Status Core Developer, 20 Aug 2018

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Status Core Devs Meeting - August 20, 2018

Description

Agenda and meeting notes from the call can be found here: https://hackmd.io/VN4rIR3mTxiUfujlDRNYhA

A

Right cool welcome everyone to the second status, Cordova meeting, I, guess just briefly on the agenda: there's three pots, so one this is in terms of specific topics that people want to talk about, so we have first light clients and then some stuff around desktop and mobile code sharing and testing Hardware wallet, secure it security and first step towards the doubt, and then, after that we have stuff like in general, going through different sort of teams and so on and see if there are specific things, people want to talk about and just a hack MD link, if you wanna, add details to notes and so on.

A

Thank you, Andy for taking notes and just briefly lesson learned from last time that we want to try to focus discussion on stuff things that are future oriented and not to give updates about the past, but more things that require people to be here and and we'll also try to squeeze out some sort of decisions about things that we want to do as outcomes. And if anyone said something we can add it at the very end and so I guess do we have any any I mean any specific things.

A

People want to say before end or any questions. Anything.

A

Right cool so first one like clients, less and ulc, and so the context is that right now, there's a bit of a disarray with respect to these efforts and it's not clear what team is opening it. Then what deliverables are, and it's one of our it's the top procure that we have to have 10% of users using this, so we should be clear about so what we're doing here, though, for most of short and long terms, longer timescale, so I guess the some wanna start talking about it.

A

The Igor OR or Eugene Oh, Boris or Adam, maybe.

B

You see car is only you still unit, so we have dropped off. Intervention you'll see in application now agrees lead on this one. So we decided to start this one next week.

C

Yeah, we thank sure we're the coal today with Boris about lesson. You'll see what to do and with it and yeah I think, with this week, I'll wrap stuff up with mobile, like topics and they yeah.

C

We can start first three unless like making it sink, and so the plan is like first iteration to make less sink again and then enable it as an super work-in-progress option is the app, so people can switch to it if they want to and then test it thoroughly and figure out what exactly the problems are, both with performance and functionality and stuff like that, and then we'll go from there like it depends on which issues we'll find and how long it takes it's a little too early to say, sell a robot, but somewhere along the way.

C

We need to fix the most annoying issues with less than reach also make a you'll see an option in the UI, and then we need to polish this and then yeah, and then we might only be a work in progress. Yes, functions in service, so before actually I mean like until we reintroduce less and there's not that much to to talk about something specific, because there are other things that might be wrong, that we need to fix.

A

Okay, cool: it's a good idea. The focus is on first getting less interrupts and then not go from there. So not hot pile of drugs. Yeah.

C

The first, the first iteration, is getting less as an option. Inside the application, I mean.

D

C

Would take like I, don't know as they used to add ulc as an option too. Then it makes them threaded as well. It's not them I mean it's also depends on. Will you'll see go to like upstream or not like. If you hear that upstream like next week, then why not have two option.

A

What's and what's our plan for nodes like they were running in a cluster or we use them self external, and this is how does it interact with the discovery work that litigating this working on? My.

C

Thoughts again initial, they might change right now is to use the same cluster, and there is one performance related notes that why I want to do this, because, right now, we already have all these connections because of the whisper nodes. So if we use less on the same note as we use whisper, then it won't be like additional network connection and that might save battery and things like this, because if we go like later, we need to switch the proper discovery and yellow, but right now just to reduce the scope of light regression.

C

Telescope, like especially like performance issues, would probably should we use the same connections that we already have. I'm gonna do have power status cluster, so essentially the first step would just remove impure equation right. We'll still have a we'll still have our cluster, but we'll also need to run less notes on the cluster. On the same, like on the same addresses that we have for bit notes right now, so that that's my initial idea.

E

So so also Dimitri started working from looking in a bit node because yeah, the other alternative is that we will run Ali as servers on our site and that's I mean that's not a long-term solution. This is just we are not I mean we are not. We don't want to become like another source of lillius servers. Servers that it sounds are lies and so on. So we can run their budgets. That's a lot of maintaining work and that's we have just one DevOps guy. It is just gonna, add more name and yeah.

E

So I I don't know how looks how it looks like we're hiring. But if you want to put more pressure on on that team, so they have up steamed, then this integral right. Now, it's just like crazy, so another one. Another thing is that we will try to. We can try to integrate with with node and that's kind of being research for Dimitri, but that's still not really clear. Yet how we could do this.

C

Yeah but mind you was to have like our current quarter as a temporary solution, because it just makes it faster to reintroduce back long term shows I mean otherwise it's it's not what we want. We want decentralization and we've kind of centralized everything.

C

E

I'm just saying that it might be a little bit, it is just another thing that we would need to do on the back inside and I guess on the like cluster side, but I think it has not been started yet. So this work should be also started.

E

B

E

Thing is that we could potentially get Ali s servers from the 3m cluster I was looking into that, but the problem with that is is you need to wait like at least 10 minutes to find something, so it can find something but yeah it's like there's a lot of time to get even one working, alias server. So I guess that's just not gonna be an option.

F

E

F

There's an interesting option in mist right now, where it uses in Fuhrer when you start yet in order to get like balances and stuff like that, and in the background it starts looking for good nodes and in connecting to the whole they're in network through less or through whatever you happen, to choose like life, client or or anything else.

F

So that's one option forward, but another thing I wanted to mention is this fact that whatever we want to do with our cluster in Fuhrer is likely doing better right now in terms of using a centralized solutions to sort of redoing their effort and building up this huge monster on our side, seams seams riffing inefficient, like from from from from a who, does what well perspective yeah.

C

Like I was exactly saying that it's a temporary solution, so we don't need to be like good if it's an option, that's like with a lot of warnings and only for some people, then it's like we don't need to be as good as in fear right here.

C

While we tend to release early authority to like production for you like not doesn't like option for some. You know people, then we definitely need to have proper discovery like real solution. So with oh, it's all good ideas for production right. We need to start the bit simpler, yeah.

F

But but but it's not only about delivering like a cluster of nodes that can serve the stuff. It's it's about us. It's posing ourselves to a huge amount of risk, risk of being attacked, risk of being taken down risk of being. You know all these things, yeah.

E

Right but I think that this idea of the earliest service running in our class or just for testing purposes, never gonna be used by by releases right. Am I right or I.

C

Was thinking that we have this option any releases, but it's not on by default, so it's um somebody wants to test it and we'll give a warning to him. That's okay, if your enable is everything, oh great, so it's so while it's like a work in progress feature, we might use it but like to claim that it works. We with definite issues which the.

G

Impact associated like I, like with the additional risk of adding distance for the cluster we already have, is it? Is it that much more because, in terms of like a additional risk on our side for testing purposes, if you already have the cluster running, what what else are we exposing by adding this functionality.

F

So what I'm getting at is is that I don't mind us running a cluster, but we should not be running these ocean. It should.

G

Not be a that's a point of functionality for the application, it's more along the lines of a testing environment for us to make sure it works. Well,.

F

It doesn't even have to be that, like it can be a cluster that runs there in order to improve the etherium network as as a whole right, so we can be buried. There's this tiny little spoken in the in the whole system, because we have an interest in the network running well, it's.

G

More like a bootstrap to the network versus the network itself. Well,.

F

No, it's just like one more one, more partition right that that's the thing and we'd be an equal participant amongst many right, but if, if we start building too much functionality, which is status specific, such that only we can run the nodes and only we do run the nodes, it's there. Where we become that single point of failure and at that point.

F

If you're as not that bad.

A

Can we paralyze this work so because I've ever hunted a lot of the problems with Elias are not necessarily only what sort of the nodes but also times running it on a phone and and sort of when to sink and specific a per whatever is give me, API fix and so on. So maybe we can paralyze it.

A

So this swarm, that's being formed, can focus on that and then there's another piece of work with the center around discovery and that nodes and these country rings, maybe that make sense to be owned by ATP ATP a team. So these two things are not. They can be done in parallel right, yeah.

C

A

Disagree with that no I feel like that can be captured, but.

C

I think one one thing that we need to the second one. The discovery probably should start a bit later because it's still nice to know so what I like, if we want to avoid this in, like it anyway, but we start over optimizing control the optimization problems that we don't have like. Okay, we only seen we decided that out of a sudden, we only think on Wi-Fi one. We don't have problems if you don't like 3G and start developing something like this, that we don't really need to develop.

C

So unless this like premature optimizations are because there are a lot of cool ideas, how can we optimize that? But we don't really know what we need to talk to my son unless we start really working with it. So as long as this one is addressed and yeah, it's, it surely should be paralyzed, because it's like two different tracks, one is how to discover, and also how to be one among many and another one how to make the protocol itself work and work on on a mobile device and how to consume this, like whatever.

A

So in interest of making sure that we cover all the topics to someone sort of strongly disagree with this approach or think, there's something big that we're missing.

E

I just want to say that it's gonna be a bit more work. If you want to even have like a few Elias servers on our site, so yeah I think yeah I, just I guess we need them to at least for testing purposes, so we should start earlier. The case is that, even if you run them and just make them a part of a dream network, it's gonna be like all the slots gonna be taken by some like other peers, because it's just like that area network is just running to you.

E

Ids servers, so there were gonna, be need some mechanism that just make sure that there are slots for our for status, apps like basically to connect to because otherwise it just it's just not gonna change anything. So, there's little more things on this side. If you want to run those notes but I think it's it's okay, I mean in my opinion we should never go with that to release, but just to have something to start looking at well, yes, mobile and generally, like general, it has the whole solution. That's just that's needed pretty much. You.

C

Mean that to keep it a night list, but never keep them in their mid field. Yeah.

E

I would never put them into a list in the left like in there please race in the better.

C

I mean for the first few iterations maybe make sense, and we can always revisit that. What.

E

F

E

Have some mechanisms.

E

Yeah, it's being somehow Aled.

F

What do you guys think of this idea in general that we actually treat in Fuhrer as as one option amongst many and the application, just you know, connects to uses them interchangeably.

F

Whatever option it has a.

C

A

F

Yeah, like I, don't know it like it connects to in Fuhrer first and then it connects to the network and and and and technically it should be getting the same information from both and then it can make like a smart choice between jus, based on whatever right, based on in Fuhrer being up and working based on inferior feeling getting similar information to to what the rest is networks giving like the mrn.

F

So we trust in Fuhrer nodes, just as we would trust any other, less node right and then we can also play with things like when we're about to make transactions. We'd beef beef it up when we're not making transactions and just looking, we wait to tone it down. We can look at options like I, don't know if in Fuhrer goes down and then the app just keeps working because of the nature of how it switches between reasons yeah.

E

I like the idea, because, for example, with note integration, you need to pay for that, so you need to have first, some channel to pay for product service and there's a couple of options. I guess that the guys behind people are considering, but I will have I believe that in the future it's gonna be work. This way that you actually need to pay for the service like alias servers, because no it just it's currently visible that it's not possible to just have like a enough number of alias service running.

E

So even if you want some, you need to pay for that and I believe that that's gonna be an option that you use in flora and then you just switched earliest server that you bought so so.

A

Just in the interest of time, maybe we can just sum it up.

A

We can continue discussion afterwards, but just sum it up, there will be a swarm with origin and an Igor and possibly others, and with the initial focus of getting less interact, and then you will see, and so on, also explore the seamless integration of things and then it'll be a team wall sort of support and, and there will be continuous passion about exactly how the discovery process will work, whether it will be on by on in release or not, and also at what point, who we can sort of had this be graceful.

A

Is that roughly accurate or some some viewpoint? That's missing from that? You.

C

My fear sounds.

F

H

Oscar, can you just repeat who's gonna be in that swamp.

A

I think it's in there idea, but whatever it's an idea, they've got point being mostly that it's one that that swarm is responsible for it and that's most of what I want to get at. So it's not like three different teams, still kind of like looking at each other. That's this form! That's in graph right now to eight you know whatever it's called tweet nine I, don't remember.

B

A

To nitrate, yeah yeah, all right, cool, okay, next one, a desktop mobile mobile, app code sharing and automated testing and the context for this is that we recently started sharing code base between desktop and mobile and your some childhood illnesses. Issues relate to this just the goal. Just general discussion and outline next steps. I believe max was due to at least.

D

Yes sure so, basically, this movement- this might be like a beginning of movement to have one single platform. I mean have one single code base to support different platforms right now, it's mobile and desktop.

D

It looks like that for ecosystem, how how similar platforms are working for others. Software products desktop is pretty essential, also like mobile, and can do things and have such features. That can cannot be done on mobile and this, why pretty important of users and such more movement also can be like start steps to have a code base that once we will need to be able to port for others, platforms like embedded platforms or anything else which we will have in mind in future.

D

But at the moment, once we have started to share the same codebase desktop builds. They started pretty often to face mostly, we believe, because currently we don't have any kind of unit testing and any kind of end-to-end testing for functionality which we share. These mobile builds so you'll be great. If we can, mobile team can suggest based on their experience.

D

For example, will they be able to cover all.

D

Functionality or most of functionality, if for desktop, builds, we will implement and run on every github PR unit tests for closure script code, or maybe only a closure script code is not enough to cover all possible issues: runtime loader script issues and additionally, additionally to Z, we will need to have separate end to end tests for desktop also. So it's like the question for suggestion to mobile team.

B

So I can't answer for desktop and turn test. Automation, I already started research and for now for as the main goal just to replace our mobile driver, which called coach named IBM driver, V's destined driver, so we will be able to reuse the same logics as we using contest's for mobile in desktop and so far we tried three drivers of them, not acceptable solution and I.

B

Don't want to share why technical issues but I can, if somebody interest, so it was like I. Don't drive Urfa mark by Adam and securely securely, basically, is a magic ignition based tool and we are not going to keep a screenshot of each element. It's a really heavy and it's not possible this. Our actual test design I, was a racial automation framework. So all other projects seems to be dead, but with some contribution we can make them alive and so far we are going to try a couple of other drivers and I believe.

B

In the end of this week, we will have PLC for this optimization running on Marcus, 6 or the worst case scenario and on Linux. That's it from end to end test site regarding desktop automation,.

D

Sounds awesome if.

B

You have any ideas experienced, not only you. Anyone here with desktop optimization tools, bad experience, good experience, gesture and shot would be really useful. Thank you.

D

Now, thank you great then.

A

Hey drew, did you have, maybe it's not the same, but I remember you had some test of experience. Maybe does a specific platform.

B

Yet not sin in web related to platform is it was more dotnet and desktop? Okay was it window, so it was to dot, coordinate, I, believe Windows, not not yet dotnet core, no okay! So yes, but from my perspective, we really should take care about running unit tests. First of all, we have them. We have them and we can catch that errors on before uncanny intent test and we have that massless builds when the market is green and github, but actually they failed.

B

So before an intern testing, we really need to fix, builds and to run unit tests and I think that it will be more quicker than doing content test solution.

D

That sounds reasonable and when we are talking about unit tests it another question appears is that if we will run the same similar unit test for desktop position, we will mean that we will run the same set of test bus for mobile and desktop, and it seems that we need to move to some common set of unit tests for both platforms.

A

So some clothes should've here who was looking till maybe this is something I wanted to close tabs on desktop. You could look into because that seems like where they already know me. This.

A

I guess nobody's on the call.

D

We probably should prioritize this because it's pretty essential to work on.

A

This is something blocking the you know of the in terms of course, development in vessel team for this right now.

D

It's not working, but like every two days we catch something that breaks runtime for desktop bills. On my dream bills.

A

So what would happen if closure, desktops or pads unit tests that run Kubo, and then we can enable it as a blocker? So we don't get pull requests. We don't get a good pull or get smudged that block desktop unit test. Would that be a reasonable first step that the desktop team could could do yes,.

D

It's priority: I'd video, guys.

B

Sorry for interrupting, but right now, I can see that we are building closure and we are, we will run unit tests twice like would be great to make this simple, similar pipeline for both desktop and mobile, not just for separate pipelines.

A

A

Jacob I think there's some your audio sir. Would you mind adding these what you said as notes to that can be sure welder.

A

Do we have anything else with respect to desktop mobile stuff.

A

All right, cool next up, Hardware wallet and the light question context is we want to support hot wallet for calm and, if possible, this requires a lot of coordination across hardware development, security UX. It was not captured in current goals or priorities among teams and, additionally, was to have to sort of hardware products.

A

So the goal is is to discuss roughly what we need any unknowns and start to figure out the team as well point people for this may be a rough timeline and also understand maybe relationship between the two wallet products, I guess, yeah, takeaway and great.

I

A

This is a GUI.

I

Speaking last week and I'm going to product manage these hard words, so I'm, together with miquellee today. So what we have for today is a light version of this hardware that they blocked by me. Kelly. It's avocado applet, that by the way we are starting a security audit of the applet code by an external company.

I

It's starting right now, so this Java con is pretty simple. So I have a card with no MMI its contactless running on NFC, so we will focus the clients integration on first, maybe or so with the PC. If we source us beat who smart card reader and so the the applet is called complete, it's documented on github or the api's, and so what's at stake now is to start the integration in the client.

I

So there's the world who have a first version of the product distributed at DEFCON, so end of October with revegetate deadline.

I

What we started to do is to write some of the basic user stories we want to enable, and there are actually quite simple. There are three things: setup of the art were dead, signing transaction and lagina setup, meaning we want to enabled a user to create a new status account on his hard wallet or import an existing account on the other at signing transaction, meaning that to sign a transaction. The user we'd have to tap his card on the back of his Android phone and enter the pin code.

I

The card instead of entering the the current password and something is enabled login into statute account with the hard wallet. So instead of selecting an account and input, the password, the user on the login screen would just have to tap his card and enter the pin code. So these are basically the three user facing scenarios attacked when able and integrating the client there's. Another thing at stake on which we have to decide quite quick is how the applet is noted.

I

On the have occurred, there are two options: a first option where the manufacturer, when we solve these records with load applets directly on the card and the second option where the clients would take care of this at first setup. In this case, the client would upload through NFC applet on the Java code.

I

So there are different user interface security and feature consideration regarding this choice of how the applet is. Did we win discussion that with Micheli, and we will share by tomorrow a document with some of the conservation to take her into account to share with you so that we decide on which path will go since impact, of course, on the train situation?

I

E

That's basically.

I

What we're trying to do there are some application on the client side for me to discuss. I, don't know. If that's the moment right now, like one I have in mind, is the fact that the user would not have any password to mean to define and remember any password how we will manage that the client I imagine the password is used to store some secret that has on the client and does that need to remain at once.

I

The private, key and whisper key on the avocado, and maybe the other thing to note right now is also the whisper key.

I

The java code can derive the whisper key directly under java code, so that would be a way to not have the the whisper key anymore in flash. The client and wanna keep it in the ram of the client by getting it from the java car. So and that's all rotoscope that's been discussed now so.

I

There's a document with the user user scenario: that's been shared, then that's open for discussion and when I guess we we need to set up a team working on that with some contributors from the client side and from the user user experience team and with the McCleary myself at least so, let's open for discussion, and so this is the other one like. Maybe we can discuss that before before we briefed about the hardware?

I

What else Pro, which.

A

Again, yeah we can skip the pro thing for now: I guess yeah. Thank you.

J

A

Yes, Andrea: did you start looking into this or I? Wasn't yeah yeah.

J

So I I started working on this couple weeks ago, just some hours and then I stopped, because we had some security bug to fix or follow. But the first thing I've done was checking a PR that is already opening the go, etherium repository and changing some parts, especially in the Association and the pairing code, and then after these I would like to understand.

J

What will we want to do like having a fork of death or like contributing on on that PR? But I don't know if they are still working on this and, of course it depends on our like timing and everything so I think these days, I can maybe focus on this.

A

Gorem, what are your thoughts drums of closure, client interview, yeah.

K

Well, first of all, we need to reduce the scenarios to the bare minimum, that's that that applies to both the closure and the overall.

K

You know look at the thing so when we said set up, signed, transaction and login like which ones of those are actually the bare minimum separate.

K

If you had, if you had to remove one from those three look, finally getting more, if.

I

We were to remove one I think you should keep signing up the transaction log in, but then there's the question of the password excellent.

K

So we are back to the password, so you mentioned that instead of password in might use pins.

I

Any comments, and yes, okay,.

K

So basically, the user would have to use bin instead of a password. It's.

G

A combination of both the hardware and the pin- yes.

I

K

Yeah, okay, so, basically, if say, your car gets stolen. Without your phone, the user or the attacker wouldn't the thief. We need to figure out the pin still right, yes, which I assume would be a bit easier. It's it's a numeric pin like for for bank cards right so.

I

K

Okay yeah, but my other concern is that we would still need to support the password login and the password sign transaction and all the general stuff, because not everyone is going to have these cards well.

I

The user chooses if he wants to set up an account with a the regular way with the best or if he has a card, then he can set up his account on the card. But in this case it's would be much better not to have a password anymore. Otherwise it's going to make too many things for him to remember. I forget.

G

Who made it, but someone made really good flow diagrams of the user experience for both options depending on and when you'd make those options and what what part goes away, based on which part of the step you're in I think that'd be they'd, be useful to put in the notes. So people can look over that. Yes,.

I

G

Think it would be important if you get rid of the password to have to push forward on the the biometric data for unlocking yeah the terms of a usability standpoint having the card out every time you want to use, the application may be a bit annoying. So if we have, if we get rid of the password having something like facial, recognition or or fingerprinting, could unlock these types of things, while using the pin plus a hardware wallet to do any type of value exchange.

J

G

True, but depending on what we're doing.

E

G

But, depending on what you're doing, if I guess it's it, what are we separating valuable right? Is unlocking the app value in itself or sending a transaction value so we're at what point do we have to enforce hard security versus just getting into the app? Because you know you're either biometrics can open up your phone. Your biometrics can do a lot of things already.

G

If it is, is sending chat messages, the the buck stops there in terms of security, or is that it's a fingerprint enough? If you have all three that should be enough to send value or do anything with the wallet? No access in the wallet should have access. You should need all three, or at least the pin, plus Hardware wallet, but maybe sending a public chat may not need that much. That's open for discussion on what we, because most important.

K

Yeah regarding the attacker being able to coerce the user into you know using the face or a finger or something like that. If the attacker is in a position to do that, you can, you know just use an iron rod and convince the user to give up the password.

E

K

Depends on how skillful the authorities are in.

A

The interest of time do we do we want to decide on roughly how we want to proceed, because it's all security considerations and things we come into detail, but just I guess from Arabs or people perspective like what do we? What do we need to move this forward if andreas looking into it from the go side and agora? How does it look from the closer side and also chad from the you excite.

K

Yeah all right, basically, what we need to is a timeline. The timeline with you know realistic expectations with realistic resource plan, and you know so. The only way we can find this is with a timeline timeline. For example, when do we get the cards? I know that the prototype cards have been already mailed, so I expected they would be around. For you know, in a couple of days you tried you had you had some question about the delivery of the of the promotional cards.

K

We mentioned something that we are going to hand out like a couple of thousand cards. Do you have any update on that yeah.

L

Akeelah, we can probably talk better to that all.

K

The way, yes, all right.

I

We need to define how we load the applet on the cart, because this has a direct impact on how we source them, to whom, so that's the first I think that's one of the first team discussion we need to have is decide how we do the rapids, because if we decide who have the manufacturer load them, then it's almost sure we won't get products for the DEF CON.

I

The structural decision to be able to source code I think that's one of the first step and then regarding the the timeline.

I

So there's all the sourcing of hardware parts which I would manage and let's try to define a list of features and discuss with you the implement end time implementation timelines. You see to put up together this over project timeline.

K

Okay, so the very sound to me is that directly on the critical path is actually the decision how to load? Yes like okay, so if the manufacturer loads, we are not going to make it. Basically, that's that's your yes.

I

I think we are not going to be able to make it by your manufacturer. We could do a hundred cards by ourselves program them ourselves, but more of a workshop exercise than the manufacturing of final product.

I

K

So basically, the only way you can see this to be done in a timely manner on a thousand scale is with a bootstrapped loading. Yes, okay, so we have okay, the first decision on the on the UI side on the mobile their side, the majority of the the major dependency that defines the timelines is actually the selection of the use cases. So I, guess that that you know kind of depends on this decision as well.

I

K

And there is there anything that affects the the go side. Yeah I was.

J

Also thinking in general that we were, we started discussing about the coupling the the keys, the wallet key and the whisper key and I think we. We need it for this because we're talking about having, like the the whisper key, only memory on the client, but for now, since the the key is the same, it would be basically the main key, which is also the wallet key in memory, because you need to sign the messages when you chat so I. Think one Dependencies finish in the starting and finishing the decoupling.

J

The the keys and I know that I think there's a lot of work, also in the protocol and in the chart, because we need to like add some features like asking the like. The wallet address request, because it's not derived from the whisper key anymore and and then also compatibility between different versions. So I think also. This part will be long and I. Think it's me that if I am just directly like the way way, yeah.

A

Doesn't want to I'm sorry good yeah.

K

Sorry, just one extra question for in-depth into I'm, not actually I'm, not too familiar with the direct consequences of the decision over the loading decision, but intuitively it seems to me that if we, if we do the bootstrapping option, we are also, among other things, going to have more custom code on the client side, so that can also affect the timeline.

K

You know, with the important difference that it's being done with the resources on our side, so we can actually, you know, help give a choice so and yeah I'm not sure how familiar you are at the moment, but is. Is this something which we should be concerned about.

J

Yeah also this also these parts will be like a lot of I mean more work. I, don't know exactly now like how long it could be, but for sure I mean it's.

A

Maybe we can folk tale conversation about this. Absolutely call to someone want to own writing up a swan proposal for this, so maybe we can top Baltimore I. Also think that we probably need a another. You, like a closer developer, probably gone on the team.

A

A

Essentially, we could talk about it outside I. Think of Cora and Andrea can also help you. Essentially, this is a proposal with the goals and and the iterations and what roles are needed. It's in github status, I am / ideas, corn and ready when I help help up yeah.

K

Sure sure yeah, basically its scope definition, is something we need to do anyway. So yeah, nothing, weird, yes and.

A

We also need we need close developers at correctors, yeah.

K

A

So if yeah I guess yeah we can we can talk about that outside cool.

A

A

I see it just gets missing me: you I'm trying to promote them to be a panelists but I'm failing Johnny. Do you wanna help, thought and double clicking here, nothing's working.

A

There's some wonder how you promote people.

A

In also fill in the rumen.

H

It might be because there's 25 participants on the Skoll, which I think is probably the limits.

A

D

A

But but they are in the coal, there is not. The panelists should.

H

Have a sittings right.

D

Well, I can jump off and join youtube.

A

Just all right, next, one security.

A

Yes context: yeah, we were yeah, we yeah, I guess, do you want to take away Corey yeah.

G

So if I haven't talked to you already I'm Corey I'm, the new security guy, which means I'll, probably be bugging everyone about what you're doing and why you're doing it. I think I've been trying to do a lot of like polling around the company to see what people are working on, what they, how they've used security in general, because I was brought on with the assumption that there isn't a lot of security practices in place amongst the company.

G

I I'd say that there's no like formal protocols and checklists and methodology for doing traditional, quote unquote security, but the backdrop of security amongst the company is really high. It's like I'm really happy that I didn't walk into a bunch of landmines and I think that everyone has the basis of what security is. They just don't have the confidence to say they have it.

G

I'm gonna be working on trying to put forth a lot of stant like checklists that go in place of what you're currently doing so you can feel more confident about taking off all the boxes that respect security and whatever aspect. That is because it's a pretty broad concept and I hope that y'all can help me do that in a lot of ways, because I can't do it myself and that's going to be. Basically, if you have any concerns comments, questions send them to me or if something happens online you can always send it.

G

An email to security add status that I am for any type of like quick response. I'd say uh see: I wanted to bring up the concept. This is kind of a I say a mind frame when thinking about security, that I don't think people have ever been exposed to as the packs, the pyramid of pain and that's when looking at potential threats with an accompanying.

G

This is built for I, guess: traditions like looking at traditional, centralized infrastructure and looking you're looking for things that to see if people are attacking you or not, but it could be easy for me to easier to share my screen or I. Put the link in the in the agenda notes to slack. If you want to bring it up yourselves, what do you think is easier if.

A

When I show your screen, I think that that would Greek. Okay, let me see here make sure I have also flooded. Recording okay,.

G

So the pyramid of pain made sure my screen real, quick.

G

G

Is that, yes, you buffoon.

D

G

Great Pyramid of pain was brought up from our security research or a long time ago. I think our on 2013 or so, and the concept is relatively simple. If you look at the graphic here, you have the types of things that you can. You can see indicators upon in your network or or your infrastructure that show that someone's someone's doing something malicious and at the very bottom. You have something you have things that are everywhere yet easy to change and at the top you have TPP's.

G

It stands for tactics, techniques and procedures, and, as you move up this, this pyramid, you increase the difficulty to change that type of indicator, which means that you make it more and more and more difficult for. If you can find these indicators and and remove them from your network, you make it more difficult for the attacker to change what he does so, for instance, hash values say: you'd, find a moshus file in your in your network and you didn't make let everyone know that hash file exists across the network and you get it out.

G

It's very easy for the attacker to change something about that file and then change the hash files, so you're not really tripping them up in any way shape or form.

G

So you can find these things easy, but the attacker just changes it and keeps going about whatever he's doing and then, as you move up to this, this pyramid, you continue along that that same type of thing, so at the very top, which is where you want to be finding things, is your you're identifying the tactics, techniques and procedures and attacker uses to attack whatever you're doing? And you know, if you ever to mitigate these types of things, then you're able to you're.

G

You force the attacker to change his behavior, which is the most difficult thing to change, and what I the reason I bring. This up is a mindset of if you're, building, something or making a change. You want to look at the types of vulnerabilities that are introduced or are already there and think about what what gets left over what you can look for to find in a Vegas misuse of whatever you're doing and I try to be. As general in term here, because that, although this is a developer call, this applies to UX.

G

This applies to all kinds of things, because it's a decentralized project and I just wanted to kind of bring up this idea of, like as you do things, I want to keep I want you to have some type of back-end mind set of what does this?

G

Do it from a security perspective and how can I find those types of artifacts and where, on this pyramid, would they would they lie not necessarily specifically in terms of domain names, IP addresses, etc, like, although that may may be appropriate for what you're doing more along the lines of if I can detect these things? How difficult is it for the attacker to change that?

G

Does it lie on the bottom in terms of hash values, or is it up to the very top in terms of behavior, and we wanted to try and always do things slash, be able to see things and change things that are at the very top here? Because that's what keeps people from if you change an attackers behavior more often than not they're just going to leave you alone, because they're lazy and they don't want to change their tooling I'll. Stop there and kind of ask question or answer questions or get comments.

G

Is this a concept that is useful to you? Do.

A

You have some examples of how something like this might apply to things we are doing or planning on doing it that well.

G

That's actually the next. The next point that I wanted to bring up. This is a traditional way of looking at things, and there isn't a lot of stuff out there that points towards how this changes for the type of development that we do and I want to I'm writing about this I'm trying to put it on a framework that makes sense for people, so that I have hard concrete examples of how to apply this same type of concept to decentralize.

G

That building, because it's very different and the attack service is very different and it's a it's. It's kind of a new thing. There isn't a lot of stuff out there that I found at least that that explains this type of stuff, so I think it'd be worth while to even brainstorm on the types of things that we do, that I want I want to rebuild the pyramid for what we do in general.

G

If that makes sense, this is something I've been talking with the other security folks in the community to try and to try and help do so as I as I build this out. I hope with. If any. If you have any ideas or ways of doing that, to send them my way to help me kind of build that out more robustly, I I want to push forth this idea and in our community.

G

You so I guess not I'll uh I'll develop that more concretely and ensure it next time I've been riding on a pretty thoroughly.

A

But I guess also: this would be an opportunity if anyone has any questions around social security, and these things did so big opportunity for as well and I. Think we have this security scam with. Yes are good.

H

Yeah, forgive me my new penis on this, but didn't I ego start a security champs procedure and what are the overlaps between what was happening there and what Corey is doing now? I.

G

Trying I'm taking over a lot of that security champion stuff, so that Igor doesn't have to wear that hat, as least as heavily as he used to I'm, still trying to figure out a way to build that in so that it's not in an incredible amount of time for the participants of security champions. I want to build. I would like that to be something that we do and I'm continuing, building the framework of how what that looks like they went through a UH a threat, modeling exercise with incorporating the ENS name.

G

They actually came up with some good results and I was really happy with it, and I want to make sure that we can get do things like that and I'll be building on the framework they don't. That's a good part of what I think I'll be doing, is building framework or tool sets that allow everyone to make better decisions about what they're doing and that's gonna be done through the initiative that you are started with security champions.

G

If that makes sense, I just don't want it to take up all of your time, because I know all of you wear different hats. But I don't want this to be a heavy hat that you have to wear granted, but at the same time, security has to be important. An important aspect of what you do.

M

I have a question: yeah cool, so um at the top of the pyramid, was the tactics, techniques and procedures I mean? Are you talking about they're, basically like the digital fingerprint of one unique hacker, or is it more like schools of hacking? Oh yeah,.

G

That's a great that's a great question so, like fates, think about so right below the pyramid. I'll share my screen again so.

G

Right right below the TTP's, you see tools right, and so this is what, if you were to find the fingerprint of a various tool and able to stop that the attack. The underlying attack could possibly still be happening in your in your network, but you've only mitigated a specific tool that does that attack.

G

So what the attacker would have to do to change if he's using that tool, has just changed the tool set, that the toolbox to do the same type of attack a better way of defeating this type of thing is to find the underlying attack and mitigate that and the example that most people use here for doing that is say: identifying people doing a pass, the hash attack or someone takes like hashed credentials and you and reuses those hashed credentials to get authenticated on the server and, if you're able to identify that, then you mitigate all of the possible tool sets that are capable of doing that, as opposed to just one, and so you you've, you've mitigated that entire attack service, as opposed to just one tool tool that can do that attack and that's a better way of protecting your network.

G

If that makes sense, yeah.

M

So when you, when you say, tactics, techniques and procedures, it sounds like you're really saying that it's like the underlying vulnerability and if you can fix.

G

That's a good music, that's the I think a better way to put it it's and it's it's more along the lines of a behavior for an attacker which corresponds to an underlying attacks like some people are really good at X type of attack and use a various amount of tools to do that type of attack, and so that's a behavior slash underlying vulnerability in the inner system. Gotcha.

M

F

G

I, like that, I'm.

H

Using my hands.

D

G

No one can see it, so that's really good cool. So another thing that I wanted to bring up I've had a conversation with a company called frets tack which looks at cloud cloud infrastructure monitoring, because we don't have a large security team looking at the infrastructure that we currently have, because we do rely on some type of centralization at the moment and always will have some form of infrastructure.

G

I want to make sure that I at least understand what's happening on an infrastructure from a from a threat perspective like who's trying to get at the network who's moving around what users are on, etc and I've set up a call with a company called threat stack that will give us a demo as to what they're capable of and why we may want to do it I'd be curious to see I'll just share my screen again to show you their website I'd be curious on who would like to join that call, and then we can have a discussion on whether or not we think it's a reasonable thing for us to do, and if we need it, I.

B

Can't I'd like to drink that.

E

As well, I've had it set up once, but I was too busy. So.

B

G

Is that can I can I can.

B

G

B

G

Okay, yeah I think you definitely are, should be on the call, because you're, the infrastructure, guy and I want and you'd be you'd have more of an opinion as to whether or not this is useful useful for us and whether or not we can just roll our own, because there's certain things that I can do that do some type of this. But it's it takes me away from it takes it takes my time away from other things.

G

I should be doing if that makes sense, like I'd, say the entire, offering that this company has is equivalent to about a full time employee, with infrastructure, as we currently have and I want to make I want to see if that's reasonable, for something we need, because over time time goes on, that reliance on infrastructure should go down based on the principles we have as a company and what we're trying to do as an application. That's.

B

A good point just.

A

Sake is also interesting and I I might be into this one interesting. Instead,.

G

Okay, let's get a list of people and I'll include you in a and the emails and we'll set up a time for that, because he's basically just waiting for me to respond to him with who I want to join and I've.

G

Given him a rough outline of what our infrastructure is, so that they can create a demo tailored towards what we would expect from their company on what we have, and so this is, the compass is the see you can look at a kind of a an idea of what an attack is, and this is like an AWS attack. Someone gained something gaining access to credentials across eight of Atos infrastructure.

G

And so on and so forth and then you know, but you don't.

A

G

Lot of this stuff that may not be it may not work well for what we do. We don't do traditional app development or at least at production, and things like that, but we'd also have credentials that we care about and I would encourage you to go to the website and then look at the video that shows you this stuff and then for the demo we'll discuss whether or not it's appropriate for us.

A

Awesome cool: is there anything else like questions having another security before we move on.

D

About parameter pane, interesting if to replace a attacker, for example, developer of a service. Other part of pyramid will still look the same, for example, because developers select tools, select some procedures to implement the rest of system part. So from that point of view, developers should aware of attackers can behave, and this will enhance selection of those procedures and overall security of the system.

D

G

That's a that's a really! That's a really good point: it's a nice! Something I want to reiterate: I'm, not the domain expert of what you guys do and so I need your help, because you understand how things work better than I do for whatever specialty you have, which means that you understand the types of tools someone can use or the types of vulnerabilities that exists with what you specialize in, which helps me understand.

G

The overall like risk risk service of the application like I can't do everything I need your help and your specialties to unidentified these types of things with me, so that we can help. We can all mitigate him if that I think that's. What you were saying is that right.

D

A

G

Move on, oh one, more thing, I think the last thing I have for now. There was a like a social engineering attack that was attempted on with with Kim, as she found an email on her spam folder.

G

That, basically, was someone pretending to be Jarrod asking for her phone number so that she can do a task for her really quickly her personal phone number, and what this was an attempt was getting access to her phone number, so they could do social engineering to then transfer over her SIM card, potentially to then gain access to whatever she has SMS two-factor authentication with or whatever her her phone number had provided in terms of security.

G

She didn't, of course, follow up on this and she reported it to us through the slack, but I want you to be aware of these types of things, and if you see them, please send them to me so that I can investigate further.

G

If you have any questions on what that, what that attack is or like the idea around it you're more than welcome to ask it's just I'm, going to be sitting like a PSA around of like a write-up of what this is, what it potentially can do and how to mitigate I got a what to do about it in case you accidentally fall for something like this or what it exposed you to.

G

So you can be aware of it, so it's actually happening this is that hurt Kenneth status, that I am and there's somehow and other people are getting into that at least the protocol in which we create emails for the company, which is basically first name at status. That I am and then broad, spanning out these types of things just see if they can get access to some of our underlying credentials through social engineering attacks.

G

So be weary of those things you might find them in your spam folder. You might not I think we should also look to put in places like certain amount of procedures or or general guidelines of things we will ask for and the official channels in which we asked for them like for what like, for instance, Jared, will never ask you for your personal cell phone number through your email.

G

If that happens report it it's not something we should ever do, and so I don't know if there's basic guidelines or rules on these types of official channels for certain things, if there's not I want to make them so that everyone at least knows where to go to find out. If something like this happens, and it's more questionable than this.

A

A

All right, mmm moving on um so first step towards the Dow um I guess: there's some work around there, belting it up and there's also one okay are around liquid touching and then there's also been some conversations and experiments around get coin, and so on, I'm, not quite sure exactly what we want to get out of you this, but maybe stop it spreading some context. I think and you might have some and then a grape is a grain.

M

It is yeah yeah.

A

Yeah, so do you want to take it away what you want yeah.

M

Yeah so initial investigation into the voting that, basically it's not so much about features or doing anything new, it's it's really about trying to discover like what is the reason that people would feel compelled to vote. So that's what's currently happening now. There's a survey, that's been created by Patrick from user experience, design and that's gonna, be going around the community. I've got a meeting with hutch tomorrow to send it up, asked to you to basically try to figure out. You know why would someone pay gas fees in order to vote?

M

My suspicion is probably around the allocation of funds, but one about the date that, through a three-hour survey, yeah and then I was investigating how to create kind of unchained consequences that the idea being that certain having an untrained vote is very expensive and if it is only a signal to the to the research foundation, it's a very expensive way to signal whether there's other channels such as reddit or, in fact, the the feature nomination with the feature voting kind of app with inside of status. Looking at it as well yeah, not not a written.

J

A

I'm curious a bit about if we're all looking into the delegative democracy and so on, I know, Ricardo was doing some work, it's único! That's that's something we're looking into or was there something on that, for example, to deal with sort of voter turnout in these types of things. These.

M

Conversations are happening. What I'm, what I'm, looking at I think what I'm trying to avoid is building. You know adding on complexity onto the initial voting float before understanding. You know what is the things that's that people will vote on, but yeah I do believe. Ricardo's investigated, there's a state's discussed and there is a understanding. Work is gonna start happening on the Dow when Jared gets back from from holiday.

A

So we don't want to put it forward right now with respect to liquid pledging, and these other things, but I said wait. Is that an idea.

M

Yeah that's my conclusion. Just if we Const I mean first things first to try finds you know what is the reason that people will will vote or walk what topics I. The reason for that is low. Voter participation is I mean not just amongst the status voting that it's low amongst pretty much every single carbon voting or on chain voting. That.

A

Okay, Andy something's around a gate coin.

H

There's no particular effort that I want to give just right now, there's some other people that I need to speak to before we speak publicly and just on the voting stuff. You know I tend to agree with Graham I. Think that he's approaching this in quite a sensible way. It's it's it's a really difficult topic. A lot of people have taken it on and I think that he's caught inside.

H

That's if you are gonna require people to go through the pain of voting on chain and sure there are ux things that we can do to make that easier and more smooth. But it's always gonna be a little bit of like added extra effort for the user. Then you really need to have unchained consequences and deciding what those untrain consequences are is quite difficult to do.

H

Graham I think that ask is righted, but probably the best kind of an Train consequences are likely to be found in some of the democracy work that Ricardo's doing and perhaps liquid pledging you know so there's some kind of financial results on chained, even if the actual results of the votes is not necessarily reflected in the decision that gets made on chain. If you see the the distinction that I'm drawing there, the get coin update Oscar will come next week or two weeks from now or other I.

G

For those interests that I interviewed get coin last yesterday for one of my podcasts and I'll be releasing this week, if you want to hear more about how they work and what their kind of underlying principles are, you.

L

Yes, I could add something that falls into this bucket wishes. You know right now we're deploying contracts to test net and main that, and you know, we're taking the first logical progression in decentralized governance with that which is, is it moving to a multi sig wall?

L

That's what at right now, we believe is still a secure way to do things and sort of there's that sort of the state state of where things are today and then we're talking about the Dow and things like liquid pledging and and there's this big gap between that vision and where we are today right and like what are the actual features or the progression.

L

That's going to fill that gap, and so I guess one one question I've been asking is like: if you could build one feature in 60 days or less, that moves a multi-sig wallet towards being a Dow or a Dow component. What would it be and there's very little feedback on that right because it turns out very few people if have ever even used the multi-sig wallet or use one on a regular basis, so something we started doing?

L

Is we have like a just a working group of people that use use the multi-sig to administer some of these contracts, we're deploying on test net and just seeing like?

L

What's the process and what's the pain points in trying to bring people together to actually administer these contracts and I think you know there's little by little some learnings and we're sirness we're seeing pain points like the coordination of has to be done, offline or off chain just to get a vote or you know just get things done with a multi-state wallet, so I started a channel called multistage evolution if you're interested in participating in these test net, you know boats and using a multi, sig wallet or just the discussions.

L

Come join us there because there's really, you know very little feedback, because it seems like there's just not enough people who even use the multi-sig wallet and if that's the case, how can we really expect people to you know, do more sophisticated forms of governance?

L

L

So that's that's a.

A

Great point: yeah.

F

They're about coordination, there's just I- remember this group that just published an article where they're using whisper for coordinating transactions. Basically it might be an idea they get up if it's of any interest.

L

I'm completely open we're right now. What we're using is basically we're using the gnosis multi-sig wallet. We haven't even used the the more read the safe that they released. So there's definitely a lot of rooms for progression from there.

H

Yeah I mean I think that yet six point is a good one, then, like first of all, the bloom guys very good at what they do, and you know it's. It's interesting that they're using whiskey and the way that it was originally intended, which is to gossip information about various different aspects of your identity around their provider network, so that they can sort of do kyc and only have to submit one transaction on to the chain to confirm that you are who you're claiming to be and barry.

H

I think that's I mean to potentially answer your question, which is a really good one about like what is the one thing that you would builds in the next six months to try and move from a multi-sig while it's to a full-on dow, it would certainly be some kind of whisper notification system that would improve the coordination that we're.

H

You know the coordination costs and problems that we're really running into with the multi-sig wallets that we're running for ye NS like if I could get a little notification in my status, app to say: hey Barry wants you to sign this transaction.

H

That would be really beneficial.

A

In my toes respected a multiset like it might be training thing, so maybe we can host some super basic tutorial session or whatever people can join just in terms of actually using it. I know if that would help, maybe explaining sort of the connection with governors and so on in terms of getting more people to play on with this thing, in terms of you are like about about what the conditional bit of time I agree with what Dan said.

A

I also think that getting some kind of integrating sort of ideas in swarms and having some kind proposal based around that, even if it's like so, if you remember dad's presentation before, would like request for information requests of procurement, and he said two things I think having like a very basic kind of thing. There will be very useful. It's something that we we barely started doing back when there was like an open bounty team, but we didn't really do it seriously. So I'm revisiting roughly those ideas, but doing it right could be interesting as well.

A

And it would also play the liquid clenching kind of things. You could imagine that I don't know, maybe they're hollow wallet or whatever right, like that. It's um there's some bondage, because we want to do this thing and then it becomes reprioritize because there are funds associated to it example.

A

A

Does anyone else have anything on this specific topic.

M

I'm not sure if this is the same topic, but the signing of the principles and I mean is there gonna be any or development required on that.

M

Similar topic, in the sense that it is part of a Down, the sense of green to your Constitution.

A

I think it's not not determined right now, I'm very you updated their version of how to sign it and believe you.

L

The there's a there's another version that we can use, but it basically just signs it using but basically signed it using only web three as an option, but I think both both versions will, you know, will create a valid signature.

A

Yeah, yeah and guess in terms of moving forward without them is specific to sign or not. Since I we can wait. Contingent comes back, but at least that seems to be the decision right now well becomes revisit this and started earlier. That seems like it's ending, oh.

L

Maybe something sorry Oscar to come back in I got it something that actually I did just see. I. Think like was the discuss. There was like an update in discuss and I. Think I didn't see it initially, but I, just like just soared, is Jared posted a kind of reply and I something he mentioned, that I guess I didn't think about was I, guess it eventually there will be an actual smart contract that validates the signature and that someone signed right so right now we're all just submitting it into slack.

L

You know, but at some point in the future, I guess there'll be some kinda, be like an actual smart contract that that validates and stores the signature on chain.

I

A

The main purpose for the graphics, intro kind of thing that was more about making sure that there was something that we individually like was sort of agreeing to and at that I was at the bar in terms of reaching consensus, as opposed to just his finger.

A

A few set of people pushed out, and it was like a marketing, so-called marketing message that was double sort of the main thing and but but in terms of actual consequences, there are no specific consequences associated because it's not part of a greater design right now, but we can still reuse a signature. So we can. We can redo it as well, easily I think there's.

G

Also, I guess the aspect that the address you sign with isn't like what what ties that to us as an individual, we don't I, don't have a specific tie that address that I use other than it's one of the I. Just like might address. That's in my slack yeah.

A

Exactly, and that might be something where we want to have it- be more privacy, preserving where you could have like a persona that you interact with it's, not necessarily your human, normal name or whatever, and that's especially with Adele in some ways. What's.

G

The point I mean it's either, like I attest to these principles and I'm, showing that I'm doing that and will.

A

Want one point would would be that if you to be sued an ominous with respect to inject weeded out, so you wouldn't be able to get any funds unless you do sort of things. That's one one idea, whether that's the right path or not. This remains to be seen, but that's an example of how that will be able to go.

A

A

Okay, I guess time: is there something that people? It is something that people feel like it missed, especially from sort of the part to which is more sort of general through specific teams as on? Is it something that someone thinks is they want to talk about? They'll, bring up or get some kind of decision or discussion about.

F

I'll just briefly mentioned that the next is there a hard Fork is coming up for those that have missed it. I, don't think that should be anybody on this call, but in case anybody's good, so there's a couple of VIPs going in there. There's an aetherium death call meeting soon check those out because they won't be perfect. This.

A

All right cool, so I, guess with that. You finish up this meeting, recording on YouTube hasn't been like here, I'm, not sure, and also, if you want to add things to notes thanks too much time before taking ups, and thank you Johnny for recording. um That's it for this time, I'll put up a github issue for next one in two weeks as well.

A

Thank you, Ron, bye,.

B

Terry, oh thank you see.

D

You guys thank you, bye, hey.

B

Guys, good right, Oh.

B