 Which brings us to the last part of today almost made it still half an hour to go I hand over to Stefan and we will have a panel discussion here And Stefan is going to invite some of the people here to come in front and we will put up a couple of chairs and then try to have a discussion This answers my first question if we do it standing up, so no we will get about we need about seven chairs We can do it standing up perhaps So we are seeing each other where we are urged to make it short and So, okay, I am please Emily Peter Roch Alexander how it is already there and oh my please I would Like you to join me and Yeah, and so I think I Don't need to introduce anybody else except from Peter. So Peter maybe How else you hand over the microphone to Peter? So he can introduce Himself and his project shortly so that everybody it'll be it'll be sure nobody else really introduced himself either My name is Peter bloom. I coordinate a project called Rizomatica biggest thing that we do is work on GSM policy Regulation for community networks, and we also run a fairly large network in the box network in Mexico Thank you, so We just would like to close the session so standing up with some gymnastics no we are Just I have From today, I took back even myself. I'm not too deep into technique technology I take a lot of information today with me still I think for me there some Partly software some partly technical question open and I would just like to ask the people here at on the panel because so we we have I would say we have partly operators here we have No, we have real operators here to I would say Peter and rock we have Base-stage manufacturers here. We have people here who are working on the management side. So I think we have most of the Users for the Osmo come second on the end users but the users of the Esmo comes tech Osmo comes tech from different areas here, so And the first question I think halted may not even answer because he answered it a lot today, but Perhaps starting with Emily. What are your expectations? To Osmo come so technique Technologically or socially or however sure. Um, well, I'd say first off we're very grateful that we can use the stack and We're very we're very thankful that it's such I don't know there's this this community that can support us in and using it so I think one thing in particular that that we would like to see from from the Osmo come project is is Quasi stable to stable release pattern or some sort of not pattern necessarily, but some sort of Package or binary that was that was blessed as stable Or quasi stable and our reason for that is because as we're trying to develop hardware With software that is itself developing on a nightly basis We we kind of end up with a lot of unknown states And we get into scenarios where it's not clear to us if the if the root cause of the failure is the change in the Hardware the change in the software stack or or combination of the two so we end up pinning it to Certain get commits But that kind of becomes a little unwieldy and it's a little hard to communicate State verbally and if we could just say yeah, we're working on release, you know x.y It would be it would be very helpful. Have you wish for a frequency for that? I I don't think we have a strong opinion. I mean, I think you know something on the order of months Not years and something on the or and you know, I think not days. Yeah Okay, Alexander, maybe the same question to you. So it's also for example Release dates or let's say more distinguishable release numbers. It's this important for you or what are your aims? Yeah, I was actually is working. Yeah, okay. Sorry. Yeah, I completely support the idea of releases and Again for us as a hardware manufacturer, it's much easier to what would be much easier to deal with some releases. I mean we actually maintain our own fairways master branch on some of the Osmo-com packages, which we stick to because we know they work because we've tested them and This is like our Releases, but it would be nice if there is like official releases Which can be in blessed and then we can just use that Yeah, so in Perhaps one one other question and continuing that to Peter then is is more stability important for for you or more the upcoming Rise of feature sets. What is the more important part for you in the future? Both obviously I think stability is the key we've been running a network now for Three or four years on this and I think it's definitely improved But definitely when we started things were much shakier than they are now. So as things get towards becoming more stable It's made it a lot easier for us to do our job because we're generally working in you know resource deprived areas where You know, it's just hard. It's hard to do anything. It's hard to run a network too so and especially with few resources so as as stable as the The underlying software could be for us I would say is even more important than having Then you know fancy new features like if we can make sure it works It does all the basic stuff it needs to do and then know you can be able to tell folks who are essentially Our customers communities and end users that you know on this date. We're gonna have this functionality That would be fantastic So even if it's long periods of time before that if there is some sort of clear ID and that the roadmap is tied to some Yeah, some dates great, but for us stability is at least for now the number more important. Okay. Thank you Omar you touched the Osmo comes back I would say from a quite different layer with CCM so What are your experiences and what what is the expectations you have in the future? because of these experiences it's been overwhelmingly positive actually coming from a a You know working with we work with open BTS initially and This is a lot more stable like you see for instance how quickly phones attach to the network and just the A number of crashes is much fewer We don't run into like You know bad state with The entire system going to into lock And I want to see us moving Further along this trend. I want I think this has the potential to be something. That's you know carrier carrier grade we're right now running it with a carrier and I Think stability is super important. I know you can't design us. You can't design software. That's free from failure So what I think would be great if we could focus on making the system as a whole resilient so that if one component goes down the Network can sort of resurrect itself and Restore to a state. That's you know fully operational and I think that's part of what Community seller manager tries to do and saying that more natively in the stack I think that would that be a huge boost for stability In the way for example that we heard today that there maybe could be plans to migrate to to better documented or more integrated control interface Could this break partly the interfaces you're using right now and how much should then the stack be Backported in a way or what do you it's all could you handle that if we have API changes there We've we have a lot of integrations, you know system. We only move the stack to system com last year But we have pretty good abstractions that We can switch like one of my work items is switching from VTY to using the control interface So You define good interfaces. It's easy to change the underlying transport so It's all open source if there's new features, you know love to see those Contributed back by the community as well. Okay. Thank you How about would you like to add something to the roadmap you told us about this afternoon? Not really I mean it's No, but I mean it's it's I touch those topics I think I mean talking about more stable releases and we talk about Osmo TSM tester as a tool to have like You know nightly end-to-end verification with with test coverage. I was talking about more unit tests So that clearly goes in that direction. I didn't speak about releases You know being a die-hard developer Releases and maintaining stable releases is like is to me is a burden. I have to be honest But I mean we have to find ways to do this I mean doesn't have to be me doing that, of course We can find somebody who gets excited about maintaining stable releases and backporting fixes and so on I'm perfectly happy to to to have that in the project And the other point is also regarding controller phase. I mean we also discussed that that basically yeah We if there is requirements and now that your code is released We can see what kind of interfaces you need and and basically improve on that also a third part that I also touched in the Which maybe was not as explicit in the roadmap in terms of the resilient the system restarting itself and so on I think if you look at the Osmo FSMs in the new VRM SC code, you will see some of that Of course, it's all still see and if the process crashes the process crashes But at least the individual sort of logical state machines. They all have like a parent-child relationship like in a Unix process model even inside one process and if a state machine dies or something like that Then the parent gets notified and it can recover and as always like you can have a default timeout And then the the children state machines get killed and can recover and so on so that I think there is some some of that in In the Osmo FSMs as well. So that goes in this direction So rock do you the last Statement about that How frightened are you because of the split of HLR to MSC does it change something to your structure have you to adapt? Well, I'm not completely the best one to answer there because I still take a blame for having run an experimental GB proxy for two years Which meaning which put us into a serious amount of troubles after at the same time I think it's one thing which is really important is really to have the daily the daily test Against against real hardware because I remember that something like two years ago There was a major change in the way where the the BTS was sending the SI The SI information that broke everything that was not 1800 megahertz And it's meaning it's so it's the type of things that could be discovered if we do if we do test meaning on the very Yeah, daily basis Against real hardware. Yeah, okay, so just perhaps just keep the mic and We are stepping over to to my next question only to left so In beginning how I'll mention that maybe the community is not White as widespread and active anymore since as it was in the beginning because maybe that 2g is just outdated for the people are not attractive. Do we have a chance? To to make it more attractive to get outside people Entering it again. So maybe because of IOT things could could this be a direction? What what could we do? How could we do marketing? I'm not I'm not so sure that actually it's it's less You know, if you if you look at the number of projects, I think our ad was mentioned something like 72 The issue is that as we have that huge number of projects. It's really difficult to bring New meaning new people in yeah, I mean if you want to deploy the entire open BSC Meaning Osmo or most more come environment. I mean today you're talking about a lot of Meaning a huge amount of effort In the past all you had to do was basically to compile Osmo need be Fine and I find an IP access BTS and you were you were good to go So but I think it's something that we might see again with the the project that you launched with the With the 3g BTS meaning where we we will see people taking those meaning buying those again Run run them and make tests and have fun with them. Yes Maybe this would be good switch over to Harald. What do you think about the three three point five g acceleration project? So boxes were sent to have good expectations, or do you think oh you will never hear anything again from the people? Okay, so maybe just I do the politician thing. I don't respond to your question, but do something else first So more interaction what rock just said The many projects I agree we have many more projects in many more areas though Which means I mean we have all kinds of projects that are not in the scope of GSM So people are excited about all these other areas, but GSM sort of is the old Stuff that has been around for almost 10 years now, and it's not exciting anymore So I don't really think that yes, there's more people doing open source mobile stuff under the Osmo Come umbrella or outside of it, but in terms of the GSM projects I'm not so sure and in terms of 3g. I think it's also too late to get many people really excited about this now Even with the accelerates 3g 5 project and so on so my expectation of the acceleration 5. Yes, definitely I see we had lots of interesting applications from people in the community who want to do useful things who also some of them have Started and we can we see some progress there So yes, it will help us to get some more community involvement, but I don't think it's a Solution for the overall problem Okay Well, just just one thing at the same time Meaning this is not a normal project in the sense of it's not Libra fees It's not Mozilla Meaning it's we're talking about a radio. It's licensed Meaning it's not something you can easily try at home Exactly, so I think we we have to live with the fact that the community is never going to be gigantic. Yeah But Maybe then if you just make the round once again to to oh my yes, the the CCM is Yeah, a little bit the same in this area So you have to find people who are people organizations who are contributing to that and how to spread the word Yeah, so if you want to get more people on the platform you have to make it Attractive to them easy to use and offer them use cases that are We can incentivize them to spend the time to set it up so People I think more people run services like this that manage your networks that becomes possible to See this in more use We'll see where How community seller will grow over the next few years That there's more policy involved in that than anything But I think for the average user if it becomes available on more distributions like I think It's become a lot easier to get up and running now with With Cismacom with the Debian builds that are Are the nightly Debian bills that we have if you can also go out and reach the Ubuntu users the Documentation That helps a lot. So right now there's the auto-generated docs that are published like being able to reference that made it a lot easier to expose a lot of the functionality that We integrated into CCM so Yeah, just the more documentation there is the more people discover use cases and be able to Use it for for their own needs That's right. So So so maybe this would be even you have to in your setups. You have to train people to use At least deployments and perhaps to use also part of the Osmo comes back and configure that so is Documentation is a very important key to Yeah, I don't think that's that's not really what I wanted to talk about. Yeah, I think it's definitely important Of course documentation is important. Yeah, I mean yes, but I'm gonna take it away from the technical piece for a second because I don't really know anything about the technical part But what I do want to say first of all, I want to congratulate the community for having this day Before the developer conference. I think it's incredibly important. I think that it's something that should continue The idea of there being this community is incredibly important I think that the community itself though and we're sort of talking about more from a maybe technical standpoint from looking at it as an ecosystem, I think it's in a an Interesting place, but a complicated place wherein the stack as has gone beyond, you know Being a hobby and has become the basis for a number of businesses that produce hardware that produce software Social enterprises or whatever you want to call like us at production networks even And I think that that dynamism is really important, but also introduces lots of complications into I think what was sort of ideally or Was conceived as sort of this pure, you know, this is just our common Infrastructure that we all share so I'd like to see Somehow more opportunities to have some of these other discussions more opportunities to for folks to share information with each other Beyond just the technical piece and so for example, like we do a lot of work on on regulatory aspects And I think there's people in this room who would be great use cases for us to Like how can we all help each other make sure that this thing Osmo come stays Strong and stays vibrant and maybe doesn't grow tremendously, but it continues to exist and that sort of common Base from which we're all working continues to grow become more stable become more feature rich and so on and that each of us Here has something to contribute to that. Maybe it's a patch. Maybe it's not maybe it's inventing hardware Maybe it's working on, you know regulation or whatever any of the other people in this room might be working on so I guess I just want to say that There's a lot we can do to strengthen the community that goes beyond just sort of Commits to the code, right? Yeah, that's right. So maybe we could start in a Real-round table once a month in Berlin. So unfortunately you're traveling from Soap Road I'm not sure or extend this event to another day or have a different type of format for half a day Or there's many you know lots of I have ideas that we can talk about great Emily you have some additional ideas to yeah, I mean, I think one thing Facebook could do in particular is help with exposure because I think the the target Users of of these stack and contributors to the stack are are are perhaps different people but but as we get like I Mean as it's going more and more towards towards businesses, you know, if If within the telco ecosystem if there's a More recognition of this is a very viable option of how you could run your network I think that that would then kind of continue to increase its profile and encourage developers who perhaps have their own forks or You know are are working maybe In a different in a different domain to to move over to this domain and contribute back because it's now become a project with more visibility And I and I think that's something that you know is part of the goal of our project under the telco infrastructure project tip At Facebook is to help kind of grow this ecosystem and we do recognize that Facebook has a unique You know, we're in a really privileged position. We get a lot of visibility And it's definitely something that we can work to to share Thank you. I think that would be a good opportunity So Thank you. Yeah, so I was always trying to Let's say Push the idea that Osmo commutes more Promotion and more marketing and you know, how old remembers I brought this topic up and every as much if con previously and I Try to do my my best and I'm glad that we now have like partners like Facebook Helping with this because I do believe that, you know us open source is basically working on the laws of big numbers So on the hundred users there is one contributor at best So the more users we have the more contributors we have So to have more users there should be more visibility. It's, you know, it's typical typical process. So and Harold recently wrote in one of his emails that in Germany you say that I'll say self Promotions So motion stinks. Yeah Yeah, so so self Praising self-praise things right and I guess that was kind of Unfortunately, I would say curse for the project because I think it needs more self praise So Compared to I believe failed project open BTS We should see the lot of promotion and still a lot of people are referring to open BTS even though in technical technically I believe it's far Lower than than Osmo come right now. So there should be more Promotion, so I would appreciate it If everyone in this room and everyone watching us and reading us will go out and speak at the conferences You know mention the project and One interesting thing which happened open BTS is an orally book published So if there is someone want to want to volunteer to write a book about Osmo come I believe that would greatly help promotion of the project so That's all of work But it's doable any volunteers free drinks this evening Maybe not now, but I just want to put this in I would say in people's mind that this would greatly help the project because a lot of people start from a book and And general documentation is very important like Osmo come documentation is very hectic as we all know so That would be great to improve and but to go back to the question of How to get more contributors? I believe so again another thing Harald recently wrote in an email is that You know, we are working on this project professionally so and we have to eat so we need to get paid which means that like as The number of contributors will also grow as more money gets into this Into this project and so this is The other side of this it's a community project, but it has a commercial side The more money gets into the project the more contributors can be paid and more code can be written. So Let's make more business with this, you know, let's deploy more base stations, you know Let's let's get this into production and let's get more money flowing into this community. So more people can Have not only have fun on their free time, but have fun on their pay time as well So you already answered my last question, which was How can we do real open source? So no open core model no crypto software just do real open source and still make a living Not driving as I hear sometimes not driving big cars not sailing the biggest chips and I was actually Nothing I wanted to say is that when I just started working in open source solar communications, I wrote a blog post saying solving an inefficiency of of telecommunications and I think this is partially what answers Partially an answer to your question is that To the communication industry is very inefficient because every company reinvents a wheel reinvents all the standards and as Harald mentioned It's like thousands of pages of documentation. So it's thousands of lines of code So every every company in the industry Invests million of dollars in re-implementing this wheel So what I really want Osmocom to be is I wanted to be the last implementation So no more after that so Well, I think I don't think anyone else will implement gsm after us. So I think you're right But I hope it will go beyond gsm, but I think that could be a good marketing standpoint so yeah, the point is There There is an efficiency and as a commercial company We want to have a good solution So for us, it's important to have this community-driven project because it helps us reduce our costs in terms of investment in our NZ because this is a common platform like look everyone has gsm Right just having a good gsm stack is not a differentiator for us. So We want to build on top of this and that's what we are doing It's actually part of what Harald's asking. So why is the number of contributions decreasing? Well, partially it's because the code is good Yes, not not at all not at all all the places and we'll discuss this tomorrow But in general it's good enough for many people to just run it To build on top of that Right That's what rise a matica has done with their Building system is everything they just built on top of it. That's what CCM is doing That's what we are doing. So yes, we are contributing in places where code doesn't work But in most cases it just does work Just just a short short edition you earn your money by selling BTS's Using the Osmo comes back. That's so that's that's how you how you could For to contribute to the stack. Yeah, okay It's just to give to give the audience also the idea how we make money because it's it's for me Since since I'm doing open source is the oldest the question Where to make the money from so it's always a question of a little bit of subsidizing and so it's in a way It's also you're selling the hardware to subsidize your contribution to the stack Yeah, you can you can put it this way, but we are essentially I'll say scratching our own each in most cases So it's not so much of Subsidizing this development as much as it's like solving our own problems with the money we earn from Selling base stations selling support to this and networks You know building building networks because we're not only selling base stations We're actually going out and okay works for operators As well as like full-plated networks and all we also sell they just SDR hardware So like we have like multiple streams just selling hardware selling base stations selling networks But yeah, because we have the need for the software We obviously yeah have to invest in some of the development related to it So Emily house Facebook making the money to I think Going to the question of like how how can you make money at open source? I think I'm probably like one of the least qualified people here to speak to that because there's a lot of like Entrepreneurs here and and people that are that are making money for themselves where I work for a big company and I I just I benefit from from what they have done So so I so I don't have any answer to that question, but I do think that there's a lot You know Then the there's a need, you know, I think a lot of a lot of maybe a lot of software solutions Are are more incremental or more? Luxury, but to connect people who are like unconnected or very under connected is is definitely like a huge It's not incremental. It's it's it's it's you know a huge leap. It's it's exponential So so I think you know some of the things that Alexander spoke to you know, just more self-praise more promotion it won't only serve this community and this project and the software but also the people who are her unconnected currently and you know that Resomatica and and fair waves and Facebook are trying to to help connect Okay, thank you Peter I mean in your project you're even not Well, we have a different business complete different so we're basically a not-for-profit organization, so we Survive as it were by you know making enough money to continue and pay people and do what we need to do Generally through grant funding. We're looking into more sort of like development aid funding I think that for this group actually as I've been here today in the last couple days I think that that's a piece that we should look more into I think that the the sort of Other than a sort of aberration of what Facebook is doing. I think that the like corporate The amount of money that the businesses that are using this are going to be able to generate and put back into the community While it has been an important lifeblood is never going to I don't think is never going to be a huge amount until all these other sort of pieces get put in And I think that there might be an opportunity to look at other types of funding streams that aren't sort of based on this provides value to a business who then Sells a product and then put some sort of portion of that money back in or yourself subsidizing by Essentially selling software that uses the stack. I think that that has sort of some sort of a horizon to how much that's ever going to be But I do think that there's really important Shift going on around the connectivity of people who are unconnected Which is over half of the of the planet and 2g Is a great solution for those people and there are Increasingly interesting business models and there's increasingly Becoming opportunities for small operators to partner with large operators or sort of not-for-profit operators to begin expanding and that Creates a lot of value And I don't think right now we have any sort of concept of how we can capture some of that value and stick it back into this project So the like contributions that we've tried to make here, which are you know minimal But we've always tried to at least Help out with whatever we can have come from from grant funding that we've gotten to do work And we had a need and then we sort of put and pushed money into the the development of certain pieces of things that we needed I think if we were more targeted about that we could actually probably raise some pretty good money from Even the German you know you could talk to the German government about this the development aid of the German government I know I People are very skeptical about this, but it can actually work because this is becoming a priority for governments And it's all about the way that you frame it to them So if you say hey look as part of what we need we need to you know contribute X amount of money into this thing So that we can more efficiently use satellite backhaul For this and that's going to require this many hours of development work And you know these are these can be multi-million dollar projects in many cases So I don't have exact plan for how this can happen, but I do think it's a space to look into Because this really is beginning to fit into larger sort of almost geopolitical Ambitions of different nations and even companies as well that they're Attempting to to connect people and how can we also make sure that that's happening with open-source software and it's happening You know through means that it will give people more control over over that destiny Maybe that's not a bad approach to really Attach in your countries where you're living and working to to get funding from governmental funding We have not even looked too much in that as being a small company there But I think for the project itself for the sake of the project it will really help Yeah, so I Deployments I think are one way that will You know get more people using the code base and hence get more developers so Facebook's in unique position that It has good relationship with operators and through these deployments that We're doing with with the osmocom stack We Get them to think about different ways of operating their networks that can potentially change Regulation in the future that makes more deployments possible so Even with the last few sites that that we're seeing like the operator was Actually themselves they were able to roll on more sites because they're involving the community And this is a stack that can be managed by the people living there And there is that there is a need so the people in these communities It's effectively We're able to lobby The regulatory bodies to Sort of hasten the process of setting up a new site so The more we see this the more we see community stake ownership of their networks And we see deployments that are credible with operators and we're promoting the osmocom Solution for its stability We create precedent in other places And there's certainly I think the incentives are there like telecom is It's huge in terms of its potential for You know operating a business So if we can put that those tools in the hands of people You know and make it simple enough for them to run That will Have a huge impact in getting more people connected But this means then the revenue stream for the people who are developing the code would be by supporting the operators or How could that work at the end well the people So there's there's need you know need needs the mother of invention. So When people are actually able to deploy their networks and they have a need that is gonna differ from the community That's right next door. So You're gonna see this is free open software. You're gonna see Organically people contributing back Okay, to the stack and they're doing it because they're Financially driven to do that. Okay their needs. All right All right, and rock if you're if you're in pressure we have to stop here, but maybe I Think rock you have you have a working business model how you work. So Tell me a little bit. Well, I think actually we almost have all the same meaning I Facebook is going to deploy base stations in order to run a communication service We deploy base stations because we run a communication service So I think the way meaning this business is going to continue to exist It's because you have people operating those networks That have a need for support and those are going to contribute to by paying back the BTS vendor or By going back directly to funding or small come If if if they have the level Yeah, the technical level in order to To do that directly Okay, thank you Your statement and maybe a final word also and no statement. No statement. No statement I have an important announcement or question though, but maybe if you want to say something to close up and then oh The question about how many people are going to the restaurant. No. Yeah. Yeah I had to ask also I can ask me But you're going to do this first let me even if it Self-raising stings because I'm from the same company how thanks a lot for Doing this the whole day. I learned a lot and so I just hand over to you for the final words. Thank you Well the final words Yeah Definitely it I think it was a success to have this day I'm not quite sure whether the talks were too introductory for some of the audience I had sort of the feeling that maybe we should have been a bit more technical But I mean that's feedback you can provide basically offline until the next year when we try to do this again The reason that we started with a single day. I know of course It's sort of stupid to travel Intercontinental for a single day event The reason why we did it for a single day was basically because we had to try it somehow and you know Renting a venue and preparing and speakers and so on and so on for multiple days We have no idea if anyone ever shows up is sort of Risky, so we thought well, okay Let's do it for one day this year and actually to be honest until a couple of weeks ago It was looking like we have to cancel the event because nobody really registered Until like three weeks ago or so and that's of course doesn't give us any real planning horizon and Also in terms of capacity may initially was like oh my god, you know We have like seven people in a room for 60 or room for 50 and now we end up having more than 60 people and having to Tell people no you cannot come because the room is full So what I would like to see is a little bit more Like planning when registering for conferences and events I mean I attend a lot of technical conferences and I don't think I've ever registered on any of them without a couple of months of Advanced ticket booking because it's not even available the ticket on on a short notice in many cases. So Yeah, whatever Yeah, make the conference expensive well, that's something I didn't want to do but anyway So that's just some explanation to to some of the comments that we have been seeing so definitely I think with the basis of this year we can do something that's more than one day Next time we do an Osmo conference and Okay, yeah, again, I had my thanks in the beginning this morning, so I'm not going to repeat myself Thanks to everyone But now we have one final important question who here will be attending the or yeah Who will be attending the social event? Let me just take a quick head count is Steve yeah, can you please count?