 Okay, welcome. This is exactly the number of attendees in the audience from my first talk I had about Debian Mids. So I'm used to this because the number of people who are interested in medicine and free software is very small. It's a very small intersection. So I'm used to this, but so I will introduce you to maybe some other people on the screen. What is the idea behind Debian Mids? Debian Mids was triggered in Bordeaux. It's DEBCON 1 in France. And I have seen that the French guy was starting to translate the installation instructions of some medical program. And I think translating installation instructions is stupid because you just install a package and that's done. And so I decided to have a talk what would be my idea how medical software should be installed. This was basically the idea of Debian Mids and it was over some wine. I prepared this talk and finally I did the first announcement in 2002. So we have now 16 years of Debian Mids. We are doing yearly sprints and you know Gunnar Wolf is running around here and when he's seen that we are doing a sprint in Brittany and then walk in the winter into the water he thought well I thought you are packaging medical software but actually you need medical attention if you are doing this crazy stuff. So yes it's in some way a little bit crazy to work in a field where people are used to pay something and evaluate things that are paid very high and you do it for free. So this is the reason why we are also such a few people here. But the main thing in Debian Mids is anyway about biology, microbiology and life sciences because there is some educational background, some university background where free software is established in medical care. Who is the Debian Mids team? I did these graphs over several years and it was started in 2009 also I started these graphs because I wanted to find an answer who is in the team because it's not really clear. We have a lot of people in the CISA maybe 40 or 50 people but they are not active or so I decided to make a graph of the 10 active uploaders. The interesting thing is not that I'm the most active uploader, the interesting thing is that we have not only one uploader we have other teams which is not so then settled we have even more than 10 active uploaders but these ones are the active ones and I think what you can see here is that since 2011 it became somehow more and watch this 2011 here also in other years it is this is graphing the discussion on the Debian Mids mailing list which is also a user mailing list so there are users who are asking for software and here also you see in 2011 something happened also here right and this is a Debian Mids developer mailing list it's not so clear this year 2011 but here you see also something these are the people who are hunting bugs you see also 10 people who are quite active there are some people who left the project for instance Charles Plessy who was in the beginning very active the yellow one became less active it depends from his work and what time he can spend on Debian Mids and this is what people are committing to the Debian Mids Git now on CISA you see also there are at least 10 people who are doing quite regular contributions and well now before I come to this graph I want to explain this 2011 since 2011 there were more activity this is the time since we are doing yearly sprints since these people not actually this 10 people but we are 10 to 15 people who are meeting face to face mostly in the beginning of the year in January or February and this I'm showing this because it's quite important that you meet in person like here at the Debian conference a very small auditorium but it's important that people meet and work together and sit in one room together don't care for any real life stuff just sit behind their computers get some good talk together and that's very healthy for the project the next graph shows how many packages get attention by how many people so I would like to get this graph rather like this says a package is never maintained by only one person but rather by two or three persons and get a graph shaped like this because if only one person is committing to a package this is actually not team maintained it is one person who belongs to a team but it's working alone on the package I need to admit that this graph only shows contribution with more than five commits so maybe there is a single commit from a person that is not calculated because it's too much noise otherwise so five commits is the limit when I said yes this is a real contribution and then this graph looks like this we are doing better than other teams which have quite steeper slope but there are teams in Debian which really have only very few packages which are maintained by only one person actually if I'm honest about the package maintained by two persons it's mostly because other people started the packaging and then I continued because they managed it so this is the graph which shows the so-called tasks in Debian made we have different tasks which is the blue one the number of packages for the biology tasks that means if you are biologist you have a lot of packages I'm sorry for the scale it's about here 600 I don't know it's not shown here on the screen so 600 packages for biologists or if you want to develop biology application you have also 100 packages this kind of biopilot and so by the way I know your background what's your background are you biologists or medical stuff so can you say my words? you are biologists ok so then maybe we talk a little bit later so in principle Debian made something for you which you can use I also have prepared slide for about 10-15 minutes and then we can talk and you can tell your wishes and your expectations just for a short introduction here we have some packages for epidemiology it's more or less statistics some are packages we have a few packages which can be used in hospitals hospital information systems we have something for medical imaging but you see the main focus of Debian made if you just calculate the number of packages is in biology and also here you see well we have some increase here but the 2011 is reflected from here after we meet the curve also is quite steep so meeting of people has a good effect the counter-indication is the meetings are basically done by biologists and in the other fields you don't see this 2011 king so it's really important that you meet well something which I really liked is some very good advice by John Medoc Hall I was in Heidelberg in 2015 as a Debian conference and I told Medoc that after 15 years working on the Debian made project or on the topic of Debian made people in my institute which is actually a medical institute which was my motivation to do all this they started using it and he said what only 15 years you young guys should be more patient I should hurry up since after 15 years I might be dead but you have so much time and this was some advice I think it's kind of a wisdom right well 15 years seems to be much and you are even younger than me but in principle every good thing takes time and it's better than if it's used after 15 years if it's not used at all or so so yeah something for my background I'm a physicist by profession I have no idea about biology no idea about physics no idea about medicine I'm just able to do packaging I have a technical background I can do code in some programming languages but I need the input of the biologists they say this program makes sense this is helpful for me and yeah so I'm not using the software I'm just building the packages so this is another interesting thing I'm very proud about because since 2012 I'm running a so-called project mentoring of the months it means for one month some student who wants to learn packaging can ask any question well there are no real stupid questions but questions the student might consider possibly stupid can be asked and I will answer and help to learn packaging I require this question all in public on the mailing list because other people will probably learn from it and have the same questions can learn from it so this is the number of people who finally made this mentoring of the months you see it's not every month three or four per year and I also recorded if the package was finally uploaded which was not really the case for all of them they stopped at some point in time and not all of the mentees really remained in Debian Mids so which I would love to have a better rate here but anyway we have some people who keep on working on this who keep on contributing and finally if we don't actively approach newcomers then we get nobody but we got some people due to this effort and I'm quite proud of it and they try to continue this so not only because of this effort but also because we try to reach out to people we did the sprints Debian Mids project attracted one Debian developer per year so if you, well, what's Debian Mids inside the Who Will Debian project if you would assume Debian Mids wouldn't exist Debian would be most probably the same because Debian is a very universal system it doesn't matter if it has some medical applications or not on biology application or not maybe the biologists are a little bit happy but because the project exists it attracted some people of this field and I'm, well, kind of preaching every year that more people try to follow the idea we are the, well, the astronomers are now doing a good job and the geographers also doing a good job to follow the similar route and attract more people so if you imagine that we in this very unimportant project inside Debian gathered more than one percent of the Debian developers this is quite a lot since 19 of these developers confirmed that they are only Debian developers because the Debian Mids project exists and we have about, I need to check the statistics how many Debian developers we have so this is quite some good statistics for this kind of outsider project and even 10 of the 19 above extended the activity so they work now in different fields of Debian they came over Debian Mids and now are doing something else and 14 of them keep on to being active inside Debian Mids the conclusions of this mentoring effort is that the time I spent to mentor people is really worth the effort and since I have shown it's not every month that I have a student, the workload is beautiful for me and the good thing is the students are reading current documents and since I have read documents about 20 years ago so when I became a developer I could even learn from the students because they ask me questions and this document is, I can find this and it's interesting that I've learned something and the major advantage is that we are usually teaching upstream developers of some biology code or so and they know the software they have written and we know how to package and so we pull the knowledge and at least 50% of the students have strong connection to upstream or themselves upstream or sitting in the next office or so so then we also are doing some software liberation which means a lot of upstream have no idea about how to find a proper license and so we have some wiki page where we just put a record if we approach some authors please use a free license for your stuff and we have one part of this table has the success where we convince authors one table is showing the current process with links to the mails we have and so this talk is online so you don't need to write down this link and click on this, you can find it I'll show the link later yeah so far what I'm also like to say is I'm very proud that we are a real team and I found this team definition by my work in the David and me team since for me as a team if I wake up in the morning and realize that somebody else has solved the problem I was not able to solve yesterday and I really experienced this it was, I was sitting in the Saturday night, midnight on a problem I wasn't able to solve I wrote an email to the list and when I opened my laptop the next morning somebody said, no I solved the problem that's really cool, this is an experience I wish that everybody could have I really like to share this so if we want to make some notes for today somebody could fire up this if you install Gobi and go to this address then we can do some notes if you have some notes maybe we are really a few I expected an outreach student of mine here but she has probably not yet arrived I don't know what's wrong because she is from biologists from Russia and is not here if not I can put it later on the screen again this is in principle all of my slides here you can find all my talks including this one with the PDF and now I can open this for discussion maybe we put one mic to you and you ask questions and I try to answer we are not many, we don't need to stretch the time into eternity but please take the mic maybe you introduce yourself as a biologist or everybody introduces himself and then we talk a bit do we have another mic? I got it say your name and your profession it is okay? please say your name in the mic what's your expectation why you are sitting here and have you heard about Debian made before or are you just at the conference and say oh there is something that might be interesting for me could you please speak in the mic maybe you can also introduce you for the mic and then you as well and then we chat a little bit yes it's good to know more about and you maybe as well because others might watch the stream yes actually I'm here because I want to learn about how to write a book okay and my major is Artificial Intelligence so I don't know much about much about medical things as I said I personally also don't know so much about medical things but I could imagine that your field of work could also have some influence into biology maybe you can also use artificial intelligence to do some gene sequencing also doing other applications which are doing this in the recent actually in the medical field there are many more deep learning applications in this field such as medical image segmentation which means to find to find something strange in a medical picture such as cancer the location of the cancer but I'm not familiar with that image that field so maybe I'll talk about it in a book that I'm preparing yeah, that would be interesting yeah do we have some responses from IRC? no response so I don't know what if you want to stretch this even more you can always approach me we can do it directly here I'm hopefully not scaring you away because maybe a little bit famous also I just learned in the talk before that I have I'm the most active uploader I have 2,000 also packages before this was also new to me now but it's a lot of stuff because we have in David I made about 1,000 packages or more and probably I have touched all of them at some point in time actually now because of the move from Alio to Salazar everything is to touch and I do it more or less automatically so it's not so it should not be so impressive because it's not so much work once a package exists to update this and so it's kind of a routine but did you work at all under Linux before or just so you can test that is okay just please try talking closer to the microphone maybe we finish this streaming and all you have a question what made you to work on the medicine package specifically because you are a physicist right can you please repeat what made you to work on the medicine package like you're a physicist right well the thing is that while I'm a physicist I'm working on a medical institute it's in the German Robert Koch Institute which is doing research for the German Ministry of Health about different diseases and we are doing we are also this gene sequencing stuff is quite common to find out some issues in diseases and so it's quite natural to do all this biology stuff and they hire IT persons and if you are a physics then you are kind of a definition IT person half of my of the students of my group are working on IT and so I'm just not a physicist anymore actually but IT scientists and well it was my motivation to say if my institute might decide to use Linux then this should be Debian and Debian should be really fit for the job which we need and this is actually the case because most of the stuff which is used in my institute was packaged before they started using it and what was not yet packaged I just did it not completely because something is more complex but it's getting better and better so I'm packaging at request if a scientist approached me I need this or that software then I'm just doing it if I manage with the help of several people because in fact I need help of several people inside Debian I for instance have frequent trouble when packaging with Java software and I'm asking the Debian Java team and it's there is also what might be interesting for you and mailing this Debian mentors explains or helps how you can solve packaging problems and I'm a frequent poster on this list not because I'm answering questions well sometimes I'm asking but I have questions to my fellow Debian developers so even if I'm 20 years a Debian developer and have done a lot of packaging it's also done by the help of random people who are helping and it's also an interesting message I think for you who are newcomers and to get help if you know how to find the right channels this is also some recommendation I usually give in this mentoring of the months I ask you a question on Debian mentors I can't answer it myself this is it's kind of an experience to learn how good this works you write an email and after a few hours you get an answer with this solution really amazing also for me every time if I forget we have now the switch from GCC GCC7 to GCC8 and some package does not compile and I don't understand C++ ANZC is okay but C++ and if you have an error I just ask there if I cannot find the solution by Zimbabwe and I get the answer in a couple of hours and then I upload the package and the fame then goes to me because I have so many package but the fame should also go to the people who helped realizing this upload okay yes maybe we finish here if there are no questions you have a question? I'm Yao actually behind the camera I'm actually interested if your if your packaging efforts is mostly into some compiler some compiler problems that is related to medical yes I usually take the ideas what to package from people's medicine and biology in the next we have about deviant science I'm also working deviant science but not so active deviant science is more general for sure biology is also a science and we have common problems and in deviant science we try to find common problems and I'm also helping there because I can sponsor a package or some packages are direct dependencies and I upload also this and science is a wider umbrella and there are packages which are in any kind connected to science it's the same problem in a wider sense can you please speak to the mic? to talk to the mic yes okay this is a good question our plan was that we want to make deviant fit for hospitals I'll show you, wait a moment I have here the tasks which are part of deviant media I can also show this and here is hospital information systems and this is one kind of a database which is used by the Vista system it has nothing to do with Windows Vista it is some to run a hospital with all the information that is needed and we have at least this database system it was actually the first package which was done and only was possible because I was running this mentoring of the month because it's a little bit complex to packages it is complex to package and so I had to upstream and upstream packages so this database is definitely used in hospitals what you see here is popcorn Dero means nobody is actually using this package but I assume if I'm install a deviant system in a hospital I will not allow that the computer is falling home to deviant so we don't get the information if it's really used so I expect that it's used but I have no channel back that somebody says yeah we are using it so I assume it's used it's probably not used in many hospitals the plan was that we also packaging Vista itself this system and we have the deviant package not yet available but we did some preparation it's incredible complex to do and I have the strategy to at first the low hanging fruits and I need somebody who really really test this I can't throw a package package ready test it in a hospital no this doesn't work so we need some preparation partner and I my idea about deviant made is that we really cooperate with the people that are using it and as long as these people not approaching us can we work together to finally get this done this will not happen hospital is very complex what we have is we have medical practice systems there are this is just small helper free diams specifically in France you can run doctor's practice here at least seven active installations are using this this number is a number of installations which are really used not only installed on the computer and then this computer is not used the files are really touched so these free diams is mostly used in France this ginkgo cat x is something where you can display medical images and we have GNU made GNU made is also I know this GNU made developer you see we have a server which is actively used by 12 installations and also maybe the client has no connection to the internet and it's not reporting so many these numbers are only estimations some people say you can double the number some people say you can take ten times the number of popcorn measures so we know our packages are used by some users what you see definitely we had in the before they had games packages which have the number 1000 so it's way more games are used by more people we have a very specific field but we have users and we have active users and if you look into biology because we are biologists so we can fill the time a little bit with this information we have 500 packages in this field so you can look here maybe for instance this one I don't know how deep you are in this but all these this information here is from the package description it is usually translated I have the German look here you see this is translated this is German this is to be translated and then if you are biologists you could say oh I can translate and yet you click the button and then you can do this for Chinese we can we do the test now because we are here we use Chinese what's Chinese is this this one please help me write this one we check if we have Chinese translations we are looking here not yet it looks all very English no unusual characters for me oh it looks not so good but I can explain more what we have we have for instance you see citations this is also a Dewey and Will team has started because the idea is if we package the code of some authors the authors get some additional some additional Google hits if they are mentioned on our page so they need are it's interesting for them if we package that they help us to package so it's I have I'm very keen on that we have every package has citations and we have also quite new these tools there are some registries for biometric tools and we have also the registry entry so you can click on it so this should be a network of all the applications that Dewey and his kind of entry point for free software in medicine and biology we have also some screenshots but not so much here where we have screenshots we had this medical imaging there are naturally more screenshots I mean screenshots you know this is not so amazing it's just the text interface so if you have some if you are using some applications please make a screenshot and submit it here you can upload the screenshot and then it goes into the Dewey and infrastructures it's not just for this page but it goes into the Dewey and infrastructure is available for all packages so here you have a list you can also see here I have the information that there is some new upstream version available that means for the developer there is something to do I need to do the packaging I could even have 5 minutes left I can even show what work it's done for the packaging yeah maybe I do this then we have 5 minutes left correct? that package flexbar has a new upstream version and what I'm doing then is I have made a script it just calls or I update the Git repository GVP to see if somebody some other Dewey developer has perhaps committed something knows up to date and then I have a script which is called OT update and then I wait this is doing some stuff downloading the new version adapting stuff and if I'm lucky and the new version has no big changes I can upload so doing a new package is not necessarily a lot of work it's now creating a change hood environment where the package will be built and if I need I do not need to refresh some patches and the end of this talk it's ready and I can upload it's not so much work so this amazing number of 2000 packages I uploaded is sometimes just waiting not all the time I can ensure it but it's frequently just like this do we have more questions why this is another one you are from Taiwan Taipei are you also interested in this topic or are you for the next for this Dewey science no the tournament after you so you are interested in Dewey and Mead or Dewey and Science it is possible that you introduce yourself for the mic that mic please maybe it's switch do you see the green light talk later about this so you need to see the package is new Lydian tells me this package has not yet an autobaker game test which is not good because we try to apply test to every package and it also says that I should possibly I usually fix this in detail but since the time is running out no I will do it later I can just tell you what this script did git log made here this was original I checked out I got the new upstream version downloaded which is recorded here I've set the latest standard version and this was the only needed change to unstable and the tech is set I could now push my changes and upload the package this was all so I have not really done a lot of stuff right to get the new version packaged any questions for this just finish here ok thank you for the attendance