 And let's put it together. So hello, everyone. We'd like to welcome you to the EuroPython 2016 session. And we're going to do the talk together. I'm going to do the first few slides. And then later on, Pavi is going to join in. And of course, if you have questions, just feel free to ask during the presentation. We don't want to put everything at the end, because then we used to lose the context. First of all, a question. How many of you know what the EPS is? How many don't know? OK. So the EPS was founded in 2004 when we found that organizing EuroPython. EuroPython started in 2002. And by 2004, we wanted to have an organization that basically holds the IP rights of the conference and makes sure that it's a continuous process, that the selection process works for each new location, that the transfer of knowledge works from one location to the next. And the EuroPython Society was supposed to back up the local organizers at the time. So the way that it worked was that the EuroPython Society had the trademarks, the logo. And it was supposed to have the social accounts for the conference. The EPS used to select, just select the conference location. It did not actually run the conference itself. That was being done by local organizers. But every now and then, when we switched, we had a CFP process to then determine a new location and then work with the local organizers to actually make it happen. In 2015, we realized that the old model of doing this would no longer work out. Because we had serious problems finding new local organizers doing the conference because it had grown this big. I mean, this year, for example, we have 1,100 attendees. Last year, we had 1,200 attendees. So that's a size where you cannot easily just do the conference organization like that on the side. Here, it's actually a full-time job for quite a few organizers. And so we thought that we'd use a workgroup model to make it possible to scale this up in a better way. And the workgroups are meant to allow work to be done remotely. So there are lots of things that you can do remotely. For example, you can manage the website remotely. You can do communications remotely. You can do marketing and these things remotely. And this is what we started this year for the first time. And it has not yet really worked out all the way. But I think we're doing fine in developing this kind of model. And we're definitely going to continue using it. One of the most important things there is that we wanted to reduce the loss of institutional knowledge from location to location. Because we always had in the past, we always had an issue having the knowledge that was gained at one location being transferred to the next location. Usually what happened is that we got a complete new website software set up. So EuroPython has, I think, about six or seven different website systems right by now. All the little details that you find when organizing a conference, those details were usually not transferred to the next year. So the same mistakes happened over and over again. And of course, there's a financial risk associated with this. And we wanted to reduce that risk and have the EPS sign contracts instead of having a local organizer sign the contracts. Now, this year, we tried that. It didn't work out because we couldn't manage the VAT taxes with the EPS. We tried to get a VAT ID for this, but the Swedish government wouldn't give us one. So we have to deal with that next year. So under the new structure, the local team is meant to just do the on-site work. And we try to do most of the other things remotely. Now, in practice, for this year, it hasn't really worked out that well. So what happened is that even though we had the work groups, most of the active members in those work groups were actually members from the on-site team, plus a few members from the EPS board. But we just started this year, so we have good hopes for the next years. Right, so just a few dates, maybe, to show you how the whole thing developed. In July last year, we had the election of the new board. Then in October, we did the CFP for EuroPython rather late. By December, we announced the winner, the ACPySS, the Python, Sans, Sebastian user group. And in January, mid-January, actually, we started to work with the ACPySS to run the conference. So the conference itself, the whole conference, was set up in a very, very short time, as you can see. In January 2015, we also, because we started with the work groups, we set up all the infrastructure for the work groups. So we're using a meeting list for that. We're using Lumio for voting. We're using Trello for a few work groups to manage various tasks. We set up a pre-launch website because people were starting to ask where the website was. And we set up this pre-launch website because it took a bit longer to actually launch the final version of the website. Then in March, we managed to get the final website running. The website was using the code base that was used in Florence because that was very complete conference software. And as you can see this year, again, there are lots and lots of cool features in that software. And once we managed to get it working, it really worked very nicely. And so we didn't have to do anything much to make it work like what you see now. So then in March, we also launched ticket sales. And then we ran the call for proposals, again rather late. Then in April, we had this podcast issue, which was actually the first bigger COC issue that we had. In May, we had the FinAID program set up and announced and the talk voting was started. Talk voting was not done last year, it was done in Florence and it was also invented in Florence. It was a very good kind of, it turned out to be a very good way to figure out which talks, talk proposals to accept. It works kind of like crowdsourcing effort. So basically all the people that have bought tickets can then go to the website and vote on the different talk proposals. And then the program work group would then take those results and based on those results would then go ahead and actually do the selection. So the selection was mainly done based on the results from the talk voting. But the program work group, of course, also took care to, for example, increase the diversity of the speakers, of the number of talks that were done by women. It also took care that each speaker only, well, we try to have each speaker only do one talk so that we can get more and more speakers instead of having one speaker do five talks and the other just maybe one. So I think I'm one of the notable exceptions here because I did one talk for your Python in the Python context and I'm now together with Fabio doing two other talks, just the EPS stuff. But this is more on the side of the conference, it's not really the normal case. Right, and then in June we had the schedule online. It took a bit to get that set up, again, because we had a few issues with the website. But we finally figured out how to actually publish the website, the schedule on the website, and then in July we followed up with the Guidebook application and now we have EuroPython and today we have the General Assembly. So overall it took just six months to organize this conference, which I find really interesting. And we really have to thank the local team because without them this wouldn't have happened. So I think we should give them a good applause. So thank you for that. Right, and just to give you an impression of how the development was with EuroPython, it started very small, and with 240 people in Charleroi in Belgium. And then every year the attendee size increased a bit. When we went to Vilnius in Lithuania it went down again a bit because it's a bit far away for many people, but it was still very interesting in Vilnius. Then in Birmingham it picked up again and last year we had with 1200 we had the peak so far. In Bilbao it decreased by about 100, which is not really much. And given that Berlin is right at the center of everything and it's very easy to get there and there are very good train connections and everything so it's not surprising that Bilbao has decreased a bit. We hope to increase that again a bit by maybe a few hundred next year. We'll have to see how that happens. I think that when people go home again and they know how well organized this conference was and how beautiful the location is and beautiful Bilbao is, I think they will definitely come back and get us more attendees next year, which is good because more attendees of course also means more work, but it also means more sponsors and sponsorship is very important for the EPS because in the long run what we would like to be is something like a European kind of version of the Python Software Foundation so that we try to use the conference to make money and then to redistribute that money to the European Python community so what we ideally in the long run would like to have is a grants program where you can go to the EPS and ask for grants for running smaller conferences or running projects in Python and so this is our long term goal for this and for that we need to increase the sponsorship from what we currently have. At the moment for this conference for example the sponsorship revenue is about 175 or 180,000 euros, which is really not much, I mean if you look at Python US, Python US has way over a million dollars in sponsorship money every year, so there's quite a big, they have about well between 2000, 3000, I think it's about 2500 at the moment. So I mean compared to Europe Python, Europe Python has about a thousand, but I mean it's just quite double the size and but the sponsorship money there is, well I mean in the US it's much easier to get sponsors because they're very open to sponsoring things and in Europe it's not that easy so we have to try to talk to all the big sponsors that you have at Python US for example, whether they would like to do the same kind of thing in Europe, you can see that with Google for example it has worked out great thanks to Fabio's good connections to the Google sponsor contact, they really invested a lot of money this year, we're trying to do the same with Intel, we're trying to do the same with Microsoft and with Facebook so we need to get better connections to them, build better relationships with them and then make everything bigger. So now I'm gonna pass over to Fabio, he's gonna talk about the work groups. Okay. We have a slide. We have a slide for that, yes. I think it's on the general assembly, right? Yeah, maybe it's not on this one, let's see. One thing, a couple of things I wanted to mention, as sorry, I know Bilbao is in the center of the world but Berlin is more connected, that was the statement. And the other thing was I wanted to mention is that I had many people asking about around regarding the website and how they could, that it was hard to help and how we chose, how come we chose the Italian version instead of the last year version and the background is that we contacted both the main developers of the Italian and the German versions and based on the support because we didn't have much resources. Both were very, they said to be very honored for the request but the Italian team turned out to be ready to do more work. So it's to say that both software work great software as well. One, so the work groups, the work groups concept we had in mind is that we had one group, work group chair person or two, depending on the work group size, working members that would focus on specific topics on specific things of the conference that could work remotely and we could organize those work groups with voting members and a few other working no voting members. The teams were, sorry, were built looking to set to tackle the conference administration so having to run contracts with the venue, tickets, support one finance group to control our budget, do accountings, billings, check out the connect with other work groups that needed something from the budget. Sponsors, so for instance, sponsors and finance work very connected work groups. Sponsors of course were in charge to contact new sponsors and all sponsors logistics, explore sponsor's needs and everything, communications, do press release community relations, work on the diversity and outreach, code of conduct, follow up with meeting lists, the support, help desk, sorry, contact attendees, help with business, financial aid, handle the grant selection, set up grants, aid, organize aid, put up information marketing, do their jobs, the program that Mark, Mark and I already told about, which is repeated, the web work group to handle basically the website and everything related, the media team to do video recording and translations and of course the most important team and the on-site team to handle as a glue for those work groups and work locally on all the providers, the venue and all the needs regarding catering, printing, logistics and everything. We thought since the beginning this was quite important to have guidelines to get communications right, avoid putting people in a place where they couldn't actually understand where they could help. The guidelines are necessary for those groups also because to welcome new people and address how they could help to keep the knowledge and hopefully act as a code base to improve and then near the new additions. The purpose is that the guidelines evolve pretty much in line with the conference's additions so every year we learn something new, every year we have new members, new ideas and we really think that those guidelines should just be written by the volunteers and the people doing the conference and hopefully from the feedback from the people enjoying the conference as well. We did a lot of, we had a lot of content to Wiki and a new Wiki with the content. We set up a few procedures and documents to handle all the communication, the conference workflows and needs. So at the end of the day, this year was pretty much a dense year. Hard year but an overall I think this year was really a big step forward on building something that can stand for the future additions. We built a lot of things. Many things may not be great on the workflow side of things but we really, I think this is a real proof that we can handle this. If we could, especially the team from Yubao could organize a conference like this in six months with a lot of attention, a lot of small time and many unplanned surprises like for instance starting with a number of volunteers and ending up with a fraction of that number pretty soon. That proves that we really can do incredible stuff if we have time we can do even better. So I guess that's that. Those are Yubao Society contacts for information you can check out the website. We have a blog and every request, there are many channels you can contact the Yubao Society even through the board or the mailing list or other channels. So I invite everyone that wants to help to step ahead, help propose ideas and keep helping us doing what we do. That's all. Thank you. Anybody got questions? So thank you very much for organizing it. It's an amazing conference. I have just two things I wanted to raise. So talk voting, so it was really amazing compared to some centralized commission which was picking up a few years ago. The only thing is I think that everyone should have a limited amount of votes that you can give because otherwise you have a lot of talks that had very catchy titles. They were not fulfilling the promise and I think that the catchy title itself because of the freedom of voting for as many talks as you want I think that it biased those catchy titles towards actually being selected soon. Perhaps Alex wants to say something about that because he is one of the, well, one of the two Alexes anyway. It's just my only thing is like whenever you are going with the model of crowdfunding or crowd voting, it's like you always seem to introduce the concept of some kind of currency. Otherwise if it's unlimited, then you introduce a lot of biases and it comes from the fact that because everyone has a limited time here. So basically it makes sense that everyone can vote as many talks as you can attend basically a 20 or something and that's it. So basically I can answer that directly from the Voted Program World Group. Thanks for the import. Basically I have some similar results because but on talkwalking we spend a lot of time on the algorithm comparing and it was not just like a number of votes. It was one more complicated thing. It is an Oswald algorithm. It's more like preferences. So it doesn't really matter if you voted for many talks because we didn't not only sum up the points. We did comparison averages and said and we finally used the algorithm with also Italian brought to us. We changed it a bit. So there was no possibility of down voting. But there are some things I think we can improve on the voting process because it's for me it's a bit limited only to people having a ticket. I want to suggest that we include people probably having attended one Euro Python because basically we want to build a great conference for attendees but we don't want spamming with people making multiple counts. So basically yeah. What was good last year is that there was like some kind of feedback you could give. You could ask questions to the people creating the talk. So this year there was no... So my only issue here is like it's actually amazing that you decided to go this path rather than actually making the whole voting closed. But I'm saying like if we really want to achieve the transparency of the selection, either we go with the crowd voting and then like crowd voting that is fully transparent or like some hybrid system where you say like you can vote but actually you don't vote because after that there will be some algorithm that would do something with whatever you did actually. Whatever you vote for. It's all available on the website and it's just nothing really hidden. The way that it worked was that we basically used the system, looked at all the voting results and then the voting results were then used by the program work group to then make the actual selection. So what for example the program group did was instead of having 10 talks about maybe I don't know I think IO, which is very popular this year, they just selected a few of those and then instead gave the preference to other subjects which were a bit underrepresented in the talk voting results. So the voting results were used as basis for the selection but I think it's already a large step forward because you don't have this kind of committee decision process for the talks. We sometimes get biases from those committees into the talk selection. So that happens sometimes with conferences. Yes, that's really good but if everyone is voting and it's like a decision of everyone to have certain kind of talks and I think that's a community decision. Like the community decided they wanna have 10 async. I just wanted to make one comment that takes account also like why the talk voting was introduced in 2011 and there was a progression on one side imagine handling what 300 talks that you need to review, look at it and it turns out that public voting if you see the people voting, I don't think anyone vote all the talks. I can't remember anyone that actually voted every talk. It's, you need to spend the whole day clicking, clicking. I totally agree. It's a very difficult thing and it's a very interesting topic. We picked what seemed to be the better compromise. That's actually that there are many theories about how we should do this. It's a very interesting discussion. So although the source is open source, we could try and maybe put some link regarding the algorithms we have used. If you care about that and you want to improve it, I would really invite you to make suggestions or try to improve that. We are open to improve next year. Oh yeah, definitely. I'm saying this from the perspective of a person that is willing to contribute to that. No, no, sure, sure, sure. It's not like really a completely clean discussion. It's really because it's a hard thing to do and the thing that we chose to not do everything voted is because to avoid some sort of bugs in the attendees preferences. Like we care a lot about diversity. We care a lot about giving the opportunity to everyone to talk. So people are biased by their friends, by people that they already know. So it would be hard for someone like young that don't have enough experience, that don't have enough visibility to get into talking just because someone that is really known have 10 talks, you know? So it's a hard thing to balance. It's really hard. And just to close, this year that the program work group was fairly limited. Like you can understand by people laughing, okay? So the program work group is really like more or less in the front row. So, no, no, there are more, there are more. It's just to say that we had a lot of, yes, yes, sorry, I don't know. But I can count on two hands. And we made a lot of requests for people to contribute. So really, if you care or other people care about doing this, we are. I just wanted to add one thing because we have to switch to the general assembly now. If you want to sign up for one of these work groups and you're interested in helping out and improving your Python, how it's organized, then please submit an application request for that. You go to our website, where's my mouse? There it is. You go to the website, you go to your Python work groups and this lists all the work groups again and at the very bottom, you, where is it? There you go. At the very bottom, you see this, the mailing address here, the board mailing address and then you just send an email there, say, okay, I would like to work in this and that work group and then basically we set you up for it. No, no, you don't have to re-sign up, no, no, no. The whole, the idea behind the work groups is that they persist from year to year and you can stay as long as you want to. If you don't want to be on the work group anymore, you can just tell us and then we just remove you from the work group again. So it's a very kind of open process. We were not making any big restrictions there. The only thing that we sometimes do is we ask the chairs of those work groups whether they need more help in a specific area because sometimes people sign up for multiple work groups and then we sign them up for those work groups that need more help rather than work groups that are very popular. For example, the program work group was very popular so we had lots and lots of members in it. Unfortunately not all these members were really active so in the end, even though we had lots of members in the work group, the actual work was only done by very few people and we've seen similar things in other work groups but for the program work group it was really an exception. So we're trying to change that and if you have suggestions on how to change that that would be very welcome because we don't really know how to address this. One suggestion that was made was for example to assign tasks to people and do that specifically so tell someone to do a certain task and come back maybe in two weeks and with that finished task to do it that way so that people feel addressed and don't just see something happening on the mailing list and then thinking that well someone's gonna do it anyway so I don't have to look into this. That may be one way to try to improve this. We'll have to experiment with that a bit. If you have other suggestions please tell us because this is where all volunteers we're all working together and we don't want to put too much pressure on people but on the other hand we also don't want to have a few people burn out because they have to do too much work. There was one question back there, yeah? Over here, yeah, Mashi. I wanted to, first of all, all of you have to know that I was involved in the program workgroup. I saw people corresponding in email at two or three in the morning, right? They worked very, very hard, okay? All of us, I made a very little contribution but all of you was something huge, the work that you did here. For my point about the voting that we were talking now I and I think other people were in some kind of deadlock situation. I wanted to come here, I submitted a proposal when it was the submission of proposals and my university financed my arrival here in condition that my proposal will be accepted but I could not vote if I not pay the ticket but if I don't have the budget I cannot pay the ticket but I don't have the budget because my proposal was not accepted. Then it was a deadlock. The last day I paid from my pocket for the ticket and then voted and my proposal was accepted as a poster and then I received the financial aid of the university. Then we have to find some way that people that presented and submitted a proposal but did not pay the ticket will be able to vote for all the proposals, right? And then when they have the money they will be able to pay for the ticket. It's a complicated situation, I know. It's not something so simple but we have to think what is the best solution for this deadlock situation. We'll have to experiment with that a bit. We actually, we simply took the code that was there and just switched it on and basically let the web code take care of this and I'm not really sure. I did one experiment and I found that you don't actually have to buy a ticket to do the talk voting but maybe that was just because my user account is switched on as a staff account. I don't know, could be. So we definitely have to look into that for the future. I actually didn't want to refer to that but I'll quickly add to that that other people who ask for financial aid and won't know for a while if they get it and can come here would be cool for them to also be able to vote. But what I actually wanted to say is about the work groups because I just said a look and the program work group is indeed huge online. Wouldn't it make sense to maybe kick people out again that do not contribute because otherwise the groups will increase in numbers. So number of members and it doesn't make sense to have like a huge list of people to refer to that don't do the work and it's just, I don't know, a handful of people who actually do the work. Yeah, that's a very good suggestion. I think we should probably have some kind of policy for this. Maybe if a member is not active for a few months then we maybe just then remove that member from the work group because otherwise it wouldn't make sense, like you say. Yeah, Jacob. And that's the last question because we have to start with you. A last comment about as we are having about a lot of ideas and new things. Actually another thing that we would like to is to be more, try to be more open and transparent during the process to also let people know more about how much we are spending or on financial aid, how much we spend because financial aid is a very difficult topic because it's hard to say how much we'll be able to spend. Many times we did financial aid and helped with a red budget situation. Risking a lot of things just because. No, no, no, no, no, no. I wanted to add this to the discussion saying that as much as the community understand the difficulties of doing this, the more maybe people can make suggestions or ideas about how we can tackle this problem. I would like to look a little beyond 2016. I think it's really, really important for Europe Python to have a vision about where we want to go. I'm actually somewhat involved in how Europe Python got to where it is today because I was the founder of the Europe Python Society and I influenced the decision to go to CERN, to Vilnius, to Britain and to Italy. And we had an idea of why we were doing this each time. We went to Vilnius because we want to spread Python in Eastern Europe. We need a vision in what we're doing and it's not enough just to think about what are we going to do next year. We have to think about what direction do we want to take. Something I see as a problem right now is that I find many of the talks rather boring. There are no really new, inspirational talks at Europe Python this year. And I think that's something we need to address. Yeah, I agree. I mean, we need to do this as well, but yes. Sorry? It's one opinion. No, no, it's an opinion. I fully agree, I fully agree. But I also think that this vision should be shared with more people. This year was a turnkey year and I think that we need to build something that is more than a group of people and is a more engaged community that drives those decisions, drives those decisions. It's a hard problem. It's easy to just take for granted that it's going to be a new Europe Python next year. And I would like that it would be like, not obvious, but this vision is much more open. And we have a lot of people, vibrant people from Eastern Europe. That means that we would like to have more teams proposing from Eastern Europe. We would like to have more workgroup members from that part, that region of Europe. So that's part of the plan that we are trying to bring. And I think we are running out of time. Okay, we need to start with the General Assembly.