WEBVTT

1
00:10:47.280 --> 00:10:50.069
Ben Adams: Please. I like it.

2
00:11:10.300 --> 00:11:11.330
Barnabas: Hello?

3
00:11:16.980 --> 00:11:17.840
Fredrik: Hello.

4
00:11:19.270 --> 00:11:20.450
jtraglia: Good morning.

5
00:11:21.630 --> 00:11:22.310
Jordan: Hello.

6
00:13:52.360 --> 00:13:53.250
saulius: Hello.

7
00:15:06.770 --> 00:15:11.339
danceratopz: Good morning, everyone. We'll just wait a bit to allow more people to join the call.

8
00:15:20.080 --> 00:15:22.420
danceratopz: Can you let me know if you can hear me alright?

9
00:15:24.130 --> 00:15:24.880
Marius Van Der Wijden (M): Yes.

10
00:15:27.340 --> 00:15:28.780
danceratopz: Perfect, thank you very much.

11
00:17:13.589 --> 00:17:15.069
danceratopz: Alright, nice.

12
00:17:18.800 --> 00:17:19.780
danceratopz: How's that?

13
00:17:21.240 --> 00:17:24.030
danceratopz: Hang on a minute, let me just check signs again.

14
00:17:27.099 --> 00:17:29.020
danceratopz: Okay, I think we're good to go.

15
00:17:30.210 --> 00:17:32.880
danceratopz: Alright.

16
00:17:34.590 --> 00:17:42.050
danceratopz: So, welcome everyone to ACDT number 103. Today is the 16th of November.

17
00:17:42.250 --> 00:17:49.069
danceratopz: 2026. I've just dropped, the agenda, or the issue for the call in the chat.

18
00:17:49.430 --> 00:17:58.630
danceratopz: And today we've got quite an important topic to discuss, which is basically the proposed EIP 8037 and EIP-8038 changes for Glamsterdam.

19
00:17:58.840 --> 00:18:06.699
danceratopz: As you all know, Glamsterdam brings a lot of important changes to Ethereum, including ePBS and Block Level Access Lists.

20
00:18:07.240 --> 00:18:17.069
danceratopz: But not only that, we snuck in another secret headliner called Scaling, which is another very large set of changes,

21
00:18:17.410 --> 00:18:23.599
danceratopz: But allows, or will allow Ethereum to scale from 60 million gas to 200 million per gas per block.

22
00:18:23.990 --> 00:18:38.170
danceratopz: And allow us to keep state growth under control. Unfortunately, rather late in our fork development process, we found out that the proposed changes in 8037 can break some smart contract invariants.

23
00:18:38.630 --> 00:18:45.100
danceratopz: And this affects contract behavior. So…

24
00:18:45.220 --> 00:18:54.959
danceratopz: We've invited, smart contract developers and app Devs today to come and talk to us about this, and give their inputs for these changes.

25
00:18:55.220 --> 00:19:09.039
danceratopz: It's important that we keep improving Ethereum and keeping it sustainable, but scaling isn't worth doing at any cost, obviously. I'm super excited to see so many people here,

26
00:19:09.310 --> 00:19:26.799
danceratopz: when I first got to Ethereum, I basically dabbled with smart contract development and was just super excited by just the ability to, like, deploy permissionlessly to the chain. That was really remarkable to me, and one of the reasons why I stuck around and became a protocol tester.

27
00:19:26.970 --> 00:19:35.529
danceratopz: At the EF, which is where I work now. So yeah, these changes were quite large, and

28
00:19:36.410 --> 00:19:38.349
danceratopz: Required quite a few trade-offs.

29
00:19:38.460 --> 00:19:50.200
danceratopz: So we've got a few people who are going to talk about this today. First off, Marius is going to give an overview of the current situation, and maybe the extent of these changes, and then we're going to hear from…

30
00:19:50.340 --> 00:20:00.080
danceratopz: Two smart contract audit teams, Chain Security and Dedaub, that have been investigating any contracts that have been affected by these changes.

31
00:20:02.330 --> 00:20:21.379
danceratopz: And then we're gonna hear, about the internal EF investigations, and follow that, we're gonna have, let me just check the agenda myself. We're gonna have an open Q&A, and then we're gonna see how Ethereum can better align the app layer and the protocol layer moving forward. So,

32
00:20:21.590 --> 00:20:28.209
danceratopz: Many thanks for coming today. I think Ivan recognized some names in the call, which is really nice.

33
00:20:29.140 --> 00:20:34.680
danceratopz: I'd like to hand over to Marius to give us an overview of the current situation and where we are at today.

34
00:20:35.000 --> 00:20:35.780
Marius Van Der Wijden (M): Yes.

35
00:20:36.000 --> 00:20:37.770
Marius Van Der Wijden (M): Thank you very much, Dan.

36
00:20:37.890 --> 00:20:51.609
Marius Van Der Wijden (M): Yeah, as Ben already said, multiple contracts were relying on invariants in the protocol that are modified by 8037 and 8038. So why did it come to this?

37
00:20:51.980 --> 00:21:06.739
Marius Van Der Wijden (M): We proposed 8037 and 8038 in 2025. It enables us to significantly scale Ethereum and Amsterdam by 3.5x, 3.3x, and even further in future hard forks.

38
00:21:07.660 --> 00:21:09.920
Marius Van Der Wijden (M): this 3.3x…

39
00:21:10.150 --> 00:21:19.629
Marius Van Der Wijden (M): Improvement will significantly reduce the cost for most users, so even though your cost in gas might go up.

40
00:21:19.910 --> 00:21:25.190
Marius Van Der Wijden (M): For most use cases, the actual cost to the user will significantly go down.

41
00:21:25.590 --> 00:21:33.529
Marius Van Der Wijden (M): And not only will this reduce costs for most users, it will also enable new use cases for Ethereum.

42
00:21:35.870 --> 00:21:44.130
Marius Van Der Wijden (M): The EIPs break a few invariants that we didn't think people would still rely on, so, kind of the work went on.

43
00:21:45.140 --> 00:21:55.699
Marius Van Der Wijden (M): We wanted to do significant public outreach, but due to multiple reorgs at the EF and budget cuts, the timelines for audits and outreach slipped.

44
00:21:56.290 --> 00:22:04.749
Marius Van Der Wijden (M): I personally share a lot of blame for these failures, as the last remaining X scaling track lead that's still at the EF.

45
00:22:05.330 --> 00:22:11.060
Marius Van Der Wijden (M): Fortunately, Chain Security made us aware of these contracts, and…

46
00:22:11.670 --> 00:22:16.370
Marius Van Der Wijden (M): Following the report from, chain security, we started investigating our thoughts.

47
00:22:16.580 --> 00:22:25.400
Marius Van Der Wijden (M): We will hear from Dedaub and Chain Security on their findings, as well… as from Toni on the EF internal investigation.

48
00:22:25.460 --> 00:22:44.060
Marius Van Der Wijden (M): And the EF has mobilized significant resources to make sure that core Devs and the community will receive an accurate picture of how many things are broken, what are broken, how much value is at stake, are there potential mitigations, and what the cost of these mitigations would be.

49
00:22:44.910 --> 00:22:50.299
Marius Van Der Wijden (M): Before we go there, I would like to remind, everyone of a few things.

50
00:22:50.550 --> 00:23:02.980
Marius Van Der Wijden (M): Application developers are crucial for Ethereum to thrive. Ethereum is a platform primarily built for trustless and unstoppable applications.

51
00:23:03.160 --> 00:23:07.429
Marius Van Der Wijden (M): And providing a stable API for app developers is crucial.

52
00:23:08.110 --> 00:23:18.659
Marius Van Der Wijden (M): The API that we are providing right now is unfortunately underspecified, and there are a lot of things that we know that we have to change about the API very soon.

53
00:23:18.830 --> 00:23:31.080
Marius Van Der Wijden (M): Vitalik wrote a… or published a blog post today, I think, that lists a few of those changes, like, just to call out one example, or, like, the most

54
00:23:31.250 --> 00:23:32.930
Marius Van Der Wijden (M): Important example.

55
00:23:33.280 --> 00:23:39.189
Marius Van Der Wijden (M): All contracts that rely on EC Recover, are going to be broken by the PQ transition.

56
00:23:39.420 --> 00:23:45.059
Marius Van Der Wijden (M): So… Yeah.

57
00:23:45.620 --> 00:23:58.859
Marius Van Der Wijden (M): we need to trans… or we need to, transition Ethereum to PQ at some point, in order for everyone to be… to be secure. And,

58
00:23:58.930 --> 00:24:08.639
Marius Van Der Wijden (M): there will be a large fallout from this. So there are these changes to the protocol that are coming up, that will break, certain contracts.

59
00:24:08.910 --> 00:24:27.880
Marius Van Der Wijden (M): What we have to be much better at is communicating these potential changes, and engaging with the community as well as security researchers before these changes hit testnets, so we will never be in this situation, again. We will later also talk a bit about

60
00:24:28.100 --> 00:24:33.470
Marius Van Der Wijden (M): Potential things that we can do in order to improve the situation there.

61
00:24:33.980 --> 00:24:41.469
Marius Van Der Wijden (M): But I would also, like, quickly like to comment on some of the proposed alternatives to 8037 and 8038.

62
00:24:43.290 --> 00:24:49.009
Marius Van Der Wijden (M): I've seen a few, and, like, AI makes it very easy to propose alternatives.

63
00:24:49.250 --> 00:25:09.069
Marius Van Der Wijden (M): That fix all the issues, but they do come with, significant drawbacks in… in UX for users, or for block producers, or increase the maintenance burden of the codebases, or… or are not, like, just not compatible with future changes.

64
00:25:09.710 --> 00:25:13.229
Marius Van Der Wijden (M): So, most of the proposed solution ended up something

65
00:25:13.580 --> 00:25:18.120
Marius Van Der Wijden (M): it's something that resembles kind of, like, a worse EOF spec.

66
00:25:18.280 --> 00:25:33.369
Marius Van Der Wijden (M): If we want to go into that direction of versioning the EVM, we should just do the full EOF spec, but that is definitely not something that we can bolt onto Amsterdam or even Hegota in any realistic time.

67
00:25:34.880 --> 00:25:42.670
Marius Van Der Wijden (M): Most of the proposed alternatives, also break other invariants in the protocol, which would require new rounds of investigation.

68
00:25:42.860 --> 00:25:53.009
Marius Van Der Wijden (M): And a lot of the proposals also primarily target invariants that aren't actually the ones that, where we saw most of the contracts breaking.

69
00:25:55.080 --> 00:26:12.320
Marius Van Der Wijden (M): To close this, this section out, the Ethereum community and the core devs have to make very hard decisions about the future of the protocol, basically all the time, but especially, around, around, like, these last years.

70
00:26:12.500 --> 00:26:25.280
Marius Van Der Wijden (M): Our job is to weigh the needs of all participants in the network. It's a $220 billion ecosystem, so there's a lot of value at stake, as well as a very diverse set of stakeholders.

71
00:26:25.830 --> 00:26:31.509
Marius Van Der Wijden (M): That's why this is such a hard decision, 8037 and 8038.

72
00:26:31.870 --> 00:26:39.409
Marius Van Der Wijden (M): 8038 enabled significantly higher gas limits, thus significantly lower cost for most users.

73
00:26:40.520 --> 00:26:55.000
Marius Van Der Wijden (M): But we also have to take into consideration, that potentially modifying Glamsterdam will create additional delays in the order of 4 to 6 months, for Hegota and the PQ transition and the whole roadmap afterwards.

74
00:26:55.600 --> 00:27:12.830
Marius Van Der Wijden (M): Ethereum is an unfinished project that will have to evolve significantly in the next years to face the challenges of PQ and AI, which means the amount of potentially broken invariants will increase before hopefully settling at some point in the future.

75
00:27:13.100 --> 00:27:23.699
Marius Van Der Wijden (M): And I would like to explicitly thank everyone who's coming on the call today, and also spent a lot of time in the last few weeks on this topic. Many people went

76
00:27:23.830 --> 00:27:43.320
Marius Van Der Wijden (M): well above and beyond to collect data in order to provide an accurate picture for an honest and fair debate today, and we will hear from, yeah, many different people, that have engaged in this, both in their work time and in their free time. So, thank you everyone for coming on here.

77
00:27:48.200 --> 00:27:55.000
danceratopz: Amazing. Thanks so much, Marius, for the… overview and undertakes that.

78
00:27:56.050 --> 00:28:07.269
danceratopz: Before we go to the Q&A, we're gonna, hear about the current findings that we've had, or that people have made from external auditors, and I think first up, chain security is going to

79
00:28:07.510 --> 00:28:09.070
danceratopz: Present their findings.

80
00:28:09.510 --> 00:28:14.959
danceratopz: I saw Hubert in the call, or is it Tanguy that's gonna present today?

81
00:28:15.320 --> 00:28:17.950
Tanguy (ChainSecurity): I really am.

82
00:28:17.950 --> 00:28:19.609
danceratopz: Would you like to share your screen?

83
00:28:19.760 --> 00:28:25.830
Tanguy (ChainSecurity): Yes, I made some slides, Jack, can you see the slide?

84
00:28:26.350 --> 00:28:28.340
danceratopz: Yeah, perfect. Yeah, thanks a lot.

85
00:28:28.930 --> 00:28:45.279
Tanguy (ChainSecurity): Amazing, thanks a lot. Yeah, so, hi everyone. We've been tasked by the EF to do mostly two things. On one side, find more protocol affected by 8037 and 8038 changes.

86
00:28:45.530 --> 00:28:52.680
Tanguy (ChainSecurity): And on the other side, to find more invariant or things breaking with Glamsterdam overall.

87
00:28:54.360 --> 00:29:10.999
Tanguy (ChainSecurity): So we combined two approaches. First, we had, Agent Swarm approach, which was searching protocol categories, starting from deployment list, verified source code, and public software.

88
00:29:11.000 --> 00:29:16.399
Tanguy (ChainSecurity): And we expanded then through on-chain activity and related implementation.

89
00:29:18.320 --> 00:29:31.069
Tanguy (ChainSecurity): A second search that we did was trace-based, so we took approximately one year of mainnet transaction, and this helped us find effective path across interacting contracts.

90
00:29:31.320 --> 00:29:40.020
Tanguy (ChainSecurity): We then investigated candidates by comparing Osaka and Amsterdam behavior through replays and constructed scenarios.

91
00:29:40.550 --> 00:29:46.140
Tanguy (ChainSecurity): And, yeah, then we could check the impact, and who controls the mitigation, if any.

92
00:29:46.540 --> 00:29:52.350
Tanguy (ChainSecurity): And, yeah, then some adversarial review of this.

93
00:29:52.600 --> 00:30:03.770
Tanguy (ChainSecurity): So, we do not think that the coverage is complete here for multiple reasons. For example, dormant path, unverified code, or private off-chain systems.

94
00:30:04.370 --> 00:30:08.470
Tanguy (ChainSecurity): Also, I'll come to this a bit later, but comparing with

95
00:30:08.770 --> 00:30:20.829
Tanguy (ChainSecurity): The EF investigation results, we found that our search missed findings that the EF identified, so we would expect there are also further affected systems to be found.

96
00:30:21.160 --> 00:30:29.860
Tanguy (ChainSecurity): And also one point is that we did not have assessed at all L2s here, so this is only Ethereum mainnet.

97
00:30:31.380 --> 00:30:49.319
Tanguy (ChainSecurity): So yeah, we found 152 findings, of which 76 were also found by the EF own internal search. I think this will be presented just a bit later in this call, which brought this overall search to a total of 200 findings.

98
00:30:49.490 --> 00:31:03.500
Tanguy (ChainSecurity): And here, it's important to note that one finding is not one contract, but one class of affected contracts, so if it's deployed multiple times, then this will not count in these numbers.

99
00:31:04.930 --> 00:31:24.679
Tanguy (ChainSecurity): On the invariant breakdown, so I will not go much into details of each invariant, but what we found is that most findings are coming from Invariant 4 and 7, which are from gas cost increases. So, when hard-coded gas budget and reserve are no longer sufficient under Glamsterdam.

100
00:31:25.080 --> 00:31:34.929
Tanguy (ChainSecurity): So those failures are not triggered by the new reservoir or refill mechanism, but really by the increased gas costs. So…

101
00:31:35.130 --> 00:31:41.730
Tanguy (ChainSecurity): From what we found, the reservoir and the refill are kind of a minority of finding here.

102
00:31:43.480 --> 00:31:51.959
Tanguy (ChainSecurity): On the mitigation analysis, so we found that the majority of the finding can be mitigated rather easily.

103
00:31:52.610 --> 00:32:05.900
Tanguy (ChainSecurity): However, there is still a few that requires either a contract upgrade, or in the case of a really immutable protocol, that will be a full redeployment of the system, because

104
00:32:06.030 --> 00:32:08.100
Tanguy (ChainSecurity): One cannot upgrade those.

105
00:32:09.020 --> 00:32:23.360
Tanguy (ChainSecurity): We also found some compatibility risks beyond this main finding set. On one side for, Gloas, so the consensus part of Glamsterdam.

106
00:32:23.700 --> 00:32:32.190
Tanguy (ChainSecurity): We identified, it'll have an issue across staking protocol and supporting service, where changes to…

107
00:32:32.320 --> 00:32:45.169
Tanguy (ChainSecurity): Proof layout can block staking operation, and also moving execution data can break some indexer and Oracle, for example. So for this, we will also coordinate with the EF to report those.

108
00:32:45.490 --> 00:32:50.199
Tanguy (ChainSecurity): And, the second one is,

109
00:32:50.860 --> 00:32:58.659
Tanguy (ChainSecurity): fixed gas limit for ETH transfer, so we found that some systems were assuming that,

110
00:32:58.810 --> 00:33:04.210
Tanguy (ChainSecurity): sending ETH to a new account, or generally to an account without coding code.

111
00:33:04.440 --> 00:33:13.790
Tanguy (ChainSecurity): They will assume that it should always succeed with 21,000 gas, and this is no longer the case for a new account, because…

112
00:33:14.110 --> 00:33:18.079
Tanguy (ChainSecurity): Now some account creation cost is being charged.

113
00:33:20.500 --> 00:33:32.579
Tanguy (ChainSecurity): And yeah, so summary and next step, so, yeah, with some AI-based search, we managed to find 150 findings, both on-chain and off-chain.

114
00:33:33.100 --> 00:33:36.390
Tanguy (ChainSecurity): Gloas is also, breaking stuff.

115
00:33:37.840 --> 00:33:46.159
Tanguy (ChainSecurity): We took some time to write a draft guide on how to fix all of this try-catch pattern, which,

116
00:33:46.290 --> 00:33:49.160
Tanguy (ChainSecurity): Which are the issue here.

117
00:33:49.280 --> 00:33:58.459
Tanguy (ChainSecurity): And we're happy to get feedback on this, because we feel, gas repricing will happen in the future, and people should,

118
00:33:58.760 --> 00:34:07.829
Tanguy (ChainSecurity): Make those try-catch robust in the future, otherwise we are gonna have this whole conversation once again in a few months or few years.

119
00:34:08.100 --> 00:34:17.209
Tanguy (ChainSecurity): And, yeah, we would also recommend thinking about the other EIPs, especially for app Devs, to look into the whole,

120
00:34:17.719 --> 00:34:22.520
Tanguy (ChainSecurity): Glamsterdam, and check if something else might break.

121
00:34:22.650 --> 00:34:24.639
Tanguy (ChainSecurity): In the application layer.

122
00:34:25.210 --> 00:34:28.980
Tanguy (ChainSecurity): And, yeah, thank you, that's… from outside.

123
00:34:34.710 --> 00:34:38.909
danceratopz: Amazing. Thanks a lot, Tanguy, for the report and sharing your findings.

124
00:34:42.130 --> 00:34:49.199
danceratopz: Dima from Lido asked if they could get a link to a report. I think Hubert responded,

125
00:34:49.389 --> 00:34:53.630
danceratopz: I guess you guys will share a public report in due course.

126
00:34:54.469 --> 00:34:55.859
Tanguy (ChainSecurity): Yeah, we really are.

127
00:34:56.179 --> 00:34:57.229
danceratopz: To some extent.

128
00:34:57.359 --> 00:34:59.539
danceratopz: Okay, perfect.

129
00:35:00.529 --> 00:35:08.769
danceratopz: Before we move on to Dedaub, did anyone want to, ask Tanguy some questions about their investigations and what they found?

130
00:35:18.149 --> 00:35:19.939
danceratopz: Feel free to raise now.

131
00:35:30.079 --> 00:35:43.159
danceratopz: All right, perfect. Then, I think we can leave it there, and any questions, if you think of them, we'll have time later for… in the Q&A. Is someone from Dedaub on the call to present their findings?

132
00:35:44.530 --> 00:35:46.829
Ilias | Dedaub: Hey, here. Ilias?

133
00:35:47.330 --> 00:35:48.530
danceratopz: Hi, Ilias.

134
00:35:49.240 --> 00:35:51.620
danceratopz: Perfect. Please feel free.

135
00:35:51.900 --> 00:35:53.489
Ilias | Dedaub: She goes insights, does she not?

136
00:35:58.390 --> 00:36:01.650
Ilias | Dedaub: You can see that… you can see my presentation, right?

137
00:36:02.530 --> 00:36:04.159
danceratopz: Not quite yet.

138
00:36:04.750 --> 00:36:06.589
Ilias | Dedaub: Hmm. Okay? Oh.

139
00:36:35.940 --> 00:36:42.509
danceratopz: No worries, take your time. I believe, Dima requested if, chain security could maybe,

140
00:36:42.640 --> 00:36:45.309
danceratopz: Share the report with them privately in the chat.

141
00:36:47.860 --> 00:36:50.660
danceratopz: And, Marius has also confirmed that

142
00:36:51.220 --> 00:36:57.630
danceratopz: on the EF side, we've made a big outreach campaign to Privately contact protocols.

143
00:37:00.930 --> 00:37:02.130
Ilias | Dedaub: Sorry, mine.

144
00:37:02.580 --> 00:37:05.340
Ilias | Dedaub: browser crashed. They didn't like,

145
00:37:07.530 --> 00:37:09.199
danceratopz: No worries, take your time.

146
00:37:09.200 --> 00:37:11.100
Ilias | Dedaub: I need to share my whole screen, because…

147
00:37:11.320 --> 00:37:15.589
danceratopz: We got… we got 90 minutes for the call now. It was only 60 minutes 2 weeks ago.

148
00:37:18.080 --> 00:37:19.679
Ilias | Dedaub: Okay, now you can see it.

149
00:37:20.500 --> 00:37:22.849
danceratopz: Perfect, yeah, that's great, thank you very much.

150
00:37:23.870 --> 00:37:29.910
Ilias | Dedaub: Okay, so… Hi, I'm Ilias Sadreis from, Dedaub, and…

151
00:37:30.150 --> 00:37:39.050
Ilias | Dedaub: Today, I'm gonna be presenting you our findings from the EIP 8037 and 8038 impact study we conducted.

152
00:37:40.040 --> 00:37:46.419
Ilias | Dedaub: So, brief introduction, just to set the tone. We were conducted by the Ethereum Foundation.

153
00:37:46.530 --> 00:37:49.530
Ilias | Dedaub: to do an impact study on the EIP.

154
00:37:49.760 --> 00:37:50.739
Ilias | Dedaub: So I'm saying?

155
00:37:52.200 --> 00:37:55.080
Ilias | Dedaub: 8, 878, sorry.

156
00:37:55.210 --> 00:38:01.119
Ilias | Dedaub: Many changes to gas accounting, and the info mechanisms were introduced here.

157
00:38:01.700 --> 00:38:07.280
Ilias | Dedaub: So this means that, unfortunately, a lot of contracts that rely on gas semantics

158
00:38:07.630 --> 00:38:11.540
Ilias | Dedaub: may stop working properly after the Amsterdam hard fork.

159
00:38:13.370 --> 00:38:19.829
Ilias | Dedaub: Moving forward to what we looked at, so… We follow the two-pronged approach.

160
00:38:20.760 --> 00:38:24.369
Ilias | Dedaub: So, I don't know, one part of the study focused on

161
00:38:25.270 --> 00:38:34.990
Ilias | Dedaub: Contracts with non-trivial gas opcode patterns. I'll specify… I'll define it later, what I mean with this, non-trivial gas opcode patterns phrase.

162
00:38:35.210 --> 00:38:42.190
Ilias | Dedaub: And the other phrase… the other part of the study, focused on, identifying products that was

163
00:38:42.330 --> 00:38:43.820
Ilias | Dedaub: Not default gas.

164
00:38:44.000 --> 00:38:51.600
Ilias | Dedaub: So, this can be typically constant gas, but not necessarily to external codes, which may also be affected by the

165
00:38:52.160 --> 00:38:53.719
Ilias | Dedaub: Amsterdam changes.

166
00:38:54.770 --> 00:38:57.849
Ilias | Dedaub: Both universes of these contexts were then classified using

167
00:38:58.100 --> 00:39:06.210
Ilias | Dedaub: various analysis techniques, static analysis, symbolic execution, and then, of course, triage with AI and My node inspection.

168
00:39:07.390 --> 00:39:12.949
Ilias | Dedaub: So… At… looking a bit closer to the first part of the study.

169
00:39:13.840 --> 00:39:20.810
Ilias | Dedaub: We looked at contracts, as I said, that have non-trivial gas patterns.

170
00:39:21.110 --> 00:39:24.719
Ilias | Dedaub: And that have received transactions in the last 6 months.

171
00:39:24.940 --> 00:39:27.509
Ilias | Dedaub: So what I mean by non-trivial gas patterns?

172
00:39:27.910 --> 00:39:35.420
Ilias | Dedaub: is a… Basically, any use of the GAS opcode that is not intended

173
00:39:36.260 --> 00:39:38.910
Ilias | Dedaub: To be used as a parameter to…

174
00:39:39.390 --> 00:39:41.999
Ilias | Dedaub: To forward all the remaining gas to an external call.

175
00:39:42.410 --> 00:39:47.659
Ilias | Dedaub: Which is basically the default for solidity and…

176
00:39:47.930 --> 00:39:49.650
Ilias | Dedaub: how they're not gonna align it?

177
00:39:50.320 --> 00:39:57.460
Ilias | Dedaub: So, by filtering this, we basically capture all interesting uses of the gas opcode interaction.

178
00:39:57.700 --> 00:40:01.759
Ilias | Dedaub: And in order to do this, we used our decompiled contracts database.

179
00:40:02.600 --> 00:40:06.529
Ilias | Dedaub: And the reason we use this database oversource is because

180
00:40:07.340 --> 00:40:14.850
Ilias | Dedaub: This is a highly normalized representation, so this type of expressions get in line, and it's very easy to write a…

181
00:40:15.240 --> 00:40:20.779
Ilias | Dedaub: regex to… Index through the compiled code and get the contracts you're interested in.

182
00:40:22.200 --> 00:40:26.060
Ilias | Dedaub: This search resulted in around 25,000 contracts.

183
00:40:26.840 --> 00:40:35.880
Ilias | Dedaub: For the verified subset of these contracts, we fed it to an agent workflow we have designed.

184
00:40:36.390 --> 00:40:40.759
Ilias | Dedaub: Which, and we ended up with,

185
00:40:41.610 --> 00:40:44.969
Ilias | Dedaub: A couple of address… a couple of, verified findings.

186
00:40:45.260 --> 00:40:54.529
Ilias | Dedaub: More specifically, we have 12 and 70 addresses… 12 high findings and 70 medium-high… 70… 70 medium findings.

187
00:40:54.800 --> 00:41:03.619
Ilias | Dedaub: Which are all confirmed, by the agent with, It's… EELS, replays.

188
00:41:04.970 --> 00:41:07.420
Ilias | Dedaub: And, interestingly,

189
00:41:07.710 --> 00:41:15.339
Ilias | Dedaub: It's… it appears that there are only 22 contracts in the dataset that we didn't identify, but the rest of the…

190
00:41:16.090 --> 00:41:22.270
Ilias | Dedaub: internal Ethereum foundations that they managed to find, so… This gives us a…

191
00:41:22.940 --> 00:41:28.770
Ilias | Dedaub: A bit of confidence that our exploration in this type of direction was complete.

192
00:41:29.910 --> 00:41:35.710
Ilias | Dedaub: There is still a bit of work to do, but this is pending work, I mean, we will keep updating the report.

193
00:41:36.260 --> 00:41:38.550
Ilias | Dedaub: With the latest numbers.

194
00:41:39.320 --> 00:41:42.900
Ilias | Dedaub: But this should give a high-level picture of our findings.

195
00:41:43.770 --> 00:41:48.360
Ilias | Dedaub: So, moving forward to the second part of the study,

196
00:41:49.420 --> 00:41:57.059
Ilias | Dedaub: So, unfortunately, Glamsterdam also breaks contracts that do not necessarily use gas, but maybe they pass,

197
00:41:58.280 --> 00:42:05.959
Ilias | Dedaub: explicit, non-construct gas to… non-default gas to cycles, so…

198
00:42:06.150 --> 00:42:13.220
Ilias | Dedaub: This may break their behavior due to increased gas costs, the refund mechanism changing, etc.

199
00:42:13.570 --> 00:42:18.530
Ilias | Dedaub: So, we did an extra study on this, and unfortunately.

200
00:42:19.680 --> 00:42:23.940
Ilias | Dedaub: Just to keep this dataset of, manageably, we had to…

201
00:42:24.320 --> 00:42:32.929
Ilias | Dedaub: to filter further based on dynamic features. So, we looked at contracts that have more than 100 transactions within a tight time… tight time frame.

202
00:42:33.050 --> 00:42:43.039
Ilias | Dedaub: Or had a high-value transaction within two weeks. And this resulted, again, in a 25K… Approximate, data set.

203
00:42:43.800 --> 00:42:47.690
Ilias | Dedaub: Similar to the study, before.

204
00:42:48.480 --> 00:42:53.970
Ilias | Dedaub: So This is the breakdown of the findings from this part of the study.

205
00:42:55.070 --> 00:43:01.120
Ilias | Dedaub: Unfortunately, I don't… I didn't have the space in the slides or the time to go deep into each of these patterns.

206
00:43:01.430 --> 00:43:05.190
Ilias | Dedaub: These are exactly the patterns that are source analysis.

207
00:43:06.260 --> 00:43:08.990
Ilias | Dedaub: Ehr… Try to detect.

208
00:43:09.460 --> 00:43:12.900
Ilias | Dedaub: And, for example, you know.

209
00:43:13.900 --> 00:43:21.160
Ilias | Dedaub: Conditionals on brands… on gas left, fixed reserves,

210
00:43:21.390 --> 00:43:25.140
Ilias | Dedaub: Fixed comb… combat gas. All these analysis we wrote,

211
00:43:25.270 --> 00:43:30.100
Ilias | Dedaub: On our analysis engine, to try and find affected contracts.

212
00:43:31.120 --> 00:43:34.610
Ilias | Dedaub: And we ended up with quite a few verified positives.

213
00:43:34.950 --> 00:43:37.879
Ilias | Dedaub: These are with transaction replays again.

214
00:43:39.400 --> 00:43:47.220
Ilias | Dedaub: So… This was our findings, and currently we are in the process of fading out the affected protocols.

215
00:43:47.760 --> 00:43:51.760
Ilias | Dedaub: This is a list of the protocols we're reaching out.

216
00:43:51.970 --> 00:43:56.200
Ilias | Dedaub: Of course, this is… Done in a best-effort way, just to not bombard.

217
00:43:56.580 --> 00:44:02.059
Ilias | Dedaub: This protocols with, warnings that other teams or colleagues

218
00:44:02.190 --> 00:44:10.999
Ilias | Dedaub: Have already flagged to the security teams of this project, so… we're doing our best.

219
00:44:11.710 --> 00:44:14.710
Ilias | Dedaub: Yeah, thank you very much, this is it on our end.

220
00:44:15.040 --> 00:44:15.779
Ilias | Dedaub: Thank you.

221
00:44:22.160 --> 00:44:24.529
danceratopz: Thanks very much, Ilias, for sharing your findings.

222
00:44:25.750 --> 00:44:33.850
danceratopz: Does anyone have any questions for Ilias or Dedaub regarding their findings on the EIP-8037 invariants?

223
00:44:52.550 --> 00:44:53.770
danceratopz: Okay.

224
00:45:00.520 --> 00:45:07.719
danceratopz: Yeah. I would just echo Jochem's sentiment in the chat that, very grateful to these security researchers.

225
00:45:07.940 --> 00:45:09.539
danceratopz: For spending the time on this.

226
00:45:10.690 --> 00:45:15.700
danceratopz: And, also respecting protocol time, or app protocol time.

227
00:45:15.930 --> 00:45:16.810
danceratopz: Alright.

228
00:45:19.030 --> 00:45:20.220
danceratopz: Okay.

229
00:45:21.040 --> 00:45:31.220
danceratopz: Moving forward, now we would have, Rahul, who's, from the access layer cluster.

230
00:45:31.560 --> 00:45:41.589
danceratopz: at the EF, who's, working on dev tooling, who would, come and share a little bit about his outreach to application teams.

231
00:45:42.030 --> 00:45:44.640
danceratopz: Regarding the 8037 findings.

232
00:45:45.480 --> 00:45:47.330
danceratopz: Rahul, are you in the call?

233
00:45:47.950 --> 00:45:49.029
Rahul (EF): Yes,

234
00:45:53.880 --> 00:45:59.229
danceratopz: Let me see if I can share your… Give me 20 seconds.

235
00:46:00.300 --> 00:46:01.099
Rahul (EF): I will…

236
00:46:01.100 --> 00:46:01.940
danceratopz: Sheep.

237
00:46:01.940 --> 00:46:08.089
Rahul (EF): excuse the background noise. I am actually on the commute.

238
00:46:08.430 --> 00:46:24.549
Rahul (EF): Yeah, as Marius and some of the audit companies mentioned, we… the EF has been doing their own analysis of which teams are affected. At a super high level, we've done this by replaying millions of transactions over the last years on L1.

239
00:46:24.710 --> 00:46:32.510
Rahul (EF): and also using Sourcify, to figure which invariants are broken, thanks to the effort of chain security.

240
00:46:32.780 --> 00:46:36.279
Rahul (EF): Actually, Dan, could you move to the other

241
00:46:36.580 --> 00:46:38.990
Rahul (EF): slide that I sent first.

242
00:46:39.480 --> 00:46:42.420
danceratopz: Jason, too. You're gonna have to give me a second, sorry.

243
00:46:42.700 --> 00:46:43.700
danceratopz: I thought.

244
00:46:43.700 --> 00:46:44.070
Rahul (EF): Oh, yeah.

245
00:46:44.070 --> 00:46:44.790
danceratopz: That's fine.

246
00:46:45.750 --> 00:46:52.370
Rahul (EF): I can speak about it. Basically, we have found 3 highs, and…

247
00:46:53.300 --> 00:46:56.740
Rahul (EF): Happily, all the three highs have already

248
00:46:57.140 --> 00:47:01.229
Rahul (EF): Not just been contacted, but one of them has already fixed

249
00:47:01.350 --> 00:47:05.870
Rahul (EF): The invariant breaking, and two of them plan to fix it over the course of the next month.

250
00:47:06.550 --> 00:47:07.490
Rahul (EF): And…

251
00:47:07.680 --> 00:47:24.840
Rahul (EF): There are a significantly large number of mediums and lows, as is obvious with chain security and Dedaub reports as well. I've hidden the actual number just because there's a really high amount of false positives, but we have confirmed that there are at least 25 mediums.

252
00:47:25.070 --> 00:47:31.850
Rahul (EF): That are impacted. A lot of those are actually on the call, so I'm happy to see you guys here. Thank you for taking the time.

253
00:47:32.290 --> 00:47:37.049
Rahul (EF): Yeah, Dan, can you move to the mediums?

254
00:47:37.920 --> 00:47:38.590
danceratopz: Yep.

255
00:47:40.550 --> 00:47:46.359
Rahul (EF): Yeah, so out of the 25 that are confirmed as not to be false positives,

256
00:47:46.620 --> 00:47:57.660
Rahul (EF): At least 16 have not just acknowledged, but are actively triaging, or figuring out a fix, or are on this call, to figure out what we can do, and how we can help.

257
00:47:57.980 --> 00:48:06.690
Rahul (EF): Out of those 16, 3 have already fixed, the other 8 plan to fix by the next month, or before the end of the year.

258
00:48:07.030 --> 00:48:12.380
Rahul (EF): Which is very nice if you want to… ship Glamsterdam pretty soon.

259
00:48:12.990 --> 00:48:25.599
Rahul (EF): There are still a few teams that we have contacted that haven't yet triaged. We're still working with them. Off note, there are a bunch of L2s that have independently forked

260
00:48:26.600 --> 00:48:29.820
Rahul (EF): OP's codebase or OP stack chains.

261
00:48:30.720 --> 00:48:32.830
Rahul (EF): There are,

262
00:48:33.820 --> 00:48:50.910
Rahul (EF): a good bunch of them still need to be… the OP has actually already fixed their invariants, and it's live with OP Sepolia. We just need to coordinate with a bunch of these teams on replaying on their chain's testnets.

263
00:48:53.080 --> 00:48:59.790
Rahul (EF): Yeah, this… yeah, more I can share later on how we're doing on comms, but…

264
00:49:00.250 --> 00:49:07.049
Rahul (EF): A huge shout out to all of the app teams that have spoken to us so quickly and coordinating with us.

265
00:49:07.360 --> 00:49:09.000
Rahul (EF): Giving me feedback.

266
00:49:09.250 --> 00:49:16.669
Rahul (EF): every art that helps us improve this process. Thank you. I'll stop here, and then we can talk later in the next section.

267
00:49:23.790 --> 00:49:28.730
danceratopz: Awesome. Thanks very much, Rahul, for sharing those.

268
00:49:35.730 --> 00:49:45.230
danceratopz: Did, on this… sorry, in this section, we… is Jango in the call from Juicebox? And did he want to speak up and…

269
00:49:45.900 --> 00:49:47.449
danceratopz: ask his questions.

270
00:49:49.820 --> 00:49:51.510
Jango: Hello, yes, I'm on the call.

271
00:49:52.150 --> 00:49:54.189
danceratopz: Thanks for joining, it's good to see you.

272
00:49:54.620 --> 00:49:59.099
Jango: My pleasure, thank you for hosting, and… Giving the opportunity.

273
00:49:59.440 --> 00:50:03.349
Jango: Just to clarify some, poignant…

274
00:50:03.540 --> 00:50:07.520
Jango: Things in the juicebox protocol to make things explicit.

275
00:50:07.540 --> 00:50:14.249
Jango: And how it affects, our apps. The biggest one are try-catch blocks. We…

276
00:50:14.250 --> 00:50:29.589
Jango: As an extensible protocol, we try to just be the generalized pay and cash out functions of the open internet, and so folks should be able to bring extensions on what happens when you pay, what happens when you cash out, including payouts, splits.

277
00:50:30.040 --> 00:50:30.980
Jango: Etc.

278
00:50:31.550 --> 00:50:33.199
Jango: And there are…

279
00:50:33.730 --> 00:50:50.320
Jango: many moments with, with Glam that the user can choose and tune the gas limit such that the, composability aspect fails, and then triggers the catch block, and then, the rest of the code runs.

280
00:50:51.100 --> 00:51:04.740
Jango: Which, you know, the try-catch is there such that if someone brings a composable element that is truly broken for whatever reason, the rest of the execution does, does not continue. Or rather, continues.

281
00:51:07.050 --> 00:51:26.999
Jango: which, you know, prevents someone from changing a composable element and then breaking the rest of the protocol. But this, change would basically allow users to tune, do they want to go into the composable element or not? And I feel like composability is a large, value prop of Ethereum. So,

282
00:51:27.480 --> 00:51:32.630
Jango: Certainly of the Juicebox protocol, but I imagine of Ethereum at large.

283
00:51:33.610 --> 00:51:47.710
Jango: A second point I think, is more philosophical here, and the one word that comes to mind when I think about this change, and a lot of the time I spend building protocols over the past years, is confidence. I think

284
00:51:47.950 --> 00:51:56.569
Jango: In thinking about breaking invariants, in my mind, I would be at peace with the trade-offs if…

285
00:51:56.570 --> 00:52:16.120
Jango: the main purpose of breaking them was to increase confidence. In the case of quantum, of AI hardening, all those make sense to me as an app developer of, like, damn, I may have really severe trade-offs to communicate to users, in extreme cases redeployment, since everything we try to build is immutable.

286
00:52:16.450 --> 00:52:19.099
Jango: Everything we do build is immutable.

287
00:52:19.390 --> 00:52:27.139
Jango: But if it makes a ton of sense with regard to confidence going forward, then I feel like I could justify it to myself.

288
00:52:27.360 --> 00:52:29.970
Jango: And therefore to users.

289
00:52:30.660 --> 00:52:45.600
Jango: And, I question if that confidence is really at play here. I think there is confidence in sustainability, which I think is very important, so I agree with the general essence of why these changes in GLAM are being proposed.

290
00:52:45.970 --> 00:52:56.800
Jango: But the trade-off in developer confidence in building immutable protocols seems at risk here. And I think the question I would just leave us with is.

291
00:52:58.940 --> 00:53:18.249
Jango: do we find it now foolish to build immutable protocols on Ethereum? Is that where we're headed towards until we reach a point of bedrock stability? And if so, I think it's just good to pronounce that very clearly and obviously, so that we don't end up in this miscommunication.

292
00:53:18.730 --> 00:53:19.540
Jango: Thank you.

293
00:53:24.440 --> 00:53:26.640
danceratopz: Thanks a lot, Jango, for,

294
00:53:26.750 --> 00:53:32.480
danceratopz: sharing your comment on the issue, and for coming and articulating that so clearly here today.

295
00:53:33.300 --> 00:53:41.460
danceratopz: I think it's… it's very valuable input for us all here. Would someone like to…

296
00:53:41.970 --> 00:53:46.540
danceratopz: step up and talk directly to Jango's Question and concerns?

297
00:53:51.590 --> 00:53:55.459
danceratopz: I'm wondering if, potentially Marius would like to talk to

298
00:53:55.680 --> 00:53:59.440
danceratopz: the state that Ethereum is at with regards to ossification.

299
00:54:00.100 --> 00:54:01.270
danceratopz: Or anyone else?

300
00:54:01.900 --> 00:54:06.120
Marius Van Der Wijden (M): Yeah, I… I can… I can… I can respond. I… I…

301
00:54:07.010 --> 00:54:15.260
Marius Van Der Wijden (M): I do empathize a lot with what you're saying,

302
00:54:16.680 --> 00:54:18.999
Marius Van Der Wijden (M): I think what we… what we kind of…

303
00:54:19.350 --> 00:54:27.050
Marius Van Der Wijden (M): have seen with the analysis is that it's not the majority of contracts that we are… we are… we are breaking.

304
00:54:27.170 --> 00:54:32.469
Marius Van Der Wijden (M): Or that I might need to upgrade. I think it's around…

305
00:54:32.910 --> 00:54:41.419
Marius Van Der Wijden (M): In the, in the highest estimates, it's around, 10, 15,000 contracts?

306
00:54:43.220 --> 00:54:51.370
Marius Van Der Wijden (M): the amount of contracts on Ethereum, in comparison, is around 90 million, so…

307
00:54:51.480 --> 00:54:55.460
Marius Van Der Wijden (M): Like, this is… this is a… I would say 8…

308
00:54:56.180 --> 00:55:06.080
Marius Van Der Wijden (M): I don't want to, yeah, make this sound less than, but it's, like, a fringe subset of the contracts that are impacted by this change.

309
00:55:06.300 --> 00:55:10.219
Marius Van Der Wijden (M): I do think that we need to be extremely

310
00:55:10.360 --> 00:55:21.849
Marius Van Der Wijden (M): Extremely careful when we are breaking user space, that the out… that the benefits outweigh, the risks and the drawbacks.

311
00:55:26.040 --> 00:55:33.200
Marius Van Der Wijden (M): I, for myself, feel like the… they do in this case,

312
00:55:33.650 --> 00:55:40.460
Marius Van Der Wijden (M): And, also in the other cases that are…

313
00:55:41.000 --> 00:55:43.779
Marius Van Der Wijden (M): That, that are being proposed.

314
00:55:43.970 --> 00:55:47.060
Marius Van Der Wijden (M): Or that, that are on the…

315
00:55:47.210 --> 00:55:53.140
Marius Van Der Wijden (M): timeline, or in… or coming in the future. If you…

316
00:55:54.150 --> 00:56:02.100
Marius Van Der Wijden (M): I see there's… there's, like, a… there's, like, a question about… About the amount of value?

317
00:56:02.770 --> 00:56:05.159
Marius Van Der Wijden (M): In these contracts.

318
00:56:05.360 --> 00:56:14.629
Marius Van Der Wijden (M): I think the total amount of value is… I don't have any reliable numbers on this, but it is on the order of

319
00:56:15.010 --> 00:56:21.170
Marius Van Der Wijden (M): A few hundred thousand, And… wire…

320
00:56:21.640 --> 00:56:26.200
Marius Van Der Wijden (M): At least the… sorry, the contracts that have not yet

321
00:56:26.380 --> 00:56:33.429
Marius Van Der Wijden (M): Said that they're going to upgrade, and, and, so these, yeah.

322
00:56:33.560 --> 00:56:42.760
Marius Van Der Wijden (M): more immutable contracts. And so it's very hard for me to justify

323
00:56:43.110 --> 00:56:49.309
Marius Van Der Wijden (M): Holding back this, this, like, $300 billion ecosystem, for, like.

324
00:56:49.490 --> 00:56:58.699
Marius Van Der Wijden (M): a few 300, 400, 500K worth of, Worth of value.

325
00:56:59.480 --> 00:57:01.669
Marius Van Der Wijden (M): I don't think that it's… it's…

326
00:57:01.860 --> 00:57:05.930
Marius Van Der Wijden (M): it's a great argument to make, to be honest.

327
00:57:06.460 --> 00:57:12.010
Marius Van Der Wijden (M): I don't like it either, because I feel like… no matter…

328
00:57:12.280 --> 00:57:21.499
Marius Van Der Wijden (M): how much money actually is locked in the contract. You should have a right to not be broken by,

329
00:57:21.910 --> 00:57:30.980
Marius Van Der Wijden (M): by protocol changes in an ideal world. but yeah, I think… Most of the…

330
00:57:32.190 --> 00:57:36.059
Marius Van Der Wijden (M): failures that I've seen are coming also from

331
00:57:36.370 --> 00:57:40.160
Marius Van Der Wijden (M): Their root causes, kind of.

332
00:57:40.360 --> 00:57:48.109
Marius Van Der Wijden (M): Stuff that we have… Made very explicitly in the past…

333
00:57:48.380 --> 00:58:04.579
Marius Van Der Wijden (M): that, like, make pretty explicit statements in the past that these things are going to change. Like, you shouldn't be hard-coding gas values in your, in your contracts. I think that is, that is,

334
00:58:05.230 --> 00:58:12.299
Marius Van Der Wijden (M): Most of the failures, actually, is, like, people, still hard-coding,

335
00:58:12.870 --> 00:58:21.290
Marius Van Der Wijden (M): Gas values and still relying on the current gas schedule. And, that is something that we have

336
00:58:21.970 --> 00:58:33.029
Marius Van Der Wijden (M): said, for a long time that we have to change the gas schedule, and we will have to update it, and it has to be updated again with ZK EVMs, and

337
00:58:33.620 --> 00:58:40.599
Marius Van Der Wijden (M): And maybe in the future with RISC-V, and so on, so… Yeah.

338
00:58:50.980 --> 00:58:53.800
danceratopz: Thanks, Marius, for chiming in.

339
00:58:54.120 --> 00:58:56.080
danceratopz: Toni, you have your hand raised?

340
00:58:56.860 --> 00:59:12.900
Toni Wahrstaetter: Yeah, just as a response to Jango, because I think, what he said is very valid, and should be taken seriously. It feels like we must differentiate the discussion a bit between breaking, things because of hard-coded gas versus

341
00:59:14.340 --> 00:59:24.120
Toni Wahrstaetter: The more complex, breakage, which is, like, invariant breaking and gas left, for example, being able to increase,

342
00:59:24.510 --> 00:59:43.549
Toni Wahrstaetter: it's a very different, breakage than just hardcoding gas. And one thing I've been hearing a lot in the past weeks since we… since chain security brought, brought the topic up, and people have started looking more deeply into the topic, is that it's just very unclear right now for app Devs

343
00:59:44.000 --> 01:00:01.809
Toni Wahrstaetter: what things core devs think about potentially breaking in the future versus what are actually the hard guarantees that won't be broken. And it's easy to just say everything around gas, but probably there's a lot of assumptions taken around gas that are also,

344
01:00:02.170 --> 01:00:05.300
Toni Wahrstaetter: Very set in stone versus others that are, like.

345
01:00:05.730 --> 01:00:10.790
Toni Wahrstaetter: about to break with every potential repricing we have to do. So, one…

346
01:00:10.890 --> 01:00:27.269
Toni Wahrstaetter: action item, I think that many people, have already, voiced, and also I saw it on the ACD agenda, would be to just be more clear in what are the things that might be breaking in the future. Vitalik already started,

347
01:00:27.700 --> 01:00:33.740
Toni Wahrstaetter: listing some of those in his blog post, and I think we should just make it a more formal process.

348
01:00:34.210 --> 01:00:37.639
Toni Wahrstaetter: and define those a little bit better. And this might…

349
01:00:37.930 --> 01:00:41.250
Toni Wahrstaetter: It might not help with the current situation, but at least we can…

350
01:00:41.420 --> 01:00:43.719
Toni Wahrstaetter: Do better in the future on that one.

351
01:00:52.430 --> 01:00:53.270
danceratopz: Thanks, Toni.

352
01:01:05.740 --> 01:01:08.990
danceratopz: Jango, you had, a question in the chat.

353
01:01:09.310 --> 01:01:13.539
danceratopz: Regarding the change, and Milos wanted to ask.

354
01:01:14.090 --> 01:01:17.389
danceratopz: whether this is similar to LZ Compose from Layer 0.

355
01:01:17.710 --> 01:01:20.129
danceratopz: Which, in that case, should be affected as well.

356
01:01:23.420 --> 01:01:28.100
Jango: I'm not sure about the Layer Zero, aspect,

357
01:01:28.100 --> 01:01:28.590
danceratopz: Okay.

358
01:01:29.020 --> 01:01:38.070
Jango: But, yeah, I think, like, in general, when I think about the fringe con… contract, Sentiment.

359
01:01:38.800 --> 01:01:43.019
Jango: I think we know that the majority of contracts published on Ethereum are

360
01:01:43.170 --> 01:01:53.779
Jango: copy-paste forks of standards, and those are standards for a reason. They give us confidence over things like ERC20s and the basics.

361
01:01:53.850 --> 01:02:11.240
Jango: But looking at that as the… as what Ethereum is, and the fringe net new, trying to create new types of… of things, creative things that actually touch business, that touch the real world, that haven't been done a thousand times before, and maybe don't have the full…

362
01:02:11.490 --> 01:02:24.930
Jango: capital at stake volume yet, but are working towards it, or maybe previous versions do, in the case of Juicebox, like Constitution DAO, Assange DAO, Tornado Cash Legal Defense, like, all of those projects have tens of millions of dollars in them, though the protocol has evolved

363
01:02:25.310 --> 01:02:34.769
Jango: Due to security reasons. Those projects being safe, but, you know, trying to expand to accommodate general business onto Ethereum.

364
01:02:36.630 --> 01:02:49.639
Jango: we are doing net new things. It's not just off-the-shelf OZ stuff. And it's… it's a… it's a delicate thing to shake the confidence of doing net new things on Ethereum.

365
01:02:55.080 --> 01:03:00.260
danceratopz: Yeah, thanks, Jango, I appreciate it. I would just… I just wanted to ask you if…

366
01:03:02.150 --> 01:03:05.719
danceratopz: from Marius's and Toni's answer, if…

367
01:03:06.560 --> 01:03:12.420
danceratopz: We can do better as an ecosystem to bridge this gap between protocol developers and app developers.

368
01:03:12.950 --> 01:03:22.680
danceratopz: And… From the protocol developer side, we can improve our processes and communicate more clearly, and certainly earlier.

369
01:03:23.100 --> 01:03:28.370
danceratopz: Would… would that maintain your confidence in deploying on-chain? And…

370
01:03:29.440 --> 01:03:29.930
Jango: Yeah.

371
01:03:29.930 --> 01:03:30.610
danceratopz: work.

372
01:03:30.830 --> 01:03:50.720
Jango: Yeah, I mean, like, to be honest, I think we're all very similar as people, as engineers, you know, we're very focused on our thing, you know, our part of the puzzle. And we all, you know, have edges of our puzzle that talk to each other, but we're so focused on what's right in front of us, and we take kind of the rest as assumptions.

373
01:03:50.910 --> 01:04:03.659
Jango: That, you know, we lose sight of a lot of the communication puzzle, we delegate, we kind of… there's information lost at the edges. So it feels normal to me, you know, I empathize with that, a lot.

374
01:04:03.900 --> 01:04:09.200
Jango: And it's tough to… to kind of end up in a settled state, like, tending towards

375
01:04:09.200 --> 01:04:33.510
Jango: zero instead of trying to do more and accommodate more, more voices. You know, I think about this from our protocol perspective, it's like, there's users here, and there's a world of users, like, addressable market to come, and making decisions that go after, the addressable market, but, you know, don't fully attend to what's right, kind of already here and committed and fully, like, on board with all the trade-offs,

376
01:04:33.580 --> 01:04:39.729
Jango: As stated… as stated, is a tough thing to negotiate, especially at scale.

377
01:04:39.940 --> 01:04:57.219
Jango: So… I… I just want to say that we're on the same team, we're on the same page. There's obviously communication stuff that we can do better, there's assumptions that can be made clearer so that we can move forward with that confidence. but,

378
01:04:57.450 --> 01:05:10.750
Jango: it's hard to put yourself in the kind of shoes of the puzzle piece next door, and I honestly want y'all to focus on your side of the puzzle piece as much as possible, because that's, like, dope, and I love being able to focus on mine, too.

379
01:05:10.870 --> 01:05:25.260
Jango: It's just when we're then selling to users, not in the name of a sales process, necessarily, but in the name of actually having beliefs that we care about, that we're trying to, kind of…

380
01:05:25.490 --> 01:05:27.290
Jango: reverberate.

381
01:05:27.540 --> 01:05:40.190
Jango: so that others understand it, too, and reach that aha moment. And then when, you know, you've done all that work to get to that aha moment, and then certain bits feel fragile after the fact, it's… it's really hard to then

382
01:05:40.240 --> 01:05:52.840
Jango: kind of retro-justify things. And I'm sure, like, what I'm saying here is not anything new, but it's the sentiment that I think is important to internalize.

383
01:05:56.080 --> 01:06:02.920
danceratopz: Yeah, absolutely. I'm very glad you came here to share that today, and it's been taken on board.

384
01:06:03.800 --> 01:06:18.169
danceratopz: We're definitely going to have a section a bit later in the call, about how we can better align our protocol work with the app layer. I think Toni already alluded to it before.

385
01:06:18.490 --> 01:06:22.299
danceratopz: We can definitely do things better in the future, and we will strive to do that.

386
01:06:24.640 --> 01:06:25.560
danceratopz: Marius?

387
01:06:26.930 --> 01:06:39.650
Marius Van Der Wijden (M): Yeah, I think… Maybe part of the disconnect also comes from that like, the core dev community has…

388
01:06:39.810 --> 01:06:45.230
Marius Van Der Wijden (M): Made some of these decisions in the past already. That…

389
01:06:46.450 --> 01:06:52.890
Marius Van Der Wijden (M): where… I don't know, with… 29, 30,

390
01:06:53.200 --> 01:06:57.440
Marius Van Der Wijden (M): With the increase in state access costs.

391
01:06:57.580 --> 01:07:09.170
Marius Van Der Wijden (M): Where we knew, like, there would be some contracts, breaking, so… For us, that…

392
01:07:09.640 --> 01:07:21.759
Marius Van Der Wijden (M): like, that assumption that everything is, like, every single contract on Ethereum, 100%, yeah… is immutable, isn't… is…

393
01:07:22.420 --> 01:07:27.830
Marius Van Der Wijden (M): It's not… yeah, it's just, like, this, this,

394
01:07:28.000 --> 01:07:31.230
Marius Van Der Wijden (M): This kind of assumption has been broken.

395
01:07:31.830 --> 01:07:39.500
Marius Van Der Wijden (M): much earlier already, and we have kind of learned to live with it. Now…

396
01:07:39.700 --> 01:07:44.110
Marius Van Der Wijden (M): We have to br-, like… I think we have to break it in a…

397
01:07:44.550 --> 01:07:56.259
Marius Van Der Wijden (M): slightly bigger way than back then. But, yeah, I still think, like, we should be extremely, extremely cautious when we do this.

398
01:07:56.870 --> 01:08:02.829
Marius Van Der Wijden (M): And, only ever do it in, like, the most important things, and…

399
01:08:09.450 --> 01:08:10.250
Marius Van Der Wijden (M): Yep.

400
01:08:10.730 --> 01:08:18.309
Marius Van Der Wijden (M): I'm… I… I don't… I'm not good app dev enough to…

401
01:08:18.850 --> 01:08:24.210
Marius Van Der Wijden (M): To say whether there is, whether this means… like…

402
01:08:26.120 --> 01:08:29.800
Marius Van Der Wijden (M): You cannot create any save, try, catch.

403
01:08:31.990 --> 01:08:33.279
Marius Van Der Wijden (M): Anymore?

404
01:08:33.920 --> 01:08:41.590
Marius Van Der Wijden (M): I'm pretty sure you still can create something, but yeah, it's,

405
01:08:46.149 --> 01:08:57.710
Marius Van Der Wijden (M): I think part of the disconnect… what I'm trying to say is part of the disconnect happened much earlier, where the Core Devs community is, we're already, like.

406
01:08:57.920 --> 01:09:02.970
Marius Van Der Wijden (M): two, three, forks ahead, and we… we see…

407
01:09:03.120 --> 01:09:09.040
Marius Van Der Wijden (M): All of the things that will break, and we have come to terms with it, because we… we think

408
01:09:09.160 --> 01:09:16.470
Marius Van Der Wijden (M): The, the… And justify the means, to some extent.

409
01:09:16.689 --> 01:09:24.340
Marius Van Der Wijden (M): And… and… and that's where we're heading. So, and we… we cannot…

410
01:09:24.520 --> 01:09:28.699
Marius Van Der Wijden (M): just ossify the state of Ethereum that is right now.

411
01:09:33.920 --> 01:09:35.240
Stefan Bratanov: And can do if…

412
01:09:45.620 --> 01:09:49.400
danceratopz: Thanks, Marius. Marius. Did anyone want to speak up?

413
01:09:49.660 --> 01:09:52.100
danceratopz: On those points further.

414
01:09:56.670 --> 01:10:05.860
Rahul (EF): On a… I had a question for Jango, I guess, on a more… Timely note, where…

415
01:10:06.400 --> 01:10:09.270
Rahul (EF): What are your next steps on the Glamsterdam?

416
01:10:09.390 --> 01:10:11.520
Rahul (EF): Breakages of your contract.

417
01:10:12.110 --> 01:10:16.320
Rahul (EF): Is it, something that can be fixed, or…

418
01:10:22.260 --> 01:10:27.940
Jango: So, yeah, it can be fixed in the current version of the protocol, and everything we do,

419
01:10:28.370 --> 01:10:38.260
Jango: tries to be the last version of the protocol. Every version that has, preceded has been a security, you know, like, the recent

420
01:10:38.650 --> 01:10:47.029
Jango: The year has been AI hardening of the protocol, as basically everyone's been doing, making sure we're the good guys to the punchline before bad guys can get there.

421
01:10:49.060 --> 01:10:56.730
Jango: And feeling really, really good about how everything was deployed in June, ready to scale, ready to pour gas onto growth.

422
01:10:57.110 --> 01:11:05.210
Jango: And now, having no confidence to, to even kind of,

423
01:11:05.580 --> 01:11:21.170
Jango: assure current users that they should continue running their business on the protocol. And they obviously want to run immutable… they want to run their business on immutable protocols because… for obvious reasons. If…

424
01:11:21.170 --> 01:11:40.580
Jango: there's, you know, there's bedrock underneath. They wouldn't want to do so if there wasn't, but, you know, Ethereum is the one beacon of potential hope that we may have bedrock to run business on top of. And I guess not just business, but campaigns, fundraising, legal defense funds, etc.

425
01:11:41.380 --> 01:11:51.179
Jango: If we were to redeploy, the funds at risk right now on the latest version are, you know, in the fringe category, about $200K.

426
01:11:51.350 --> 01:12:10.299
Jango: Not to mention, you know, developer costs, audit costs, deployment costs, etc, but still very, very meaningless compared to the great scope of Ethereum. The biggest thing is going to continue to be the confidence thing, which I think is really the value prop. It is the wealth that is

427
01:12:10.300 --> 01:12:16.470
Jango: Ethereum. And if I were to deploy another contract with, you know, adjustments that make

428
01:12:16.610 --> 01:12:21.609
Jango: try-catch composability work under glam,

429
01:12:23.480 --> 01:12:41.769
Jango: I'm then hesitant to still sell the same immutable hope and value prop that I think we can be, and I think we should strive to be, because of the same reasons, you know? So it's kind of this ongoing, doubt that then we're gonna then gonna

430
01:12:41.990 --> 01:12:50.870
Jango: cause, and I think there are certain precedents that have already been set, on breaking invariants that

431
01:12:50.990 --> 01:12:54.779
Jango: Were trade-offs nonetheless, but may have…

432
01:12:56.060 --> 01:13:14.509
Jango: not been as, as consequential to the theory that is Ethereum. I think composability is theoretically one of the most important things for app developers to build on the open internet, to try to make the open internet compete with private alternatives, which is, like, my mission. And…

433
01:13:14.690 --> 01:13:25.220
Jango: It's, you know, there's maybe, like, a check that can come through the door that can make the system whole and allow us to spend the next few months on redeploying on a next version.

434
01:13:25.220 --> 01:13:42.310
Jango: But does that then entail, finding ways to do so, and adding back governance, let's say, for, for, evolution? And the thing that I most despise is governance. Like, I think it is the fragile bit of… that I've learned doing Dow stuff for the past 5 years, is

435
01:13:42.400 --> 01:13:43.780
Jango: governance is…

436
01:13:43.900 --> 01:13:58.290
Jango: a huge, huge suck of energy and intellect, oftentimes. And what we really want is immutable revenue systems and, kind of tokens backed by revenue that just kind of work

437
01:13:58.400 --> 01:14:00.010
Jango: And…

438
01:14:00.340 --> 01:14:08.950
Jango: Yeah, it's tough to say, like, where I will sit from how I want to spend my life force, depending on how the decisions,

439
01:14:09.530 --> 01:14:23.090
Jango: on how the cards fall with regard to GLAM, but I can, like, definitively say, like, there's not a ton of money at stake relative to the billion, like, you know, hundreds of billions of dollars of ecosystem value.

440
01:14:29.630 --> 01:14:31.730
Rahul (EF): Okay, yeah, makes sense.

441
01:14:38.890 --> 01:14:42.590
danceratopz: I think we're really… I mean…

442
01:14:43.660 --> 01:14:50.640
danceratopz: I really appreciate your perspective, Jango, and I think the…

443
01:14:50.840 --> 01:14:58.160
danceratopz: I think I can speak for protocol developers when I say that, essentially, Ethereum is nothing without the applications that run on it, and…

444
01:14:58.650 --> 01:15:04.180
danceratopz: It's… without saying… Well, you don't even need to say it, but basically,

445
01:15:04.420 --> 01:15:10.620
danceratopz: It's of the utmost important that we can maintain your confidence, right? And this is why we've opened up this call today.

446
01:15:10.740 --> 01:15:12.990
danceratopz: And…

447
01:15:13.460 --> 01:15:22.160
danceratopz: I think we can move now to basically saying how we would plan to help and better align, like, the app layer and the protocol layer going forward.

448
01:15:22.450 --> 01:15:29.420
danceratopz: And then we can still keep time for Q&A afterwards. We can do it spontaneously, intermittently.

449
01:15:30.090 --> 01:15:37.110
danceratopz: But essentially, we… We hear you, and we feel your pain, and

450
01:15:38.110 --> 01:15:44.359
danceratopz: we can… we can definitely improve and do things better going forward, and we will strive to do this.

451
01:15:45.200 --> 01:15:53.890
danceratopz: perhaps Rahul, who's already spoken to Nicole today, would just like to speak up to his role at the EF and maybe explain what

452
01:15:54.530 --> 01:15:57.309
danceratopz: What work he's doing in order to guarantee this.

453
01:15:59.970 --> 01:16:10.559
Rahul (EF): Yeah, thanks, Dan. Hello. Just to be clear, after this, would be very keen to hear from apps about any questions you have around repricing or how to flex.

454
01:16:10.740 --> 01:16:18.649
Rahul (EF): But now, so, in the last year, the Ethereum Foundation created a new cluster called the Access Layer.

455
01:16:18.930 --> 01:16:24.479
Rahul (EF): And the access layer pretty much has one goal of better aligning

456
01:16:24.800 --> 01:16:27.999
Rahul (EF): The app layer with the protocol layer,

457
01:16:28.530 --> 01:16:39.199
Rahul (EF): That could be making it… making the app layer more crops, making it easier for apps to trust, or do trustless things on Ethereum, in the same way the protocol guarantees

458
01:16:39.310 --> 01:16:41.540
Rahul (EF): Censorship resistance and other things.

459
01:16:41.770 --> 01:16:49.640
Rahul (EF): Within Access Layer, there are two teams, the app enablement team and the builder enablement team, of which I'm a part of.

460
01:16:50.080 --> 01:16:54.169
Rahul (EF): Who pretty much work with external apps.

461
01:16:54.330 --> 01:17:00.579
Rahul (EF): Building on Ethereum, in certain cases L2s, and figuring out what's missing and what we can do.

462
01:17:01.080 --> 01:17:06.620
Rahul (EF): The builder enablement team actually just completed a survey with over 50 teams.

463
01:17:06.770 --> 01:17:09.470
Rahul (EF): that the EF usually doesn't…

464
01:17:09.830 --> 01:17:15.259
Rahul (EF): have close coordination with to understand more about the pain points of building on the EVM.

465
01:17:16.300 --> 01:17:23.939
Rahul (EF): this… the… the Glamsterdam Repricing outreach was actually a very good… Example, and, like,

466
01:17:24.820 --> 01:17:28.000
Rahul (EF): The use case of our resources in figuring out

467
01:17:28.430 --> 01:17:35.439
Rahul (EF): How we can reach out, not just to apps, but the direct engineering contacts, or the security contacts and apps.

468
01:17:35.840 --> 01:17:39.869
Rahul (EF): Together with SEAL and these teams, we…

469
01:17:40.620 --> 01:17:48.980
Rahul (EF): We feel quite confident that we have either good group chats with the right people, or have access to around 3,000 plus developers.

470
01:17:49.210 --> 01:17:51.370
Rahul (EF): Building on L1 itself.

471
01:17:51.940 --> 01:18:04.929
Rahul (EF): And as we found with chain security and Dedaub's help, like, more and more teams are getting affected, we could reach out to them and know that within one day, we'd get a lot of really good response from the right people.

472
01:18:05.400 --> 01:18:10.190
Rahul (EF): Also, I'll reiterate, like, the whole reason…

473
01:18:10.870 --> 01:18:14.730
Rahul (EF): we have decided to delay Glamsterdam was to put apps first.

474
01:18:15.030 --> 01:18:19.210
Rahul (EF): Something that at least was really nice for me was

475
01:18:19.430 --> 01:18:30.200
Rahul (EF): seeing that for Hegota and future hard forks, how much work is going on in replaying a bunch of transactions and doing a lot of analysis on how EIPs will affect

476
01:18:31.410 --> 01:18:39.710
Rahul (EF): And I think, like, for future conferences, we'll try to do more talks on what invariants may break, or what do app developers need to be aware of.

477
01:18:39.820 --> 01:18:47.140
Rahul (EF): multiple months before the hard fork, as opposed to delaying it to whenever the next Devcon is.

478
01:18:47.500 --> 01:18:53.540
Rahul (EF): Also, like, There are certain things that…

479
01:18:53.830 --> 01:18:57.730
Rahul (EF): So, Glamsterdam, in my opinion, is one of the most profound

480
01:18:58.230 --> 01:19:02.639
Rahul (EF): Hard forks for app developers, in the sense of…

481
01:19:04.030 --> 01:19:13.570
Rahul (EF): stacked too deep, and code sizes being increased were, like, two of the biggest pain points of building on Ethereum for a really, really long time, like, over 10 years.

482
01:19:13.900 --> 01:19:18.530
Rahul (EF): And both of those do get fixed. There are a bunch of other quality of life improvements.

483
01:19:18.750 --> 01:19:25.419
Rahul (EF): Obviously, repricing was one of those things that we tried to do for scaling and reducing transaction costs.

484
01:19:25.620 --> 01:19:34.309
Rahul (EF): But we found out there were some security incidents about it, but we are hoping to make it such that every hard fork, we improve more and more of the quality of life improvements.

485
01:19:34.430 --> 01:19:40.219
Rahul (EF): All of that said, all I want app teams to know is

486
01:19:40.390 --> 01:19:46.890
Rahul (EF): feel free to either hop on ACD Calls or… Reach out to… Any one of us?

487
01:19:47.180 --> 01:20:02.810
Rahul (EF): If you have any feedback and stuff, and unrelated to this call, if you have more pointed feedback on how we can improve the communication for repricing, feel free to schedule a call with me. You can reach me out at glamstdammit at ethereum.org.

488
01:20:03.070 --> 01:20:05.829
Rahul (EF): I actively monitor that email.

489
01:20:06.590 --> 01:20:11.420
Rahul (EF): But yeah, if anyone else has anything else, feel free to jump in as well.

490
01:20:16.620 --> 01:20:20.930
danceratopz: Thanks, Rahul. Did anyone… Sorry?

491
01:20:21.150 --> 01:20:26.290
Marius Van Der Wijden (M): Yeah, there was, there was, a proposal…

492
01:20:26.430 --> 01:20:39.730
Marius Van Der Wijden (M): from, from donnoh, to create an EIP with all of the invariants, and, I… I don't know if the… if, like, creating an EIP is the right

493
01:20:40.070 --> 01:20:47.520
Marius Van Der Wijden (M): way of… of specifying the invariants, but I do feel like there…

494
01:20:47.770 --> 01:20:52.899
Marius Van Der Wijden (M): There needs to be some sort of documentation about

495
01:20:53.510 --> 01:21:04.549
Marius Van Der Wijden (M): The invariants that we… that we are sure will never change, or as sure as we can be, that they will never change,

496
01:21:04.930 --> 01:21:15.519
Marius Van Der Wijden (M): And the invariants that we are… that we are very sure that will change, in the future, and also links to

497
01:21:15.850 --> 01:21:25.280
Marius Van Der Wijden (M): Why they will change, and, and then, like, yeah, just… giving a…

498
01:21:25.490 --> 01:21:35.930
Marius Van Der Wijden (M): Giving a better sense, or a better look into the current thinking of core developers, around,

499
01:21:36.530 --> 01:21:42.509
Marius Van Der Wijden (M): What is necessary to be changed, what is un… like, what is… what… what is…

500
01:21:43.900 --> 01:21:47.970
Marius Van Der Wijden (M): Likely not going to change, and,

501
01:21:48.290 --> 01:21:53.030
Marius Van Der Wijden (M): Why we think certain things have to change in the future.

502
01:22:00.750 --> 01:22:05.580
danceratopz: Yeah, thanks, Marius. Mario, are you still in the call?

503
01:22:07.140 --> 01:22:07.640
Mario Vega: I am here.

504
01:22:07.640 --> 01:22:10.750
danceratopz: I just wondered if you wanted to speak up to some of the efforts

505
01:22:11.310 --> 01:22:15.020
danceratopz: that we've just started doing, as part of the specs team.

506
01:22:15.020 --> 01:22:15.460
Mario Vega: Yes.

507
01:22:15.460 --> 01:22:16.190
danceratopz: Awesome.

508
01:22:16.190 --> 01:22:18.079
Mario Vega: Definitely. Yeah, thanks.

509
01:22:18.370 --> 01:22:19.619
Mario Vega: Thanks, Dan, can you hear me?

510
01:22:20.240 --> 01:22:21.459
danceratopz: Yep, I can hear you. Thanks, Frank.

511
01:22:21.460 --> 01:22:24.520
Mario Vega: Nice, nice. Yeah, so, my name is Mario, I'm…

512
01:22:24.540 --> 01:22:41.490
Mario Vega: Part of the specifications team, for the execution layer, we play a role in testing all the protocol changes that go into the fork, namely every change in Amsterdam, and basically also every single change in Hegota.

513
01:22:41.500 --> 01:22:53.580
Mario Vega: And, yeah, I would like to start off with saying that EIP-8037 has definitely left a mark on everyone that is testing, and that is, like, working on the specifications.

514
01:22:53.650 --> 01:23:02.830
Mario Vega: I feel like now, more than ever, we cannot simply just eat up any specification change that comes our way, and just basically

515
01:23:02.930 --> 01:23:12.899
Mario Vega: not see it through the lens of the application layer, because it's, it's been, like, we've, we, our work is basically just,

516
01:23:13.000 --> 01:23:24.279
Mario Vega: to test the protocol invariants and make sure that they don't break, they don't… we don't have consensus changes, but also, like, I feel like it's also on our side to…

517
01:23:24.280 --> 01:23:42.010
Mario Vega: foresee any of these issues that might come up, and for that, we have a couple of ideas of how we can do a better job in the future, with the specification changes, and one of them is trying to see that the invariants that we tell the users are definitely not gonna break.

518
01:23:42.010 --> 01:23:54.719
Mario Vega: So this would be on top of, like, us, testing for protocol changes, also, like, paying attention on these things. We're definitely gonna,

519
01:23:55.000 --> 01:24:00.239
Mario Vega: have a more focused mindset on the user, and I would definitely, like.

520
01:24:00.450 --> 01:24:19.030
Mario Vega: for, like, us to see, specifically the CFI and the, well, the consider for inclusion EIPs for the next forks, I think we have to rethink and really clear… clearly think through if there is any changes that we are not foreseeing. Of course, we're not always going to be able to

521
01:24:19.030 --> 01:24:23.040
Mario Vega: Like, predict every single way every user,

522
01:24:23.040 --> 01:24:36.570
Mario Vega: does things in Ethereum. That's basically a very, very difficult thing to do, but at least we… we can, say that we are gonna try to, put a…

523
01:24:36.800 --> 01:24:45.350
Mario Vega: A better effort on trying to just basically pull tools in our disposal to better foresee these things in the spec changing.

524
01:24:45.420 --> 01:24:57.109
Mario Vega: Other thing is that, I would like to, for us to not… I mean, it's very easy for the specs to change, right? It's, like, one… a simple line of code sometimes.

525
01:24:57.160 --> 01:25:09.550
Mario Vega: But I would say that now, more than ever, we really have to stop and think, stop… if a test breaks, we have to stop and think why it breaks, and not simply let the AI just go

526
01:25:09.550 --> 01:25:22.800
Mario Vega: go nuts and fix it and make no mistakes. I think we should… it is part of our job to stop and think about every single test that breaks with every single specification change.

527
01:25:23.030 --> 01:25:24.520
Mario Vega: Yeah.

528
01:25:24.750 --> 01:25:27.330
Mario Vega: in the, in the ACDT…

529
01:25:27.510 --> 01:25:44.640
Mario Vega: Agenda, we linked two examples. I think those are not final, I would say, so we have, it's just ideas, but I think we're gonna keep working on iterating on those to make sure that we have something different in the specs, and some new tools to help us out with this in the future, yeah.

530
01:25:49.390 --> 01:25:50.280
danceratopz: Awesome.

531
01:25:50.410 --> 01:25:51.820
danceratopz: Thanks a lot.

532
01:25:52.870 --> 01:26:06.029
danceratopz: Mario, for sharing that. I can just drop, so I did… thanks a lot, Ryan, for the question in the chat, and there's definitely action items, and there's definitely work ongoing to communicate better and investigate these things earlier.

533
01:26:06.280 --> 01:26:08.430
danceratopz: In future work, and…

534
01:26:08.790 --> 01:26:17.789
danceratopz: basically, as part of the specs team, we're obviously… we write the… well, just so you know, we write the, basically the canonical sort of reference specification in Python.

535
01:26:17.930 --> 01:26:31.440
danceratopz: Which should be easy to read. And then we use, this specification to generate test vectors, that are used to, verify client implementations. That's basically our work. So we're… we're obviously very close,

536
01:26:31.750 --> 01:26:48.680
danceratopz: to the specs, without… needless to say. I just copy and pasted this comment in the chat. Basically, there's a couple of PRs here, they're open at the moment. One's from Toni, who's, works on architecture in the call, and another one is from Felipe A. on the specs team.

537
01:26:48.970 --> 01:26:52.789
danceratopz: Mario has another PR open that can be used

538
01:26:53.250 --> 01:27:02.170
danceratopz: To verify correct behavior against our canonical specs. So we're directly taking measures right now to… to try and improve the situation.

539
01:27:02.840 --> 01:27:05.570
danceratopz: And…

540
01:27:05.740 --> 01:27:13.199
danceratopz: Anders in the chat actually said that defining… so Anders is a protocol researcher, so he's at the other end, so if you have

541
01:27:13.640 --> 01:27:21.350
danceratopz: The users, the app layer, the client Devs, and you have the specs, which is our team, and behind that, you have,

542
01:27:21.480 --> 01:27:36.159
danceratopz: the researchers that are basically coming up with the best way to try and improve Ethereum. They also are grateful to have these specs implemented in Python and concrete, and Anders all just said that basically specifying these invariants also helps their work. So…

543
01:27:36.260 --> 01:27:38.760
danceratopz: This is something that really helps,

544
01:27:39.240 --> 01:27:48.040
danceratopz: researchers and programmers across the ecosystem to do their job better. So it's in all of our interests, and obviously, without saying, Ethereum's

545
01:27:48.290 --> 01:27:51.479
danceratopz: I would also just add,

546
01:27:51.650 --> 01:28:03.329
danceratopz: Oh, sorry, I just copy-pasted the wrong thing. Let me… let me grab the other thing I wanted to paste. So, pcaversaccio, who's, like.

547
01:28:04.130 --> 01:28:16.010
danceratopz: An EF board member from earlier this year, basically has written up his own take on this on Twitter, and also dropped it on the… on today's call issue. I shared the link in the chat.

548
01:28:16.430 --> 01:28:19.220
danceratopz: He outlines, basically,

549
01:28:20.370 --> 01:28:32.319
danceratopz: a whole list of things that we can do to improve the situation, and I think we've spoken to most of them already in the call today, but he calls for us to formalize invariants

550
01:28:32.650 --> 01:28:36.810
danceratopz: To incorporate them into the specifications, test them continuously.

551
01:28:37.150 --> 01:28:41.529
danceratopz: Test deployed applications and make compatibility explicit.

552
01:28:41.790 --> 01:28:44.999
danceratopz: And, certainly the…

553
01:28:45.510 --> 01:28:53.740
danceratopz: first three points, are being actively worked on in our repository, which is, Ethereum execution specs. So,

554
01:28:54.270 --> 01:28:57.680
danceratopz: We've… we've heard the call, and we're working towards it.

555
01:29:13.330 --> 01:29:18.680
danceratopz: Okay, does anyone have anything else to add on that point?

556
01:29:19.440 --> 01:29:21.650
danceratopz: Or any other open questions?

557
01:29:29.880 --> 01:29:38.300
danceratopz: There's an awful lot of chatter in the chat, and it's a bit hard to keep up with, but if your question's really important, then please raise it. Very happy to have it.

558
01:29:39.250 --> 01:29:43.610
danceratopz: Floated in the… In the chat.

559
01:29:44.840 --> 01:29:45.780
danceratopz: Nicole?

560
01:29:46.630 --> 01:29:49.110
danceratopz: Trying to keep up right now.

561
01:29:49.900 --> 01:29:57.569
danceratopz: So the last question is from Justin, who's a protocol dev at Besu. Are we still shipping 8037?

562
01:29:58.520 --> 01:30:00.470
danceratopz: I think…

563
01:30:01.080 --> 01:30:14.180
danceratopz: That's… that's not a decision that I would like to make in this call today. I think we need to disseminate the information a little bit more, and I think chain security and Dedaub will share their reports.

564
01:30:14.540 --> 01:30:19.830
danceratopz: And we'll heal back and be in touch with protocol, with dApp protocols.

565
01:30:20.570 --> 01:30:23.580
danceratopz: And I think we can…

566
01:30:23.890 --> 01:30:29.580
danceratopz: But from what I can take from the call, I think… that…

567
01:30:31.320 --> 01:30:38.070
danceratopz: This is not a great situation to be in, clearly, and we can do better, and we're striving to do better in the future.

568
01:30:38.940 --> 01:30:41.570
danceratopz: But I think…

569
01:30:42.460 --> 01:30:52.339
danceratopz: that it's quite difficult to… to remove 8037 without a serious delay to Glamsterdam, and I'm talking, sort of, of the order of potentially half a year.

570
01:30:53.080 --> 01:30:56.389
danceratopz: It's also a very important,

571
01:30:56.740 --> 01:31:01.400
danceratopz: change to make in order to keep state growth sustainable on Ethereum.

572
01:31:02.630 --> 01:31:08.190
danceratopz: So… I…

573
01:31:08.350 --> 01:31:15.720
danceratopz: I really appreciate the app developers to come here today and express their opinion, and in particular Jango for speaking up on behalf of them.

574
01:31:15.900 --> 01:31:17.310
danceratopz: But we will…

575
01:31:18.110 --> 01:31:25.350
danceratopz: We'll move forward, and I think Ansgar's got his hand up. Ansgar's ACDE moderator, the execution layer, core Devs call.

576
01:31:25.510 --> 01:31:30.770
danceratopz: I think that's something that they'll discuss on Thursday. Ansgar, would you like to chime in?

577
01:31:31.250 --> 01:31:32.539
Ansgar Dietrichs: Yeah, can you hear me?

578
01:31:33.090 --> 01:31:34.060
danceratopz: Yep, I can hear you.

579
01:31:34.060 --> 01:31:40.689
Ansgar Dietrichs: Perfect. Yeah, I mean, I just wanted to basically, like, plus one what you said just now. I think it is clearly, like, a…

580
01:31:41.060 --> 01:31:49.990
Ansgar Dietrichs: not ideal situation that we find ourselves in, and I do think this has kind of, like, been iterated a few times today already. I think, A.

581
01:31:50.100 --> 01:31:55.270
Ansgar Dietrichs: things definitely didn't go ideally to even get to this place, but I think more importantly, also, like.

582
01:31:55.480 --> 01:32:09.880
Ansgar Dietrichs: this is not going to be the last hard fork that, whether it now goes ahead or not, that will make changes to the execution system. Some of these will just be inevitable with quantum and everything coming up.

583
01:32:09.990 --> 01:32:20.880
Ansgar Dietrichs: And we just need to get to basically now take on that challenge of just being much more proactive, much more actively communicating, working with the ecosystem, like…

584
01:32:21.170 --> 01:32:34.559
Ansgar Dietrichs: making sure that for future hard forks, everyone is on the same page. Early on, we understand, kind of, like, the potential things that will be impacted, how we can minimize that while there's still room to change EIPs, all these kind of things, so I think

585
01:32:34.560 --> 01:32:47.790
Ansgar Dietrichs: all of this, we are all coming out of this situation with this kind of shared conviction on doing this, and I think today's call is also a great step in that direction with so much participation, and I think we all agree that that needs to improve.

586
01:32:48.470 --> 01:32:53.179
Ansgar Dietrichs: In… in terms of Glamsterdam itself, I also just wanted to…

587
01:32:53.340 --> 01:33:02.920
Ansgar Dietrichs: to say that today was not meant to be about decisions. We have the ACDE call on Thursday,

588
01:33:04.710 --> 01:33:09.139
Ansgar Dietrichs: I just also wanted to stress that there is still a strong…

589
01:33:10.110 --> 01:33:26.440
Ansgar Dietrichs: desire, at least, to try to move ahead with the fork as it stands today. And that is not coming out of an ignorance of the cost, both in terms of direct affected projects, but also, like, the costs of, say, weakening the trust in the invariants of Ethereum.

590
01:33:26.440 --> 01:33:31.970
Ansgar Dietrichs: But it's coming just from a place of, like, just the trade-off, and the very real cost, like…

591
01:33:32.060 --> 01:33:39.869
Ansgar Dietrichs: especially given that I think Ethereum is really trying to now move fast with the post-quantum transition, and the execution layer side of this is the most

592
01:33:39.870 --> 01:34:04.520
Ansgar Dietrichs: urgent, because obviously, on the Consensus layer, once you transition, you transition, but on the execution layer, once you allow users to transition their wallets and their accounts, it will take a very long time for all users to do that, because you also don't want to rush users, right? Like, you want the entire toolchain, the wallets, everyone to be fully ready for this transition, so basically, like, having the starting point of this, of this process as soon as possible, with basically the protocol supporting full post-quantum accounts.

593
01:34:04.750 --> 01:34:19.930
Ansgar Dietrichs: I think it's really, really important in this moment in time. And so, because of that, just the cost of making changes to Glamsterdam is just so, so, so high, because of that specific kind of context that it's in. Doesn't mean that

594
01:34:20.110 --> 01:34:35.709
Ansgar Dietrichs: it's a locked-in decision that we won't, 100%, but I'm just saying, my plea, basically, would be for everyone affected here, directly or indirectly, to at least… to basically, like, I think we should try very hard to see, can we still go ahead with Amsterdam as is?

595
01:34:36.240 --> 01:34:56.219
Ansgar Dietrichs: the call on Thursday will be focused on that question. If at all possible, I really think it would be incredibly valuable for Ethereum at this moment in time, but then again, only combined with us learning the right lessons out of this going forward. So, that's all I wanted to say. But yeah, on Thursday, obviously, we will have the full room to go into this, and there's no

596
01:34:56.220 --> 01:34:58.820
Ansgar Dietrichs: Pre-made decision here in any way.

597
01:35:04.200 --> 01:35:06.290
danceratopz: Thanks a lot, Ansgar, for clarifying that.

598
01:35:06.410 --> 01:35:13.320
danceratopz: Hubert from ChainSecurity would like to ask your questions? Are you… are you ready for now?

599
01:35:13.490 --> 01:35:21.669
Hubert (ChainSecurity): Yeah, so, the… my question is, what prevents the Ethereum Foundation? I mean, we now have seen

600
01:35:21.790 --> 01:35:26.260
Hubert (ChainSecurity): Two different teams, kind of doing this kind of impact analysis.

601
01:35:26.430 --> 01:35:33.320
Hubert (ChainSecurity): with different tools, we've, previously seen Ethlabs do a great job with this, impact analysis.

602
01:35:33.680 --> 01:35:37.250
Hubert (ChainSecurity): What prevents the Ethereum Foundation from maintaining this

603
01:35:37.650 --> 01:35:46.299
Hubert (ChainSecurity): capability to do this impact analysis. Whenever they do any kind of protocol change, because.

604
01:35:46.300 --> 01:35:48.690
Trent Van Epps: It's really, it's really hard to hear you, you.

605
01:35:48.920 --> 01:35:51.120
Trent Van Epps: If you can move closer to your mic or something.

606
01:35:51.750 --> 01:35:52.540
Hubert (ChainSecurity): Sorry.

607
01:35:52.540 --> 01:35:58.870
Marius Van Der Wijden (M): I think I get what you're trying to say. You are asking why we don't have this capability in-house.

608
01:35:59.050 --> 01:36:14.629
Marius Van Der Wijden (M): And the answer to this is we started building up this capability in-house, but unfortunately, due to budget restrictions and reorganization, we had to let people that were working on this go.

609
01:36:19.030 --> 01:36:21.660
Hubert (ChainSecurity): Thanks, sorry for my audio troubles.

610
01:36:22.460 --> 01:36:26.710
danceratopz: No problem. I think we heard it. Thanks, Marius. Toni?

611
01:36:27.810 --> 01:36:39.140
Toni Wahrstaetter: Yeah, just to add to that, so this was something that was, I mean, originally initiated by Ethlabs, by Francesco, and led us to discovering, like, all the…

612
01:36:39.270 --> 01:36:46.710
Toni Wahrstaetter: invariants and all the protocols that were, affected initially. So we did this for the first time, and we never had, like.

613
01:36:46.870 --> 01:36:49.040
Toni Wahrstaetter: Issues that were that deep.

614
01:36:49.790 --> 01:37:04.360
Toni Wahrstaetter: Of course, AI also involved a lot in the recent years, making such broad analysis even possible, and I think in the future, or I'm very certain, that we will just keep doing exactly that, like, letting agents,

615
01:37:04.860 --> 01:37:11.610
Toni Wahrstaetter: Manually parse through all the contracts and find issues. And at the same time, we should probably also engage

616
01:37:12.080 --> 01:37:18.889
Toni Wahrstaetter: Security auditing companies earlier on, because they have also a lot of insights into coding patterns.

617
01:37:19.120 --> 01:37:23.729
Toni Wahrstaetter: Yeah, I think this is… will already be a great first step to avoid.

618
01:37:23.830 --> 01:37:26.149
Toni Wahrstaetter: Such issues in the future.

619
01:37:30.990 --> 01:37:40.100
danceratopz: Thanks, Toni. Tanguy echoed the question in the chat, by the way, from Hubert. So what prevents us from maintaining this? And yeah, I think that's been answered, that we can… we can move forward to…

620
01:37:40.870 --> 01:37:44.160
danceratopz: maintaining this infrastructure in the EF as well.

621
01:37:44.690 --> 01:37:46.160
danceratopz: Daniel?

622
01:37:47.120 --> 01:37:58.770
Daniel Gretzke: Hi, Daniel from Uniswap here. I understand that there is a strong desire to just keep going, because already a lot of work has gone in, it causes a lot of delays.

623
01:37:59.040 --> 01:38:11.559
Daniel Gretzke: But I really want to emphasize that 8037 should definitely be reconsidered, especially the handling of the gas reservoir, because this just fundamentally breaks

624
01:38:11.720 --> 01:38:17.439
Daniel Gretzke: Some assumptions that we have basically fundamentally catch… Where breaks try catching.

625
01:38:17.640 --> 01:38:29.379
Daniel Gretzke: And also introduces a bunch of second-order effects that create a very difficult UX for the end users of the applications on Ethereum.

626
01:38:29.570 --> 01:38:37.130
Daniel Gretzke: I do… for me, you know, it's totally fine that gas assumptions

627
01:38:37.340 --> 01:38:45.539
Daniel Gretzke: are allowed to break. And we should not, you know, be using very conservative, hard-coded gas.

628
01:38:45.660 --> 01:38:48.229
Daniel Gretzke: I think everyone's aware of this,

629
01:38:48.400 --> 01:38:50.840
Daniel Gretzke: I do think that the main issue is that

630
01:38:51.160 --> 01:39:04.479
Daniel Gretzke: when we add implicit branching based on the transaction gas that a user sets, or these kinds of things, I would… I would argue that this would fundamentally break how we

631
01:39:05.530 --> 01:39:08.229
Daniel Gretzke: Build contracts, and…

632
01:39:08.510 --> 01:39:15.579
Daniel Gretzke: Yeah, so I just wanted to emphasize that this really needs to be rethought, and all kind of also edge cases handled.

633
01:39:16.030 --> 01:39:18.899
Daniel Gretzke: Before actually considering moving forward with it.

634
01:39:23.260 --> 01:39:25.529
Jango: Needless to say, big plus one.

635
01:39:27.470 --> 01:39:29.359
danceratopz: Thanks, Tanya, for speaking up.

636
01:39:29.550 --> 01:39:31.830
danceratopz: Toni, you've got your hands up.

637
01:39:32.010 --> 01:39:32.500
Toni Wahrstaetter: Yeah.

638
01:39:32.500 --> 01:39:33.640
danceratopz: Would you want to take Simon?

639
01:39:34.080 --> 01:39:39.340
Toni Wahrstaetter: Yeah, what I can add to this discussion is that from the spec side, we have been looking into

640
01:39:39.370 --> 01:39:54.979
Toni Wahrstaetter: ways how to resolve the situation, and the problem there is basically that there isn't really a nice solution that fits everyone. So, doing small tweaks to the specs would introduce, like, other broken invariants.

641
01:39:55.170 --> 01:40:12.670
Toni Wahrstaetter: And the underlying… just to provide some background here, one of the issues why 8037 was even needed was because, for security reasons, we had, for example, the transaction limit cap EIP-7825 in the last fork.

642
01:40:12.900 --> 01:40:20.839
Toni Wahrstaetter: And in order to even be able to deploy Max Di's contract, which app developers were chasing us for years.

643
01:40:20.870 --> 01:40:40.779
Toni Wahrstaetter: To ship, and in order to deploy those, we didn't even have enough gas per transaction available anymore to deploy those. So, for example, changes like just removing 8037 would also mean we will not be able to deploy those big contracts anymore, and other downsides that will appear downstream.

644
01:40:40.920 --> 01:40:49.659
Toni Wahrstaetter: So, I think we should be very hesitant in changing the specs, just because it comes with huge cost, especially when it comes to security right now.

645
01:40:49.820 --> 01:40:52.990
Toni Wahrstaetter: And I would prefer, like, while…

646
01:40:54.320 --> 01:40:59.499
Toni Wahrstaetter: like, not closing the topic right now and moving on with scheduling Hoodi, we should still…

647
01:40:59.710 --> 01:41:04.979
Toni Wahrstaetter: Be aware that any changes we do to the specs right now come with a very big cost.

648
01:41:10.190 --> 01:41:17.859
Daniel Gretzke: I mean, generally, I'd be… I'd prefer to wait for an increased smart contract size limit, if that means

649
01:41:18.270 --> 01:41:27.339
Daniel Gretzke: you know, we can fix all the other issues. Like, all the other issues are way more important to me than the small contract size limit, which is nice, but…

650
01:41:27.860 --> 01:41:32.139
Daniel Gretzke: I think the other part would be worse, just speaking as an app dev.

651
01:41:32.940 --> 01:41:40.699
Toni Wahrstaetter: Yeah, no, I totally agree with you. This was just one of the examples that is, like, obviously interacting with this topic.

652
01:41:40.810 --> 01:41:51.720
Toni Wahrstaetter: But there are many, like, things where you can do small tweaks to the specs right now, for example, refunding into the reservoir itself instead of execution gas, or moving the refund.

653
01:41:51.880 --> 01:42:03.389
Toni Wahrstaetter: after that transaction. But in the end, you would just… like, at least that's what we have found so far. You would just introduce different invariants, and most of the things that are breaking

654
01:42:03.550 --> 01:42:10.029
Toni Wahrstaetter: like, I think 90% of them are just because of state becoming more expensive, and

655
01:42:10.530 --> 01:42:18.690
Toni Wahrstaetter: There's basically nothing we can do about that, because in the end, it's just a trade-off between keeping hardware requirements low, and at the same time.

656
01:42:18.930 --> 01:42:20.939
Toni Wahrstaetter: Limiting state growth.

657
01:42:21.740 --> 01:42:29.099
Toni Wahrstaetter: So it's just a very difficult issue to resolve, and for those hard-coded gas, problems.

658
01:42:29.360 --> 01:42:38.579
Toni Wahrstaetter: There might just not be a good solution, except us finding an actual sustainable solution, how we are handling state, but we don't have that, unfortunately.

659
01:42:51.060 --> 01:42:57.569
danceratopz: Thanks for the, question, Daniel, and thanks for answering, Toni. I think, Daniel, I think the problem is it…

660
01:42:57.720 --> 01:43:03.639
danceratopz: What we're doing is a trade-off, and it's a hard problem to solve.

661
01:43:07.950 --> 01:43:13.929
danceratopz: and in general, as Dragan wrote in the chat, there's basically no silver bullet to solving this problem, and that…

662
01:43:14.220 --> 01:43:21.200
danceratopz: In order to… to mitigate Ethereum state growth, like, any kind of 2D gas version is likely to break something.

663
01:43:27.700 --> 01:43:34.069
danceratopz: So it's… it's… it's a… it's a difficult problem to solve, basically, is the main message, and

664
01:43:34.250 --> 01:43:37.630
danceratopz: Unfortunately, there's… there's some trade-off to be… to be made there.

665
01:43:37.980 --> 01:43:39.549
danceratopz: Neville?

666
01:43:41.650 --> 01:43:51.440
neville (Dedaub): Hey, so after looking at a bunch of contracts, and also, like, speaking to affected developers, one thing we realized is,

667
01:43:51.800 --> 01:43:54.729
neville (Dedaub): There's no reliable way for,

668
01:43:55.220 --> 01:44:01.710
neville (Dedaub): For callers to distinguish whether a callee has reverted due to natural causes.

669
01:44:01.960 --> 01:44:04.640
neville (Dedaub): Or whether it was out of gas.

670
01:44:04.930 --> 01:44:07.739
neville (Dedaub): Invalid opcode or something like that. So…

671
01:44:08.040 --> 01:44:12.350
neville (Dedaub): This is something for the core devs to contemplate about, but

672
01:44:12.530 --> 01:44:19.209
neville (Dedaub): You know, consider including, like, an opcode by which you can introspect the last call status.

673
01:44:19.770 --> 01:44:24.240
neville (Dedaub): And there's… A lot more to discuss about that.

674
01:44:24.830 --> 01:44:28.090
neville (Dedaub): But this is something that changes with Glamsterdam.

675
01:44:32.350 --> 01:44:35.659
Marius Van Der Wijden (M): Yeah, someone already put in chat that,

676
01:44:35.990 --> 01:44:47.979
Marius Van Der Wijden (M): like, it's… it's a common theme that whenever I think of something, I see a post from… from Martin Holst Swende from, like, 5 years ago, who discussed the same thing.

677
01:44:47.980 --> 01:44:56.820
Marius Van Der Wijden (M): And one of his posts was about how it would be really nice for a call to give the revert reason.

678
01:44:57.760 --> 01:45:11.379
Marius Van Der Wijden (M): Unfortunately, what we have right now is that, zero means good, sorry, 0 means bad, and 1 means good, and so you cannot really introduce other

679
01:45:11.720 --> 01:45:15.960
Marius Van Der Wijden (M): Types, because there's a bunch of contracts that, like, hack

680
01:45:16.600 --> 01:45:23.210
Marius Van Der Wijden (M): Whether… whether it's non-zero, whether the return value is non-zero, and… and branch on that.

681
01:45:25.390 --> 01:45:37.300
Marius Van Der Wijden (M): I think there's… there might be a way of introducing some new operation to get… similar to… to, return data, to get the…

682
01:45:37.870 --> 01:45:43.269
Marius Van Der Wijden (M): To get a revert reason for the previous subcall.

683
01:45:43.470 --> 01:45:48.610
Marius Van Der Wijden (M): It's something that we can…

684
01:45:49.950 --> 01:45:55.920
Marius Van Der Wijden (M): discuss and talk about, but adding that to a Glamsterdam hard fork again would

685
01:45:56.290 --> 01:46:04.019
Marius Van Der Wijden (M): increase the time to ship it by, like, 3 to 4 months, I would say. So, yeah.

686
01:46:14.260 --> 01:46:15.210
danceratopz: Thanks, Grace.

687
01:46:21.200 --> 01:46:30.769
danceratopz: So, Milos just asked in the chat, I saw… Milos have to apologize, I only saw your comment on the issue, like, halfway through the call.

688
01:46:31.120 --> 01:46:36.740
danceratopz: So if you wanted to step up and make your point, you're very welcome to do that now.

689
01:46:38.160 --> 01:46:43.869
Milos | enso.build: Oh, yeah, no problem. No, here I was just asking,

690
01:46:44.120 --> 01:46:49.109
Milos | enso.build: Because, like, there seems there are a lot of comments on, like, even possible solution.

691
01:46:49.500 --> 01:46:53.240
Milos | enso.build: But the response is, that…

692
01:46:53.680 --> 01:46:59.079
Milos | enso.build: it can't be done in a reasonable time frame, which, I mean, is understandable, it takes time.

693
01:46:59.250 --> 01:47:07.620
Milos | enso.build: But what will be the issue, of postponing by a couple of months and address everything the best that we can?

694
01:47:12.490 --> 01:47:21.870
Marius Van Der Wijden (M): Well, to be honest, like, even the… this opcode to get the revert reason will not address anything, because it is…

695
01:47:21.940 --> 01:47:35.910
Marius Van Der Wijden (M): a new opcode, and it will not fix any contracts that are broken with this right now, and people would need to redeploy anyway, and so it's kind of pointless to debate this. But in general, I think there's…

696
01:47:36.560 --> 01:47:37.860
Marius Van Der Wijden (M): that is…

697
01:47:38.150 --> 01:47:56.630
Marius Van Der Wijden (M): A general feeling around the Core Devs is that we have to move very deliberately and very quickly in the next coming years, because of advanced, advancements in, quantum, computers and…

698
01:47:56.680 --> 01:48:01.819
Marius Van Der Wijden (M): In AI in order to keep Ethereum secure. So I think we don't…

699
01:48:02.050 --> 01:48:04.579
Marius Van Der Wijden (M): Really don't have any time to waste.

700
01:48:11.030 --> 01:48:23.269
danceratopz: Yeah, thanks, Marius. Ansgar also answered Milos in the chat. So yeah, that… to Jango, I think, you also received a… received a response. And yeah, basically, Glam is a precondition, so…

701
01:48:23.510 --> 01:48:27.080
danceratopz: To answer that question, Hegota?

702
01:48:27.460 --> 01:48:33.010
danceratopz: Which is the next fork after Glamsterdam, which is already in active development, and there's been…

703
01:48:33.420 --> 01:48:38.279
danceratopz: several devnets already is actually built on top of Glamsterdam, so if we…

704
01:48:38.800 --> 01:48:41.609
danceratopz: to undo Glamsterdam's work with not only

705
01:48:42.150 --> 01:48:46.199
danceratopz: I mean, we can't undo Glamsterdam, we need some parts of Glamsterdam.

706
01:48:46.390 --> 01:48:54.249
danceratopz: But if we decided to take parts of it out, we'd have to redo the work for Hegota as well. It's… it's an expensive change at this point in time.

707
01:48:54.700 --> 01:49:04.190
danceratopz: And just to echo the fact that, We're… we're basically putting, Quantum and AI,

708
01:49:04.780 --> 01:49:12.869
danceratopz: acceleration, basically at the core of our, you know, like, our security upgrades, and we want to keep Ethereum secure and make sure that,

709
01:49:13.110 --> 01:49:20.579
danceratopz: with the acceleration of mathematics, thanks to AI, that Ethereum can stay safe going forward. And I think that's…

710
01:49:21.020 --> 01:49:23.650
danceratopz: One part that definitely,

711
01:49:24.100 --> 01:49:27.879
danceratopz: influences this decision. It's… there's a lot of moving parts.

712
01:49:38.450 --> 01:49:49.470
danceratopz: All right, we are actually already over time. Given the importance of a discussion, I did not want to rush it to a conclusion. Thanks a lot to everyone who could stay around.

713
01:49:51.110 --> 01:49:58.730
danceratopz: Before we chime off, is there any other questions, or is there anything else that someone wanted to share on the call today?

714
01:49:59.380 --> 01:50:00.939
danceratopz: Before we move forward?

715
01:50:13.490 --> 01:50:18.269
danceratopz: We did have two other agenda topics. Justin.

716
01:50:18.430 --> 01:50:21.070
danceratopz: Is it too late to discuss your question?

717
01:50:21.870 --> 01:50:25.479
jtraglia: Yeah, I believe it's too late. We'll need more time to discuss this later.

718
01:50:26.780 --> 01:50:28.660
danceratopz: Okay, I'm very sorry about that.

719
01:50:30.440 --> 01:50:36.220
danceratopz: for Jochem, Jochem's got a short announcement to make that I think

720
01:50:36.620 --> 01:50:41.990
danceratopz: We can still squeeze in to the last minute, I think that's worth doing. Jochem, did you want to speak up, or…

721
01:50:42.570 --> 01:50:54.689
jochem-brouwer: Yes, let me quickly do that. So, as you all might have seen, we are seeing a large influx of new EIPs and ERCs being created on the EIPs on ERC side.

722
01:50:54.810 --> 01:50:58.390
jochem-brouwer: likely due to all the, well, the AI boom, I would say.

723
01:50:58.490 --> 01:51:17.139
jochem-brouwer: We are thinking about integrating the AI into the review process, but the question is, how do we do this without decreasing the quality and the trust in EIPs? This is a topic on the EIPIP, which is the EIP Improvement Proposals call on Wednesday. I will link that here.

724
01:51:17.280 --> 01:51:23.610
jochem-brouwer: And I would invite anyone who has ideas on, integrating AI to the scope.

725
01:51:23.830 --> 01:51:24.880
jochem-brouwer: Yeah, thank you.

726
01:51:29.570 --> 01:51:30.360
danceratopz: Awesome.

727
01:51:30.480 --> 01:51:37.129
danceratopz: Thanks for sharing that, Jochem. Please show up to the call on Wednesday if you want to talk about the EIP process and AI tooling.

728
01:51:37.550 --> 01:51:46.930
danceratopz: Alright, then in that case, I see there's still chatty going on in the chat, but,

729
01:51:48.240 --> 01:51:57.610
danceratopz: I believe it's time to call… draw this call to a close, and I would like to thank everyone for showing up and sharing their perspectives.

730
01:51:58.180 --> 01:51:58.950
danceratopz: It's…

731
01:51:59.150 --> 01:52:04.449
danceratopz: It's been a very valuable call, for me, and hopefully for many other people here. Thanks a lot for coming.

732
01:52:06.230 --> 01:52:07.849
Marius Van Der Wijden (M): Thank you. Bye-bye.

733
01:52:08.140 --> 01:52:09.370
jochem-brouwer: Thanks, Dan. Thanks, all.

734
01:52:09.370 --> 01:52:09.810
jtraglia: Bye, everyone.

735
01:52:09.810 --> 01:52:10.450
jochem-brouwer: Right.

736
01:52:10.450 --> 01:52:11.800
Toni Wahrstaetter: Thanks, everyone.

737
01:52:12.520 --> 01:52:13.270
Ilias | Dedaub: I do.

738
01:52:14.310 --> 01:52:15.410
danceratopz: Goodbye, everyone.

