Protocol Community Meeting (March '22)

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

1. 0:08

   welcome everyone to uh the first public protocol meeting my name is josh i'm the devrel team and we have bowen and jane uh here as well to kind of kick things off this is the first time we're doing this so it'll be an iterative process we'll figure it out as we go uh but yeah uh you want to introduce yourself bowen yeah uh here i'm bowen from pagoda and i beat the political research and development here um and yeah it's uh

2. 0:32

   very happy to host the first first ever public project meeting hi everyone i'm jaina proud manager from the experience team at pagoda so we're excited to be here and i'm seeing some familiar faces from our validator community so guys welcome thank you for joining and looking forward to make this even better going forward okay so let's get started uh yeah i just first just want to uh talk about uh what we're going to do today so first we're going to do an intro and then i'll share

3. 1:05

   some protocol updates and then i'll have a lot of time for q a um and yeah i i just want to mention this is the uh the first time we're doing this so i still test things out and uh i didn't prepare a lot of agenda for today so um there will be a lot of time for a free-form discussion um and yeah if you have suggestions on uh how we can improve the format and make it more interactive or more engaging for uh for people who are listening uh feel free to uh let us know uh and

4. 1:45

   uh yeah if you can either talk to me or talk to to to josh about it um so yeah so why are we doing this um i think there are a number of reasons so uh we've got this feedback uh that a lot of the protocol development right now is down kind of in the dark from the community perspective people are not really aware of what we're working on what our priorities are and how to contribute to the protocol development process

5. 2:18

   and we would like to to fix this problem after all this is supposed to be eventually a community-driven project and we should not just talk the talk but also work for work and hopefully this is going to be the first step towards towards that goal uh we may not get there um immediately but nevertheless we want to take it one step at a time and slowly getting more people engaged on in the protocol development process but yeah so regarding this meeting um

6. 2:54

   mainly we're going to to do pretty much three things so one is that we're going to keep everyone posted on what is our current priority and the focus on on the protocol development side and also we would like to have a public forum for on discussions of new portable features eventually we hope to use this meeting as a medium to make decisions and and drive um new features for forward um and also yeah obviously we want to get more community engagement that's why this is a public meeting this is uh now

7. 3:27

   just a team meeting with seeing within pagoda so one caveat of mentioned here is that since this is a public meeting we will not be discussing anything that's security related for example if some security vulnerability would not be discussing them here unless they've already been patched um and if uh yeah if someone mentioned the topic or ask a question that i think may be related to one of those things um i'd

8. 3:56

   probably just say uh yeah it's not a good place to discuss them so just keep that in mind okay so yeah let me just uh uh provide a uh high level update on what we're what we're doing uh what our current focus are um and so on uh so this is the slide from town hall i i just want uh just like reinsert it here uh in case um not everyone watched the town hall uh february town hall um so so yeah just to reiterate uh our current priorities so mostly around

9. 4:34

   optimizations for stability and performance uh so as everyone probably knows that we've had a dramatic growth of usage uh in the past uh several months and it is more and more important that we focus on uh stability and performance so that we can make sure that as the network scales um people still have us most experience of using them and wouldn't be bothered by any of the stability issues and also any of the performance bottleneck that may stop them uh from running the

10. 5:11

    application smooth stream um net to that and i think uh two important things now working on one is making uh aurora run faster uh i think ours still contribute to the uh the most amount of gases among all the applications or all the smart contracts is running on mainnet right now um and we're working with the our team uh on making them run faster i believe there are people from the oral team that's on this call right now

11. 5:41

    so yeah we're working with them um to uh to make both optimizations on the smart contract side as well as candice to the protocol to to improve the performance overall uh and so far um we've already had a 2x performance improvement just on the contract side uh yeah another thing that's another important thing we're working on is faster storage uh making uh optimizing for both storage reads and storage rights as they

12. 6:12

    are an important component uh for most transactions on here uh so on the other hand yeah we would like to increase commuting involvement as i mentioned previously uh so this meeting is one of the one of the initiatives that we're taking to uh to increase the community engagement in the protocol development process and at the same time we're going to try to figure out a process for new protocol feature development

13. 6:44

    that is more commonly driven and not so that not every decision is made by a single team within pagoda uh that will likely take a non-trivial amount of time to get to an ideal state so i think what we will do is that we'll start gradually decentralizing that and in the beginning there may still be um some some form of centralization that is i would say unavoidable at this stage i want to acknowledge that uh but

14. 7:17

    i also want to say expectation that uh eventually the goal is that uh the decision shouldn't be made in a single team or by a single entity okay so that's kind of the high level overview um and yeah so uh so here is i want to share a more detailed view on what we are kind of working on on the protocol side um so this is a screenshot of a durable that i i'm still developing so the the hope is that um this will be a place uh that where we

15. 7:55

    track all the all the protocol features that are uh that are either being developed or in uh in in the ideation stage uh right now it's not it's far from complete um but yeah we're in the process of iterating on it and adding more things there um and right and also right now uh this board is not public and um yeah i'll also work on that to make sure that everyone can get access uh to to viewing the sport

16. 8:28

    uh oh yeah so that aside uh i want to talk about uh what's upcoming in terms of uh protocol features so uh so let's start with um this waiting for release column so this essentially means that uh the implementation is done and it is just in the process of being being rolled out to test the mainnet um so this uh this thing here in christmas max gas burned to 300 turkeys uh what that actually means is that we're

17. 9:02

    increasing the maximum amount of gas that can be attached to a single transaction so previously this limit is 200 targets and we're now increasing it to 300 paragraphs there are many reasons for that but one of the reasons is that we've now observed more and more expensive uh transactions or our applications that want to perform more expensive operations in a single transaction and previously the limit of 200 tera gas sometimes isn't sufficient

18. 9:36

    and there is a need for having more capacity in a single transaction and that's especially important for aurora but also for some other large applications that sir that is also important uh so we've made this change it's already live on testnet and it's going to be live on main that is once we um rather release uh from tesla to mainland uh so that is already basically down and just waiting for waiting to be released

19. 10:06

    on mainnet uh and yeah taking a look at the current focus uh so some of those uh names are could be more descriptive uh so let's just take a look at the uh caching tri nodes um sorry for the typo here uh but what it does is that uh it's essentially an optimization to um to save some costs related to storage operations uh so we'll have this touching china that costs the charge the charges for

20. 10:37

    every single try node touched when when we execute a transaction or a receipt and and the problem is that even though in the process of of executing them uh we would actually catch the china we read um but in reality when we charge for the transaction we still charge them um again even if it has already been touched um so that's essentially a inefficient or rather not ideal way of charging gas because the execution from the point of view of the exclusion that's totally fine and nothing wrong just that the way

21. 11:17

    we charge guys does not really correspond to the amount of computation we do so this will fix that and we actually expect that this would give a pretty significant boost to a lot of transactions on here especially those that involve um storage a lot of storage operations maybe you can save up to 30 percent of the or maybe even more um for the total cost of a foot transaction uh so this out beyond 108 support is

22. 11:47

    essentially um cryptography cost functions um that would provide support uh for for private transactions on here so uh once this is um enabled uh i think we can uh yeah the use cases for private transactions would mostly be unblocked and yeah finally there's this uh feature of splitting unused gas across promises what it does is that today there is a common pattern that let's say you have a function call and then it would generate further function calls through promises

23. 12:30

    then today the common practice among developers that you essentially hard code the amount of gas that you want attached for those promises we're seeing the the function uh the problem with that is that uh well first of all you need to to find those numbers and and uh it's not easy to it's not always easy to to estimate how uh how much gas you need to attach um and also because um the amount of gas a transaction can a receipt can consume also sometimes

24. 13:05

    depends on the state uh so that is actually changing and having a static number there isn't always the best option uh so what this does is that it would allow developers to specify uh like a weight on each uh each of the promises created it was seeing the function call so that it would automatically just distribute uh unused cash within the function call to towards those promises it doesn't it doesn't

25. 13:38

    it doesn't change the kind of current behavior you can still attach a a fixed amount of gas uh just in addition to that you can also specify how you distribute the remaining cash so that that would be add on top of the static amount that you attached to a function call so hopefully that would make it easier for developers to write smart contract and not worry about their smart contracts being broken one

26. 14:02

    day if something changes on the product side or the state gets larger in their small contract yeah so basically that's kind of what we're currently focusing on then there are a bunch of things in the backlog that maybe we can discuss uh if people have questions and there are obviously more that should be added but yeah that's kind of the current state um yeah and i think that's

27. 14:27

    yeah it's all for the presentation part um and ideally there should be a section where we discuss new protocol features but since this is the first meeting i didn't actually um i don't i don't have anything prepared for us to discuss um but if if someone from the audience wants to propose something to discuss um yeah feel free to do that but otherwise yeah happy to answer questions yeah there's uh there's a couple questions uh in here and again seeing

28. 14:59

    that this is the first time we're doing this we'll kind of go through the questions and maybe if it becomes kind of more of a back and forth we can uh bring them on the uh bring it bring bring that individual on the panel and they can discuss it with you the first question we had guest fees are increasing because of increased space and time complexity of operations on aurora as more daps are created with the fees

29. 15:24

    keep will the fees keep increasing literally or will it be exponentially yeah i suggest that you read our economics white paper uh there is a description on how the gas price will change uh when there is congestion yeah essentially it will change exponentially the i mean the short answer is that it will change exponentially to deter people from sending more transactions um and then i think a follow-up question that will um the same individual why distribute gas accordingly uh accord

30. 15:57

    why distribute gas according to weights instead of having the promises detecting how much gas it needs and then pull it from the pool of unused gas good question uh the problem is that promise cannot detect how much gas it needs yeah just because it promises would be executed english in the subsequent blocks so uh before the execution wouldn't actually know how much gas should uh it would need um can the splitting allocation be auto

31. 16:27

    allocated or will i need to specify a percentage for the split uh that's a good question i think by default it's an equal split and if you yeah and and you can but you can also specify how you want to split them um okay um two questions one splitting unused gas across promises sounds awesome do you have an expected timeline for the release yes i think this should be going to the 1.26 uh release which should be on

32. 17:04

    testnet uh sorry let me just stop sharing okay um yeah i was just going to check the calendar oh when the next i think next text net release is um yeah so the next test net release is on march 23rd and the main that should be something like april uh testing that end of march and then maintenance sometime in april yeah yeah it returned yesterday oh cool awesome um what is the difference between max

33. 18:10

    burnt gas and attached gas that's a great question so some so the max burnt gas is essentially the maximum amount of gas you can burn uh in a single execution of a single receipt but the maximum attached cash is the the maximum amount you can attach to a single transaction which may result in multiple receipts uh so previously the maximum amount you can attach is 300 but the maximum it can spend on one

34. 18:43

    single receipt is 200 so there is a discrepancy there which means that previously you can that's 300 potentially you can spend up to 300 but that would be across multiple different receipts thank you um what is the timeline for increasing the number of validator seats ah that's a good question uh so yeah i should mention that because right now we're focusing on um stability and performance of the current

35. 19:14

    system uh the the phase one of sharding um is delayed compared to the previous timeline i think right now we're expecting uh maybe q3 or q4 of this year regarding the wallet team as my observation a lot of people complain that wallet is not a good ux um with issues with 2fa problems doesn't deposit deposit addresses and he planned to invest more resources in the wallet team um i think it's a little uh different yeah

36. 19:53

    yeah so yeah sorry uh this is the public protocol meeting uh for what yeah i don't think it's uh we have the best knowledge to talk about to answer that question yeah that's correct but i should reach out to the wallet team and maybe we can do some wallet uh wallet team um public you know meetings where we can have similar uh things like this where we can have like an ama and q a yeah yeah yeah yeah

37. 20:15

    rachel patrick yes um cool uh that's that's that seems to be all the questions so far if if anybody else has any questions we have a bunch of time left but uh if nobody else has anything else to ask we can uh we can oh another one came in is it possible to attach a fungible token in contract method calls um uh yeah like an attached deposit yes yeah i'm a bit confused as to what the question like maybe means it's like paying as so

38. 20:54

    yeah actually i don't actually know what what other question was supposed to yeah could you elaborate a little bit more on that question um certainly in addition so for for deposit today you can only attach near tokens if that's the question i think that's what it is yeah the answer is no uh yeah damien has a question though he didn't ask in the q a uh yeah so um yeah that's a good question i would say no because uh because uh because of the focus of the team i mean if the if the focus of

39. 21:36

    the team isn't sorry on performance and stability then we would actually push for um phase one of charting then um there wouldn't be a need to do anything in the meantime right so that's not um it's not that we view that is not important thing it's more because we always think stability and performance at this stage is is more important any other questions uh any timeline for the bs z bridge uh

40. 22:12

    it's a question for the aurora team uh actually we do have some one from the oral team here it's to come maybe you can yeah i mean that's not the this is not the best medium but not the best meeting to answer that but like things since the question is already asked we might as well try to find a way to to answer them but yeah in the future for bridge is just not there um okay yeah yes there's marcelo he should we invite marcelo to give like uh

41. 22:44

    yeah longer longer longer yeah yeah let me know more so if you want to come up on on the panel and give a give a longer update on that um okay more more questions coming in um how was the 300 tear gas limit decided on was there some benchmarking involved how was 300 targets limited yeah so uh the the short answer is that it's it's decided somewhat arbitrarily uh it's not entirely arbitrary um the reason is that

42. 23:17

    we don't want to change the the maximum um gas at uh attack allowed to be attached to a single transaction and that is 300 targets um so yeah because we because we don't want to make the the change very invasive we just wanted to draw something quickly so we just changed the max gas burnt to be 300 so that um yeah nothing else needs to be changed uh and and yeah there's another uh concern which is like if we uh increase the number even further let's say we also

43. 23:48

    obviously don't also need to increase the maximum max gas attached then um then there is a concern that you know because this essentially only goes one way you it's pretty much you cannot lower it again because then you potentially break contracts uh the concern is that if there is some some like undercharging that people that potentially attackers can exploit uh then if we cannot lower it then other than that means that uh the wiggle room it becomes smaller and potentially the

44. 24:20

    uh which basically that means we cannot lower the the block gas limit below below what's allowed for a single transaction otherwise there would be transactions that cannot be fitting in one block so there are some concerns there that's why we didn't like very aggressively increase this number but yeah there's some idea on like you know how to increase the number even further uh to to make the single trans

45. 24:43

    transaction capacity even larger cool thanks marcelo for the answer oh yeah no totally um is how will the stability um be affected if we increase the validator seats uh yeah good question i don't think the step it depends on what you mean by stability if you mean the kind of security of the network uh then this actually not affected because in phase 1 of sharding block block producers would still validate all the shards and the trunk only producer wouldn't actually sign in blocks so the security of the network

46. 25:23

    should pretty much be the same if you're talking about stability otherwise um i mean yeah a good question but we will make sure that the stability wouldn't degrade uh before we revolve the change are there any plans to encourage people to redelegate their tokens to other validators to encourage better decentralization uh yes but that that is outside of the protocol i think uh the liquid staking initiative is definitely one of them uh

47. 25:55

    essentially the liquid sticking pool like uh would rebalance the uh uh the stakes among uh among our auditors and would i think it would do something that would um essentially stake to validator with lower stake um and so on so uh but yeah that's outside of the i was the other side of the scope of this uh some protocols have starting to support running a node as a validator in docker are we going to plan on move to docker at all for the next movement for

48. 26:27

    validator nodes um i'm a bit confused by the question to be honest because i believe today we also support docker there is a document like compose you can you can use to run on your core and we even publish a docker image on our docker hub so yeah not sure what the question is about maybe if you're going to clarify a bit more cool um looks like we're at the end of our questions again

49. 27:00

    if anyone has anything else to add or have any questions um to add please please drop them in give it a moment uh are there any plans to have synchronous calls in order to simplify cross-contract calls uh do you want to touch on that topic at all yeah uh sure yeah um i would say there isn't anything super concrete right now as a proposal i think i think uh there was a forum post that i think was written by max that

50. 27:38

    summarizes the uh situation and also the approaches we're taking maybe we should how about let me just quickly try to find us and send a link here yeah there was some discussion about having a synchronous shard yeah uh should i type should i i guess i'll just post yeah just post in the chat there you go perfect and then everyone everyone should be able to see it oh it's to hosts and panelists

51. 28:20

    i'll i'll read i'll repost it no problem cool uh all right any other any other questions comments anybody has well yeah no i think this has been a very oh can you explain the function of pagoda to a five-year-old eli5 pagoda well yeah that might be yeah yeah this is not the best meaning i would say like watch our tongue recording i think julie did a good job of explaining what pagoda is there um and yeah yeah if you go to our youtube videos

52. 29:06

    yeah absolutely and then feel free to ping me on discord um as well um i can drop my discord link in here or just ask in like the the community channels for some clarifying questions if whatever answers you you don't have uh or you didn't get any you know answers out of the presentation that we did give but yeah this the recording uh tozilla's have been asking uh this is being recorded right now it's not live

53. 29:31

    stream to youtube yet um but it will be uploaded to our youtube channel after this call so cool any any other any other questions all right well um if we don't have any other questions comments oh are there any good documentation out there explaining how the virtual machine blockchain interacts uh i think there actually is not me yeah it depends on like uh uh how technical you want that to be i

54. 30:13

    think there is one um actually i need to wait is it um i'm sorry just trying to find the poor request that actually dodge that explanation yeah i think there's like a diagram uh i can try to get no there is a very detailed um explanation i'm just i'm not really not having trouble finding it um i think it's in your core so sorry one moment no problem yeah on the on the devrel side um documentation is something that we're

55. 31:05

    actively working on to consolidate it's definitely been something that we are recognizing that we have kind of our documentation everywhere and we're looking to create little areas that are more discoverable um and yeah working with the foundation on the wiki end so um yeah we have uh our new nodes documentation or we have developer documentation and then like a wiki that'll be general more high level for end users but yeah looking to ship lots of changes within our documentation

56. 31:38

    and reorganization within the coming weeks so stay tuned and if you have any questions or suggestions about that please reach out to us in discord um as well yeah shameless plug for the devrel office hours um if you have any questions about how to develop on near even like protocol questions we might not necessarily have the answer but we can definitely reach out to the team and help field or answer any of those questions that you may have um feel free to drop in we do it every day i'm actually doing it twice a day

57. 32:12

    now once in the morning uh well it depends on what time zone you're in so one that's more um uh caters towards the european africa and asian time zones and then one that's more for the americas um i will type out the there's a link from bowen on uh your question but i'll type in office hours link um feel free to click on that link you can see what times that we have our office hours twice a day and you can there's a link to excuse me add it to your calendar

58. 32:45

    and yeah that's in discord under the office hours channel in the development section so anybody on call that would like to ask any further questions or comments um any yeah feedback from today's session uh greatly appreciated uh but yeah i haven't seen any other questions come in so maybe we could wrap it up a little early um uh yeah this is this has been a great great start i i think um really really enjoy um yeah i think like yeah you're welcome um yeah i am too

59. 33:18

    looking forward to these sessions each month um it's cool to get to pick the core team's brain a little bit and ask questions uh as they come up and then uh yeah maybe we can gather some feedback from you guys uh maybe i'll send an email out to all those people that registered and see kind of you know how you'd like to see this meeting uh shape up as we move forward um but yeah any parting words bone

60. 33:41

    uh yes thanks everyone for joining and yeah we will definitely iterate on the on the process iterate down the format of the meeting and know that this is definitely not perfect the first time we're doing it so but yeah that's you know we have we have to start somewhere uh and yeah i hope that in the future it'd be uh more engaging and there will be more uh discussions uh in this meeting but yeah

61. 34:04

    um thanks josh for uh for helping hosting uh the meeting and and also help with everything around no problem awesome thanks everyone have a good day you