially with the implementation complexity and also um potentially effect on the right

36. 18:24

    performance but I think yeah those are already taken into consideration and then we have ideas on how to address them in the future okay I can chime in now so uh in general uh this being this is a topic that it's been discussed for a really long time now and we kind of I think we already understand uh pros and cons uh one thing I really like about this proposal the it's kind of what

37. 18:56

    Martin was mentioning in the in the cons is that all this complexity it's actually how you need to implement it which which of course we need to implement it in some way but the nice part is that it's not part of the protocol uh as opposed to for example the miracle tree where we need to have a really nice spec I really will spec about how the miracle 2 should be accessed and how the cost should be

38. 19:20

    managed uh these flood storage can allow other clients to implement it in a different way potentially different way so that's uh also a benefit so potentially the the code can be simplified uh and yes in general I'm also leaning towards approving the snap awesome uh I guess Marcin did you wanna I think you're kind of already spoke about his do you want to say your indication yes I am consistent with what I said five minutes ago yes I I also lean towards approving this

39. 20:02

    cool just to be super explicit yes so it sounds like we have a full majority uh vote to to approve this snap so awesome congrats and I think we can now briefly talk about what does that mean in terms of The Next Step so yeah what what are the the next steps that we need to outline here for the snap Alex maybe you have some insights on there because I I just uh tried to learn the status there uh and it seems like the implementation is already

40. 20:46

    merged long time ago and it was very well tested on staging but can you speak to it maybe in more detail that's true so implementation is well tested the stabilization pair is not ready yet but it will be really soon we were mostly like applying testing by maintaining construction at the same time we try not being part of protocol but consistently like chicken correctness of flip storage and checks fast

41. 21:18

    so as for other important point is that the storage kind of relies on the next step we want to discuss because the next step addresses some of security concerns with the storage opportunity links and as for next steps I also can mention that like there are other ideas we Implement that depends on like of the company goals but like to be transparent they say that one of the years is applying it to statesync meaning that food storage would allow to

42. 21:59

    sink to create the states to get the States from other nodes faster then the countries another idea is to apply the same optimization to rights but this is on the research stage that's two major big steps to come into my mind uh one question for you Alex with regard to the overall testing you have done try to migrate to flood storage in magnet like in a testing framework so do you know how much time will it take

43. 22:35

    approximately yeah we know how much time the migration takes so it's like let's say 10 hours for PC notes one minute 15 hours of the standard personnel and two hours work hours two days for archival news and the complexity is key because to create food storage we actually have to try it the whole try which is slow yeah but yeah I will prepared for that church okay awesome any other questions that does mean that the validators node

44. 23:16

    will need to have like two notes and swap between them in order to see to to apply the migration and then proceed execution or is it happening in the in the background awesome question um that's what we cared about a lot and you don't have to stop your knowledge actually the immigration can happen in background the only like requirements we have is that like we need to have a new binary let's say a week before the actual protocol rate happens the thing is that liquid protocol

45. 23:57

    upgrade happens then you need to have food storage but we usually give at least a week or later stop rates and in that sense it's just enough to download similarly launch it and they will Define automatically awesome thank you I will also mention that still immigration is quite heavy work so like it's again like a bit more details it's eight iron threads like because of the leading data forms and if it takes a lot of time like like I said config options

46. 24:33

    foreign thank you I'm ready to move to the next nap yeah we'll follow up async on on the next steps and uh yeah for now let's move on to net 455 uh and I'd love to invite Jacob to speak to it yes so basically this snap is about uh problem that is mostly for protocol Engineers that users may not care that much about um that is at least my impression here so let's take a step back and think about

47. 25:15

    what gas is and why this can be a problem for protocol engineers so developers have to attach specific amounts of gas when they make a function call and when they make it from within a function called the mechanatic function called call it a cross-contract call and they have to be quite precise in how much gas they attach there if they get the estimation wrong things will fail and that it can be quite problematic and in like in the worst case it's a security flow

48. 25:50

    uh where a user can enforce that part of your execution fails while other parts work and when people upload their codes to the blockchain and the contract is installed these numbers how much gas is attached is fixed we have some new tools that can make it relative but some degree it's always fixed and you can't change it anymore afterwards you can just attach more in total but each of these function calls needs to have some amount of gas attached

49. 26:25

    now what we have to do as a protocol engineer sometimes is change specific gas values when we do that all these estimations from contract developers would kind of fall apart and their stuff starts failing we really don't want to do that and it's kind of our assumption that we can't break anything so in other words we can't change any parameters unless we can prove that nothing will be broken by it and that sometimes involves going out and finding all the developers making

50. 26:59

    sure they upgrade your code and only then we can change these values foreign so why do we even need to change these values well that's because we fill our blocks or to be more precise our chunks we fill them based on how much gas each function call or other receipt Burns we take this as a proxy for how long it and how much time it takes to execute it and we use it to guarantee that we can

51. 27:30

    finish each chunk processing within a second so that we can have our block time consistently among all shards sometimes we change an implementation and that makes a certain part slower it can even be that on average you make things faster but there's weird cases where something else gets slower and then we kind of have to change these parameters sometimes we find that our implementation has weird Quantum cases where it's slower than we originally anticipate and that leads to undercharging attacks

52. 28:06

    which could be a whole topic to go into I want to avoid it here it's all written out kind of in the in the map itself but that is the motivation right like we find honor charging problems we have to adjust it compute costs pretty simple we just say we don't fill chunks by the amount of gas but by some other number that other number we can change while keeping the gas number the same so the guess number is what people care like the developers care about

53. 28:40

    that's what they estimated and if we don't touch that nothing will break but we will touch the compute cost as to how much we'll fill into the the chunk and that allows us to fight under charging it allows us to change these estimations for how long it takes to do different things it also allows us to do faster iteration on new features try them out um and the flat storage map that we've just approved is one of these examples where we need to change a couple of parameters

54. 29:13

    where like read modify flow becomes slower but in most cases it's better but to be on the safe side one we have to increase the cost of certain parameters again we don't want to break stuff so compute costs helps us to change the parameters we care about in terms of Chunk filling without breaking anything um and that is pretty much what we are trying to do here in terms of implementation

55. 29:44

    you can just think of each gas parameter that we have today it's like a list of close to like 100 different small parameters of that we can charge we kind of duplicate those and say we have a gask number and we have a compute cost number and by default these are the same but we can decide specifically to change some of the let's say compute costs when we need it and that allows us to fix under charging issues quickly it allows us to

56. 30:17

    have the situations of like flat storage can ship it without the fear of breaking anything and yeah there's a whole bunch of space of other Solutions we considered it's all in the nap and there is also a whole bunch of specific details what does it mean exactly where is which of these numbers used again everything is in the net so I will kind of close it here and answer any questions if you have them foreign thank you Jacob that was a really great

57. 30:52

    summary uh and really helped paint a good picture yeah and I think just uh like we did for the other one I guess if anyone is able to sort of Highly speak if there's anything else to add to the to the benefits and concerns that were raised so anyone from the working group um yep let me try to invite this um so uh mayor benefit I see from this uh from this proposal is that it allows to make protocol changes without breaking a developer experience which is uh which

58. 31:31

    has been so far the main uh drawback to to make this change uh faster and sometimes at some points this might be critical issues as in uh issues that bring with us to the system this has happened uh on the past I remember with the delete action um so yeah this is like a very neat idea to get rid of this issue foreign wants to elaborate there you can go ahead otherwise I can try to yeah I mean the tldr I would say is like this is the necessary evil

59. 32:13

    yeah I would love not to have to do it I would love to have like a perfect estimation of gas cost and we don't have to think about having any kind of parameters we just have one value and it's done that's because we're not living in a perfect space that's why you have to do it and as it was nicely written in the benefits this is the least like the simplest thing that the simplest thing that we can do given the time constraints and this will definitely unblock us and give us at least a way out

60. 32:44

    a reasonable way out when we see that we have a big under charging problem in our in our system so one one thing that worth highlighting it's why like the evil part of your comment uh Martian which basically like why wasn't it this done from the very beginning like well what is the main issue that this is introducing and basically it's like we are kind of open about like if you want to deduct the system we're kind of open our blue charted methods you can call to DDOS the system cheaper right

61. 33:21

    uh as we have discussed in the in the GitHub discussion um it actually doesn't uh it's not a big a very big issue like you might do the system like twice cheaper maybe not too much but we need to be conscious about that right like there shouldn't there shouldn't be a high disparity between gas cost and compute cost in general and well ideally we want them to be the same at some point right uh the net mentions uh that this party should be temporary I

62. 33:54

    guess well yeah I don't see this being temporary in general but anyway uh it's a good idea if we tried this to be the goal okay uh do we want to get to the voting or do you want to highlight any other concerns oh yeah I think I think a concern number two uh-ish maybe I wrote that yeah I I um I think this is a good first step towards uh addressing this problem but I I would yeah I would like to see in the future uh if we can do this adjustment

63. 34:32

    Dynamic racer that we're battery equipped to deal with some of the emergency cases foreign yeah also uh I want to highlight that I was mentioning the discussion that as the pr mentions that this this party should be temporary in general whoever it's proposing this uh difference between computer and gas costs should be also making some sort of proposal or at least uh having a some way to track

64. 35:09

    when is the gas cost B uh going to be the same as the compute costs like which are the mechanism that will be done to make this happen and if this is not going to happen well let's have it explicitly somewhere so I think this snap doesn't propose a way to track those maybe we need to have have the mass issues in in like in the net repository or in this same map somewhere we need to track like it's not about the disparity because this party will be like like near reference right

65. 35:43

    it's about how this will be addressed that's what I mean maybe they can be tracking the same full request or issue that it's creating the disparity itself so it's also acceptable time yeah we can write down some specific strategy one we want here I I think we want to track everything on GitHub with issues and make sure that like as long as these are different we have an issue that tracks that and maybe some tracking

66. 36:14

    issues that summarizes all of them but I think that's something we can kind of discuss offline as a follow-up that it shouldn't be difficult to come up with a useful strategy here but definitely a good input thank you yeah I would suggest to use the same map just update like create a section with the current values for now and then submit pull requests to update it with like changelog we actually have this

67. 36:47

    laid out in the uh map001 as already already mentioned we did a lot of changes there and one of them is that maps are not set in stone and we accept the reality of the world and we we will consider to add those map extensions and approve them as as a regular nap process going forward I think it's the best way for for the ecosystem to agree upon to keep track of inside the map it's all but feel free to suggest some other uh approach in the nap so you know the

68. 37:27

    future readers will know how to how to operate I'm also it might be worth mentioning that are we are planning to use this net for flat storage right uh yes so technically it's not necessary for flat storage but based on performance measurements on our specific implementation of the client we know that it's safer to use this the way we want we plan to ship the first iteration of flat storage okay

69. 38:09

    yeah yeah just one one thing that I remember from my comments that it's not enumerated in this drawbacks is the way we present this to the consumers in this case probably Developers might be also users uh to extend uh but I remember while developing on near uh like counting how many transactions I can batch together like uh in a single transaction and I was like counting gas and all that stuff but in this case I

70. 38:40

    will care more about compute gas than about uh or not not really no you should only care about casts because the 300 Terra gas limit is computed using the gas cost not the compute cost and that's all developers should care about so this is not leaking into developers that's what you're saying yes that was our my main priority here and so the question you also ask there like in the gas profile that you can

71. 39:10

    clearly using the Json or PC but we'll just not show compute costs at all it's just the guess numbers developers shouldn't even know that this exists the only place where it's visible is if you imagine that The Shard is completely full it's like fully loaded and suddenly the theoretical capacity of a Shard would be to say 900 Tera gas and not one better gas right because we would like because of these multiplications we would be like filling The Shard more right okay I

72. 39:40

    see and and that has actually two implications one of them is the obvious one you limit the throughput it can get less through and there is a more subtle one that is mentioned in the nap for this kind of easy to overeat when we adjust the gas price we look at if you're above 50 of total usage across all chunks then we have linear interpolation of like how much we increase it that will also be based on compute cost

73. 40:08

    not on gas costs so if you try to track like what is the how does the gas price change based on usage that would change I don't think developers care about that but that is an unobservable effect okay for all right any other thoughts or we're ready to move to building you can go to both India okay awesome so same drill as before who would like to kick it off yeah I can go first I in general as

74. 40:52

    Mercy put it I think this is a necessary evil I really like that way of putting this uh yeah I'll link towards uh approving this I think this will simplify a lot of process for upgrading uh near um yeah um lean towards improving it yep same here we've got to do what we got to do so yes I also approved this snap and I rethink the outros for like writing it down and you know naps like these are not easy because we know we're kind of putting a little bit of a dent in our like shiny

75. 41:32

    protocol uh but hey that's where reality meets the theory so yeah same yeah I also link towards approving the uh the nap uh I think yes uh already mentioned yeah this is uh uh what we need to have um and yeah I in support of that and let's hope we will have to use it as few times as possible yeah awesome well sounds like Anonymous uh consensus so that's great uh in terms of next steps I know we already discussed making some some updates that Jacob was

76. 42:12

    going to take but are there any other ones that we want to call out foreign I guess I can shortly comment on like the implementation is roughly ready but it's not quite as as far as flat storage in the sense that we've already merged the code into Master because we would have to change configuration files and whatnot we we have the PRS but we haven't merged yet so kind of getting that into my master room is also something we have to do

77. 42:46

    but also it's the much much smaller change it's almost no change in in terms of code right so that's something that's definitely needs to happen and then actually defining the numbers for like what we need for flat storage and coming up with this process all of this is something I'm looking forward to kind of see it happen the first time okay I have a question to both of the authors of the map like uh I if I remember correctly there was a rush to

78. 43:23

    uh approve these two naps before the next release of near core going to test net um is it still on track or if the decision was reconsidered since the time can you update the community on that you want to answer Alex or should I or I mean anyone yeah so we had actually a big discussion on that yesterday and I think the decision where that would not rush features like even if there is uh like of communities so we better uh like

79. 44:08

    carefully like consider the codes like if we are going to stabilize I mean I'm not sure how it sounds but I want to take a look at that like of this snapcode like for making a decision and uh if I don't feel confident I would not include it in the next release it's still not uh there is no final final decision in regards to that it's it's still conditioned on some implementation details uh yeah okay I'm confident in the flat

80. 44:48

    storage code that you can prepare stabilization the release date and I also want to see compute costs yeah I think my main takeaway here is these are two really great features uh and especially in combination these are kind of something we want but there's no very urgent dependency on it by anyone in the community so there is no harm in pushing it out one release the future so it is kind of more relaxed for everyone if we stabilize it right after the next

81. 45:31

    release is cut then we have the full five week cycle time of like testing it before it goes to test net and then hopefully it's just a reduced risk I would vote for that to be honest but it's your call guys yeah I mean okay also to be mortal spread I feel kind of sorry for uh like Russian this call I think having this shouldn't happen and I will be like more relaxed in the future about that

82. 46:09

    all right well we made it and it seems like the working group still survived and is still together with us on the call which means like we can keep keep rolling uh please don't drop other balls this week uh we only have Friday we don't want to do it on Friday on the positive side look the important thing is it also shows and when needed we can actually make sure that net process is fast enough

83. 46:42

    it's not considered like oh no we have to do an app it's going to take like five months and plus no it's not the case it should be also a good signal to other developers at near ecosystem that if needed we can move quick yeah that's true yeah thanks for bringing it up yeah that's actually probably a good segue for Q a if anyone has any questions or comments I do want to have a quick cyber with the working group just talk about a couple of uh housekeeping stuff but are there

84. 47:13

    any questions related to to the nips or the call that we just had you know the Q a but I I wanted to just shout out to the to the working group who were very flexible and helped us to go through the review and uh the um it's even scheduling the the the the call because it's the the main reason why we actually take more conservative uh and relaxed in terms of the timeline is that we are trying to accommodate the uh all sorts of requirements of people living in different time zones and

85. 47:55

    taking vacations and different needs that all of us need and this time was uh really fast and snappy and thanks everybody for cooperation uh thanks authors for for the maps being written in such a beautiful way and also interacting with the all the uh reviewers and getting updates quick um yeah and one thing one extra comment maybe more to the community that uh we are looking for more people to

86. 48:34

    contribute to collaborate on behalf of the developer Gardens in different roles be it um a contributor of the code or author of the nap subject matter expert expertise required from you by the working group members looking like all those roles are getting um through the time you can get immediately promoted to the working group member but uh it starts uh is as all of us started with a active contribution to the code to the ideas

87. 49:14

    and developer Gardens now has more clarity I hope um with how to ideate how to request planning how to um contribute to the to the nearest system that I wanted to invite you to read the blog posts on near devgov.org website and find all the necessary information and ask for more yeah and the Diggs board is a beautiful place too also highlight your ideas maybe something just you know came to your mind and you want to

88. 49:53

    to share it with others and let them implement it if you don't have time all sorts of opportunities are there and yeah yeah I think that was well said and I totally agree and thank you to everyone involved and it's exciting to see that we can move quickly like y'all are saying so awesome well we're gonna follow up online on the two naps and excited to to get them approved and uh thank you everyone for joining and take care

89. 50:27

    thank you