 Alright so it looks like we're recording just a reminder of the anti-trust policy notice all meetings are attended by people who are potentially industry competitors and you know make sure that you're not talking about anything that you don't want to be talking about that might violate any anti-trust policies so with that we will head to the agenda so this is our agenda for today we'll be talking about just reminding about the events obviously the people who are on this call know that the TSC has changed to a new zoom link I also put that in the TSC chat I will do a quarterly update talk about the crypto live proposal and then talk about the testnet I have some discussion about the testnet so with that Dan you want to kick us off yeah yeah though I think you've already pretty much kicked us so just adding to the usual reminders as we're going this morning that everybody is welcome in this meeting and the rest of our technical working group meetings and everybody's welcome to contribute code to any of our projects and if you're not certain how to behave in any of these forums we do have a code of conduct that you can find from the hyperledger.org webpage but in short just be respectful of everybody around you in addition to the topics that are listed on this slide we had planned on having the update from the composer team and I know from their last updates that they're going through some changes but if anybody has a connection to somebody who's actively involved there it would be good for them to check in with us and let us know what's going on there and I think probably our our largest topic for today then is going to be the update from the training and education group so I will hand off to her. Can you unshare first? I'm still getting that I cannot share while the participant is sharing. Hey, Rye, I don't know if you can hand the ball over. Oh, I'm working on it. I'm not exactly sure you got to promote to presenter. Rye, if you just unshare the agenda slide then you should be good to go. I mean I unshared. Okay, reboot. Does everyone see it? Yes. Okay, this is the update as you know of the training and education working group first regarding the working group health. We had changed our meeting time from 10 o'clock PST to 8 o'clock PST due to requests from Chinese participants from China and then for whatever reason they did not attend the meetings and we had several people who could not attend at 8 a.m. and everyone understood that this was on a trial basis. Now that the U.S. is going to be going through a time change everyone agreed to go ahead and move it back to the 10 o'clock so that some of our original core attendees would be able to participate again. If we receive further requests from participants in China then we might consider having a second once a month type of an update so that we can be inclusive. One of the key things that we've talked about in the last couple of times that I've updated you is that the name of our group is causing us some issue. I think that what happens as I've mentioned before is that participants are in the edX course and at the end of the edX course they're invited to join one of the working groups. They're a little nervous perhaps about the names of some of the working groups and then there is training and education which feels like the next step and so I would say about two-thirds of the people that join our group are coming to get training and education rather than to contribute to training and education and we kind of go around and around about that. They're looking for a continuation of the edX course as much as they're looking to participate in a working group and so we've talked about names and I think Tracy came up with the best name which is the Learning Materials Development Working Group and we would like to put this out to your group to have some discussion on again trying to find the right set of words that differentiate that we are creating and developing materials rather than this is a place to receive training. Any comments? Yeah this is Mark from the Performance Scale Working Group. We have a similar thing where a lot of people coming in just want us to solve their performance problems. You know and so I think some of it gets back to even just the website. The website still doesn't really do much for working working groups that says you know hey this is what this working group does come join. You sort of have to hunt through and find a wiki and then go read what's there. I mean we list all the projects on the front page but we don't list the working groups so you're not alone. Let me put it that way. Anyone else? I think we've gotten too little exponential back off problem there. Silas, carry on. Silas might have carried off. Was it about why? Yeah maybe I just suggest we can make the name a little shorter like the Learning Development Working Group. How about that? Okay. This is Dan. I'm all for a name change and I understand the frustration that you've expressed earlier too there. So maybe if people want to give some other ideas out on the mail list you might generate some other concepts that start off with some active term so that it's clear that it's a contribution kind of thing and not a consuming kind of thing. Okay. I thought Tracy's suggestion was good. Yeah of the two I think I prefer the one from Tracy but I kind of don't think I have a I don't know that it has to be perfect. It's more I hate to say a filter but it's more a clarifier that this is what we're doing. So obviously there are opportunities for contributions for people who are new to the space such as you know editing and things like that but we're spending a lot of time answering questions like what is GitHub and how do I connect into Google Docs and as you know we've had some momentum issues we seem to have kind of a third of our group who are creating materials and then two-thirds of our group who are because what we're kind of doing them a disservice because they're coming expecting to receive training and not only are we as a group not developing enough content but the people who are coming to receive training are not getting what it is that their expectation is and so even though we're trying to support their efforts and kind of guide them in this direction or that direction it's difficult because of the amount of time and I know that all of the groups are experiencing this but people who want to learn also need to set aside the time to learn and so I think that they're thinking it's more of a we're a train or two them and so we're going to save them time rather than investing the time in learning so that they can contribute to the group so I like the idea of putting the list out we talked about training and education development learning materials development learning development etc so we'll put those out to the group but Tracy suggested that it would be good to find out if you had any particular preferences or because you're working across groups if there were any particular words that you thought would not work as much as words that you thought might work any other comments or would you let me to move on to the next item yeah feel free to move on from there okay do we do we need to do anything as a TSC or does that up to the working group to just change the name on their own we probably agree to whatever the name is I don't I don't think we're gonna I don't expect any of us is is gonna have any reluctance to approve whatever the predominant name is okay well I'd make a motion to change it as Tracy suggested for me Kelly did you want other names considered first or if you're happy with that one we can move forward with that I'm happy with that name and also I would like us to move forward as a group with our content but we could spend three months discussing the name as well so I think that this really solidifies what it is that we're doing and it will be a step forward so I'm good to go with this one if everyone is in agreement okay so we have a vote then all all in favor hi hi I think I almost counted 11 I'll oppose anybody abstaining all right sounds like we have an anonymous anonymous sort of approval here so we'll go ahead and work on getting the name changed to learning materials development working group obviously there's a few things we'll have to change related chat mailing list and those sorts of things but we'll go ahead and get that change the next thing that we're going to discuss regarding issues is that as you know I've been working as the kind of working group administrator and I'm gonna be making a change there I'll still be involved in the group and we'll support whoever is the working group administrator I've talked to Tracy and David and we'll be leading a subgroup project on smart contracts and purposefully calling that smart contracts because we want to be able to start to expand and scale understanding of chain code and understanding of the role of the various projects in relationship to smart contracts and so what we're going to do is with two or three participants from our group or new participants come up with one use case and then when it's complete another use case and we will create smart contracts within that or for that use case across the projects and then kind of review how it is that each project was approached also there will be some step-by-step tutorials because what we're finding as a working group is that everyone wants to contribute to about articles about white papers everything from governance to what is the future of blockchain but in the edX courses and in the various forums we see a lot of struggle with the actual technical implementation and so we'd like to present a use case we talked about starting with one that Tracy had begun earlier and we will write that out do kind of a mock needs assessment talk about the size of the quote company etc and then how would you approach that or what is the relationship potential relationship with sought-to-splits potential relationship etc etc and do a tutorial and then really kind of talk about what we ran into what considerations an organization might want to consider etc etc so the use case is really a combination of a use case with almost a POC or a tutorial geared toward a portion of a POC so that we can begin to address what we see people stumbling on which is really the implementation with several years ago any moons ago when Java first came out with Sun Microsystems we I worked at a college where we really saw that it wasn't getting past early adopters at that time and so we started trying to think of how we can expand to products that are familiar at the outset and then went into the technical detail rather than starting with the technical detail and working the way to a product so with Tracy we'll be working up kind of a a list of things that we're considering and then reaching out to the other groups regarding hey we're thinking about is this the right approach or this is what we were thinking about for this use case using your your project so it'll be within the training and education work group but will obviously involve the other groups as well that sounds like that will be really valuable and since the the working group agenda is yours to define along with your other active participants there's there's no reason in my mind you couldn't consider just making this the topic for for your meetings continue to pursue it the way that you're indicating here too but do think about just setting the topics for the agenda in a way that accomplishes the things that you want to accomplish yeah so I'm gonna add to it to that you know part of the thing here was I know back in June when we were in Amsterdam we talked about kind of what's it what's it easy for people to get started with the different projects and so you know the idea here was really if we're looking across the different projects with some one one particular use case that we could apply across the board that it helps people kind of you know familiar with the different projects really understand kind of you know what the what the right choices are for them based on their particular use case so that was another kind of let's bring in some of the things that we've talked about in the past to to this new well this we named working group right and focusing on that so our overall activity in the past quarter we worked on the markdown for the edX materials and then came up with the end user and decision maker concepts list the original was begun when we had the last quarterly meeting sorry I coughed a little bit no I'm gonna frog in my throat and then expanded that a little bit this kind of concepts list has been a little bit of a round and around for us because it showed how readily and easily the group is distracted the idea is good but you can go down a lot of rabbit holes and that's what we found was taking up a lot of time which pretty much led to the contribution also of changing the names and then tightening up the agendas so at this point that particular group is also in planned work products wants to complete that and then start writing more of an academic article on the effects on a democratic society and so because we have people in the group that are interested in contributing at more of a future level or a research level that's one of the reasons we decided to go with a smart contract project so that we could really balance that with the implementation that we're seeing that people are looking for in addition I created a prototype script and a prototype animation to help with some understanding of the concepts and so the group looked at the prototype animation and made a couple of comments we're working now on six additional scripts so we'll kind of go with that style of video and then that can be incorporated you know across different learning you know whether it be something that links for read the docs or whether it be something that people take into a course project we have a small project status list and it's just kind of a contact list and then we started in order to get contributions we've got a lot more people on the mailing list than we have showing up in the meetings so we decided every time that we have a meeting at the top of the wiki we're going to kind of put current current projects and current needs and then also we've been following up with a this is what happened in the meeting and here are ways that you can contribute over the next two weeks kind of a scrum type of a viewpoint but we were getting again more people who said yeah I would really love to learn that then so that's kind of the shift that we're trying to make as far as our planned work products goes they're finishing the final edit on the 36 concepts and want to begin the article that I mentioned they're also starting a fabric concepts definition document outline the smart contracts I brought up earlier under issues and then we will by our next TSC report have additional of these animation videos we've been having a consistent approximate 10 contributors we have more people on our mailing list about 200 and we have more people that show up to the meetings but usually we've got 10 people who are actually creating learning materials so I hope to have a much more kind of a productive or more focused report next time you know I think that we're getting there it's not it's it's a good group but just a little bit not on the charter and so that's why I've asked us in our next meeting we'll be reminding and reviewing the charter so we can keep things more in going forward in developing materials mode all right well thanks for the update Kelly and thanks for volunteering your time to keep this working group going forward and just getting it off the ground to begin with now I really appreciate Tracy too she's been instrumental in getting us through that yes Tracy is very helpful all around the board and hyper ledger but don't let that go to your head Tracy but thank you for the kind words okay I think we're ready to move on then to wait then I had one question I mean this is almost speaking so I was looking at the documents that you already have started drafting and so there is this thing about like there is a document on concept and you list a bunch of things and my only question is does that how does that relate to what's being done by the architecture working group because I see some form of overlap in that their reference architecture group is defining this kind of concept right in a way which is meant to be general enough to apply to the different frameworks we at least know often and developing within hyper ledger and I'm not saying there's a problem let's be clear I was just wondering if you know those things match with one another is that something you you keep an eye on we reached out to the other groups and asked for comments on this particular list I know that the person that's been working on it has been attending the fabric meetings but I don't know that he's been attending the architecture meetings but I can ask him about that and ask him to check in with that group for sure you know and the architecture working group has already published a few documents so it'd be good if we produce new material that it's at least a line and you know I think it can have a different angle you know point of view and a different audience and there is I'm not saying this is duplicative by any means but you don't want to be contradicting either right we should be consistent with all the documents we produce sure and this is mark I had a similar comment looked like you were defining terms some of which we've defined in the metrics white paper which are also you know I think the metrics white paper aligns with the architecture group okay and I'm not I think go back and compare to see how you know whether they contradict you know I'll take an action item to do that but I didn't know in general if we want you know like one one list of terms that all the different working groups can refer to or if it makes sense to have them in each paper hi this is alish I also have a question why we are focusing on fabric concepts and definitions and not go across all platforms that are under the literature in this case it's because we have very few volunteers and because our volunteers are kind of at the level of experience where they can select one thing and go into that one thing I don't think that they have the experience to write across and so I think the idea was this is a starting place it's a fledgling group who are learning at the same time and so it's I think we have one person who attends several of the fabric meetings and that's kind of the amount that he's able to both contribute and absorb at this point but if this group has ideas one thing that we did do is we last quarter put a note into each of the working groups to please let us know if there are things that we could collaborate and contribute on and we did hear back from Iroa and worked with them on some documentation but we didn't hear back from others so it's good for this group to help us to prioritize and understand where to connect what we have are a handful of people who are saying okay I think I know enough about this now that I can start to work on it versus people in this group who have a much better overall understanding which is one of the reasons why we're changing the name of the group and all just kind of feeds back into that level of experience of the group individuals in the group yeah this actually also happens in the technical working group in China when we start the translation work we found the volunteers that are familiar with fabric but can only find a few ones are familiar with other platforms so on the end user and decision maker concepts list then I'll ask that group to talk with the architecture group to look at the define metrics white paper and to I will post in each of the different working groups you know maybe please take a look at this etc if it turns out that this is not what this group finds relevant or if it's not the right approach that's fine because one thing that it's done is kind of gotten our people engaged if it turns out to be just an internal reference or if it turns out to be nothing that's fine it's it's served its purpose which is to get people putting something down on paper and to get them looking at the different working groups at the different documentation etc yeah I do like the direction that you were setting of trying to create some some concrete examples and growing out from there I think that'll be a more productive use of viewer and the rest of your group's time then then trying to make high-level papers that might end up with some of these duplicating or conflicting concepts that the other folks here just mentioned okay and with that we should probably move along in the agenda so thanks again Kelly and then Hartz you are up awesome thanks Dan just reaching for the mute button so I guess we should just continue the discussion we had last week so I will say we have updated the proposal to reflect a lot of the changes in feedback that people gave at the last meeting in particular Chris's suggestion about maintainer lists rather than stewards was a good one and we have we've changed the proposal to reflect that so I guess I don't want to exactly rehash what we what I presented last week or what we said so I'll open up the floor to questions comments anything like that this is mark I was happy with the changes awesome does anybody else have other questions or comments or why not Montreal do you say why not Montreal I think you said Hawaii no I said Hawaii not Montreal sadly that question will have to go when we move it to the the official uh recall I guess darn anybody else I mean Chris did those changes reflect what you thought yep yeah no I'm good with it first part of Dave I was really pleased to see those modifications it really clarified that's a long line to it yeah the risk of adding to the deafening silence I I'm good with it this looks like a nice proposal okay well I see no reason to belabor this further heart would you like us to move this forward sure all right well why don't we take a vote to prove the the cryptolyb proposal so I do have a question before we vote did we decide on a name for this we have a tentative name choice it is Ursa we need to run it through with the marketing committee though and I was actually going to ask about what's the protocol for doing that so I think in the past we've approved projects and then they've gone and figured out what the name is with the marketing committee afterwards I think that's what we did with the the interledger proposal yeah bring that wrong we did yeah you know I I don't know what was done for sawtooth but I know for fabric IBM did a you know you know we did a search so we may want to do a search but you know with the qualification of hyper ledger fubar it it should be okay but having having the marketing marketing committee and probably our legal beagles approve it would be good yeah that's what I was going to ask about so if I I'll follow up who should I talk to on the LF staff about this hmm yeah the marketing committee so it's probably worthwhile just sending to marketing at hyper ledger.org it'd be dano pray and now alissa is the vice chair I already was going to send through Dan I was wondering if there was like a legal someone or something that I needed to well that they'll work with council okay just send them the marketing and you know all right so then I guess we're ready for our vote then we'll we'll get heart you'll interact with the marketing committee and we'll figure out the name but just I'll move who you'll move for the vote Chris okay second great all right uh all those in favor of the curriculum the proposal uh becoming a project within hyper ledger uh say I hi hi hi hi hi hi okay all those opposed no anybody abstaining all right sounds like we have approval um so heart if you wouldn't mind just following up with the marketing we'll go from there awesome thank you Tracy yeah once we get the the name we can set up all the uh all the mailing lists and chat channels and well I guess you know maybe the name chat channels if we have to um and uh yeah get it set up on the website as well so thanks a lot of people I'm hard did you incorporate the rfc repo in the the final uh proposal uh no but I have a link saying that you know this is what we're thinking of doing and it will probably change heavily so I implied it but I didn't like the most recent content yeah I want to make sure Tracy knew yeah I was gonna say some of the some of the proposals have a resources section at the end where they list the the requested um repos and we don't have it in that proposal but yeah we can sort out those details we at least want two or three yeah it was just a decision we made yesterday morning so I was making sure it got into the into the repo setup hey Dave what was that uh what what was the agreement so we can we can update you offline I think it's okay yeah we'll do it off yeah all right and uh Dave you're actually gonna be the the next on the list here congratulations first to uh heart and and the rest of the crew which is actually quite a few of us here it's quite a big crew so thanks everybody yeah it was uh it was a long time was it back in LA that we started that or was it even before then I believe Nathan likes to point out that it was on the bus in Madrid right Lisbon yeah it's been a long time Lisbon sorry down here yeah the double ducker yeah on the bus in Lisbon yep I will I will corroborate that story all right well it's trams trams all right uh so Dan sorry I lost the agenda list here what am I talking about so uh if you're prepared for more discussion on testnet we have time for that if that's not a discussion that you've uh that you think that we need to take more time yeah um yeah there is no agreement so far there's no idea like we haven't even settled on exactly what the needs are for a testnet I mean of course the the the story that I've posed so that we run like Kubernetes fabric and you know Kubernetes um sorry instance and we can compensate people for you know some amount of usage of that but um I don't see a consensus developing around any of the proposals so I would like to gather feedback right now from the community from the TSC about what a testnet means to them like what would you like to see because there's a huge range from you know as Brian likes to joke why can't this just be a cluster of raspberry pies although we have to do we have you know big iron running somewhere that can really do performance testing so so but Dave I think I think it gets to the sort of the the fundamental distinction of is this a uh sort of a Kubernetes replacement for what we already have with OpenStack through digital ocean or wherever our provider is I can't remember or is it intended to be where sawtooth and fabric and burrow and and and roja and so forth are all stood up and people can kick the tires as users of these things there's they're very different beasts right I mean one is an extension of our CI that allows us to do scale and performance and various other types of testing that's you know more than what we can do by just testing inside a single VM right now I think all of the CI is basically just grabbing a VM and doing Docker compose up and down to run a set of tests but it's not going to get you any kind of you know multi multi-node testing it's not going to get you any kind of scale testing or even performance because these VMs are pretty slow so I well so Chris to clarify I I saw this as more of a place to spin up an a test instance of an application under development and just poke at it I mean I know we can do it locally but there are some aspects to these decentralized applications that really need to be run like in production but so this is like somewhere between testing on your desktop and being in full production that's how I thought because the CI pipelines already have soak tests and performance tests going I mean I know the sawtube team does that and I know the fabric they can only test in a single VM because we're using open stack basically is spin up a VM you know we'll just do a bunch of containers inside there and kick the tires we aren't doing anything more extensive than that as far as I know on any of the projects I know that Tong had worked Tong Lee had worked for quite a while to spin up like four node using cello right to spend up four nodes in in our service provider and yeah on running tests and I don't know whatever happened to that because I know I'm on the team yeah that may be the case I don't know yeah well on the sawtube team we do something quite a bit more significant than that so we're we're regularly spinning up 10 node networks with with between medium and large nodes on AWS and driving max load at those for periods of one three and seven days depending on what what the purpose of that test cycle is I'm assuming Intel is paying for that Intel's been putting the bill for that so right I mean we're doing yeah and IBM is doing something similar internally where we're spinning up on the IBM cloud doing some things well and we I would add there that the indies had to do the same thing and one of the things we've noticed is that if you don't do a network that scales across data centers meaning you don't do a globally spanning network you don't see a lot of the same performance issues because you've seen it all in one data center right and a lot of the value of this sort of network is is the interoperability testing meaning that you're on a network along with other folks because if you're just doing testing for just one thing or just one set of benchmarks you know doing that privately is is usually easier than trying to coordinate with a pool of nodes that are being run for other people for other purposes and to give my perspective from uh from borough yeah we're doing something similar with long-running test networks of between like seven and 12 nodes that we might run for up to a week or two weeks sometimes what we'll do is we'll just continually redeploy a load of Ethereum contracts or create some synthetic traffic on on those yeah we're putting them across different data centers and things like that oftentimes this is revealing issues with borough so it seems to be part of borough testing and currently MONEX it is paying for that so part of it would just be it would be nice if the cost of that was born or mutualized with hyperledger because it's you know not in significant cost for a startup to be uh be running that stuff um I think yet so so for us we run on kubernetes so like a kubernetes namespace that did borough would be ideal for what we'd want I think there might be another another angle for this once these networks are running uh that maybe there's an ability to start playing with stuff um between different namespaces but the the the other thing is that unlike in TI test or benchmark that might run to completion it's useful that when these uh networks get messed up if you can manually go in and poke around so often on some shorter automated tests I might have seen some issues uh but then I lose the state um so it's kind of nice if these are semi-manual for me anyway yeah so sorry Alison I think the the key thing from my perspective is that with everybody essentially doing stuff behind the curtain of um uh you know their own corporate uh test environment and so forth we're unable to sort of leverage the community right so nobody can help come in and write tests or you know examine the the output and and contribute in some way to that aspect of things because it's not public I'd like to make it public personally um I keep pushing for more and more of our internal testing to be externalized um and with every release we're doing more and more of that um uh transferring things out into into the community but again when we start thinking about running scale of you know let's get 128 nodes or something like that running um uh and then tear it down after a while that becomes a very different prospect in terms of how you know we can allocate uh our our our test environment and it's probably going to you know be the case that again if we're if we're if we're running kubernetes for instance pardon me then you know does lf provide support for running kubernetes right or do we do we get it from you know an ibm or an azure or whoever right that's an interesting point you bring up chris because one of the goals i had personally was to use the tests to drive the development of the the standard you know kubernetes composable configs and helm charts the the effort that we began this summer to try to make these things deployable across cloud platforms much easier so that would be one way to engage the community in your testing process would be hey we're going to build this really giant you know scaling tests we need to you know here here's our kubernetes configs and everything like that and have all that be public and then spin it up on the network yeah well i love that i mean and that's you know i think you know the the six four thousand dollar question and i don't know if it's costing that much or not but is again you know if this is something where end users are not participating as part of the development of the various projects can just spin up a fabric or sawtooth or a roja and so forth and and burrow and kick the tires and run smart contracts that they've written that's that's a user test net that's like the ethereum test net right you test net for ethereum not to test ethereum is to test my smart contract and if right if we're instead saying no we're testing burrow we're testing fabric we're testing sawtooth and so forth then i i would like to see that and we should you know seriously consider if we're all doing or leveraging kubernetes to to do sort of more production like kind of deployments then i i think we should explore you know not you know how do we how do we how do we get a case you know service that we can share as part of our ci just like we're doing with i can't remember rye who it was but so chris just just to make sure um kind of understand that all of the major platforms are doing some level of large-scale tests with the resources that they're contributing uh ibm is contributing so for fabric intel is helping to support alias deployments for sawtooth we already have this sort of controlled uh long-running high-load tests that we're running what what i hear from you and nathan and others and what has been my experience is um you tend to fix the control tests but it's really hard to get the uncontrolled tests um you know where you're running on a network that drops every third packet for some bizarre reason or you've got you know some highly heterogeneous set of links in it that that suddenly expose some timing problems are when you talk about testing the platforms in this case like what you're suggesting with kubernetes is the objective here to get those to identify the randomness and the weirdness and the unpredictable situations or is it i mean the other part of this is the other random factor in here is what do people what do real people do with my application that i didn't really understand they were going to do with my application and i just need to get something out so a community can it can abuse it um well i saw i can i can only speak to the kinds of things that we're looking at and basically we're looking for the things that um you know that nathan was sort of articulating that there are certain um you know when you when you when you when you cross network boundaries when you are crossing data center boundaries when you are working on um you know with with significant scale and or performance then you start running into issues that you might not otherwise uncover if you're just doing sort of small-scale testing right um and so the intention here is to find those things in the community rather than behind the curtain of some company's own you know sort of i'll put the word proprietary and it's not really proprietary i mean but proprietary testing that they're currently paying for but that is not visible and not accessible to the community either to review the output and or contribute to improving those tests and so forth um i mean i'd love to have all the chaos and scale and performance testing done out in the open for fabric that'd be wonderful from my perspective and i suspect you know uh saw to you know everybody else feels the same yeah maybe one other thing i can add that isn't uh initially repetitive with what i've said before is um if we wanted to open up the networks that were already running which are in a way open but not advertised the an obstacle for uh to me for that is just the the security administration of of the hosts themselves so as you're thinking about what we could potentially do with your budget uh certainly the cost of of operating the systems is one thing from from like uh the cloud cost perspective but if you're looking at adding on a sysadmin or something like that think about how many of those systems that sysadmin could um protect and and manage maybe there's some sort of baseline figure where we can say even if hyperledger isn't hosting 100 nodes uh you know if they're hosting i don't know let's say just two nodes per project just to have a number there uh the sysadmin cost around that it's um i think you know you mentioned that you were hosting iris on um amazon amazon we're actually using sort of the internal version of the ibm cloud so it's you know charge back as opposed to you know we're paying a bill and i can't just you know share the ip's and stuff right i mean i need their place to bring stuff too i could it could be the ibm public cloud and but you know then there's a whole matter of billing and everything you have to sort out yeah i'm just trying to throw out some ideas for budget so regardless of where it's hosted having having somebody to help secure those that's a cost that i don't know how to deal with right now uh the thing we already talked about is the the cost of the platforms themselves and maybe you can do two different tiers there one where there's uh just say a couple seed nodes provided by hyperledger and then the rest of the community and the other companies are adding the the girth of the population and then a second tier where where hyperledger is funding a lot of nodes and that might be important for for projects that don't necessarily have as deep of pockets behind them and then there's there's there's also the sort of the cross cloud you know um you know testing that you know something that's you know where there's nodes that are running on amazon on azure on ibm on google wherever right and making sure that that all so there's a little bit of a tanda for oh sorry can you guys can you guys hear me can you guys still hear me okay hey so that the if we hired somebody a human that significantly raises the table stakes um i'm also concerned about losing the you know sort of the non-accounted for financial contributions from the member companies i would love to still leverage that so just to throw it out is another idea i don't know if it's even possible but it would be cool if the member companies that are already running testing right now to provide um cloud space for running pods so that we could have a heterogeneous pool of pods to do kubernetes deploys on so we could actually do you know heterogeneous cross provider deploys of these networks and i also like the idea of limiting this to just teams right and not opening it up to sort of the general public although um and i mean what i mean by that is like our working groups you know i think the educational materials working groups should be in there you know the documentation and you know that they're using to to teach with to make sure that the steps are accurate um architecture working groups should be able to poke at stuff along with the development teams right and i guess if we did the possible yeah sorry arnaud let's try to get in here a couple times yeah thank you i mean so no i just wanted to bring up another aspect which was mentioned in the past and that has not been brought up in this discussion at all which was the this notion of having some test nets for hackers because you know as the bug bounty program was unsuccessful and one of the reasons seemed to be that it's too much of a hassle for hackers to get even the network set up to get bounding on you know and so we had mentioned uh you know the possibility of redirecting some of the money that was initially dedicated to the bug bounty program to setting up test nets so hackers could have an easier way at you know getting started yeah um stylus i was a little surprised that that bro didn't chime in that that's having a test net in the ethereum test net sense was didn't seem to be your first priority so yeah i was going to come in to respond to a few things going back to what mick originally said so the one thing is the way that i think i would see these networks at least initially would be that human level of investigation so you often might end up writing the automated tests after stuff you discover on these longer running networks when we're testing um we end up throwing application level code at the at the chain because that's the one of the better ways to start um generating types of faults and issues so i don't know in so far as it will be possible to uh to be a little vague about what they're for initially but actually lower the barriers of getting them set up so i like the idea of a multi master kubernetes um deployment across different clouds i like the idea of a a base load being provided by by hyper ledger i think there's also like a human administrator would be really handy this so there's a few cluster level admin things there are there are pain that you could do if if an admin was managing the masters and maybe just a few extra nodes there's also things like elastic stack for logging um cabana and stuff like that just for looking at metrics all of that stuff could i would have thought be reasonably easily shared so if that was there then i think test nets whether they're for some level of application i think it's maybe tricky having like we don't want to have to provide a high level of service to just random punters wanting to deploy applications but i think we could probably have a better idea about how exactly they should be used once we started doing investigative test level stuff maybe some benchmarking maybe some like um inviting hackers stuff maybe some you know putting up some applications that that we know put strain on the chains like i don't know whether we would have to decide between those to start with but having some level basic level of infrastructure that was somewhat managed would be would be useful well hi this is this is ladder just to have my observations i think everything that we're saying seems to point us in the direction of having a governed environment somewhere in the cloud that we can call a test net but also based on the performance and scale metrics that have recently been published this is the point of that work in group and from a collaborative um an interoperability perspective if we can bring all that together to instantiate a test net that each of the projects can use that can be configured to provide a benchmarking that's is it was that Sylas who just talked about and um and the rest of the team um so literally to me that would be money better spent because you could then have a you could then have a charge back model back to the different projects as well as hyper ledger something like that would have value that can be configured for everyone um going forward thanks thank you and we're at the end of time so i'd like to quickly propose a potential right turn here as i'm looking at cooperating with some of the universities the universities in the boston area of what's called the mass open cloud where they have thousands of physical machines that um you know maybe we can get time on those and maybe get students to manage them you know if we set up some stuff with universities or something yeah i think recruiting from universities or elsewhere is great so we are out of time here Dave clearly there's many ways we can spend any amount of money that that you can uh you want to suggest on the budget so please keep pursuing that and thanks everybody for your time at today's meeting thank you everybody thanks everybody thanks everyone