 Minä olen Ilma Rilaukas, ja olen Settember, jossa olen tullut TDF-yhteyden perustaminen. Olen tullut uudelleen jälkeen, jossa olen tullut uudelleen jälkeen, jossa olen tullut uudelleen jälkeen, jossa olen tullut uudelleen jälkeen, jossa olen tullut uudelleen, jne. Tänään lopulta esima aina mielenkiintoisemmin näkyy sen erittäin ymmärrä rapuтоista. TDF on lopulta tu каналеa muutaman valtiasta 2018. Se hallitamme, että olemme oppimme sekä inducta다ista. Eli me olemme osaamista kulttuurisuus, jotta kulttuurisuus on kokemme. Me olemme kulttuurisuus, jotta kulttuurisuus on kokemme. Ja myös ottaa kulttuurisuus- ja valintirjärjestelmät, ja ottaa kulttuurisuus- ja valintirjärjestelmät. Ja sitten tulee tuntumisesta. Meillä on kontaktia jäämme enää tehdä työntekijöitä. Seuraavaksi, että kun tuntumisesta on tuntumisesta, voimme tuntumisesta ongelmasta. Se on ollut yksi kokemme, että koko tämä kokemuksesta on tuntumisesta. Jos katsotaan statistiikkiä, me näemme yksi kertaa. Stats show 38 contacts for C++, 89 for Quality Assurance, 33 for User Experience Design, and 8 for Documentation. We also have opportunities for visual design and marketing, but there was no point in including them, because visual design only received two contacts and marketing has never received contacts at all. So only five persons in Quality Assurance and one person in documentation did any actual work get plenty of contacts, but very few of them moved to contribute. When responding to the contacts in the beginning, I was immediately fondling in the dark as I was lacking the experience. The experience and understanding built up quite slowly as I could not really draw conclusions on anything after several months. After witnessing contact, after contact drift away without ever starting to contribute, it became apparent that this platform was not some magical funnel of contributors. We would have to deal with this as a problem to solve rather than as an eternity solution. The burning question is what happens between the act of contacting us and the non-act of never doing any work. The frustrating lack of data from this area meant that we had to operate on theoretical assumptions when trying to improve our approach. Let's consider the background of the contacts for a moment. Most of the people have never used LibreOffice. The ones that have heard of it are a minority. Not everyone mentions their education and word background. There are many professionals and some students. This is different from the usual problem contributors. Personally, I had no experience in quality assurance before I started branding LibreOffice bug reports. So non-professional LibreOffice enthusiasts have the advantage of being familiar with the software and having a strong drive to improve it. On the other hand, professionals are expected to be able to understand any software from the perspective of their expertise. Professionals have the advantage of knowing the tools of their trade and should thus learn our workflows much faster than managers. There is a screenshot of our C++ development opportunity as it appears on the volunteer match website. In principle, anyone browsing our posted opportunities could start to contribute without sending us a message through volunteer match. The opportunity descriptions include the necessary information and links to technical documentation. In practice, there have been maybe one or two cases where someone appeared in the chat without reaching out to volunteer match. This is how I responded to people inquiring about bug testing in the very beginning. I'm happy to hear from you. Have you already created an account at our Godzilla and the link? Do you have experience in quality assurance? If you want, we can chat in real time and then chat information. I typically reply to the contact emails on the same day as I received them, but I think it's pretty important to be quick in this. So here is an evolution of the previous email after some months of tweaking. So it goes like this. Triasing, analyzing newly arrived bug reports is by far the most time consuming activity for us. We get anywhere from 600 to 800 reports per month. Here is our guide for getting started with triasing. It has a quick start guide I made that is a good fit for beginners. After you have tested a bunch of bugs, according to the beginner instructions, you might start studying a more complete guide to triasing and then mention about communication opportunity. Some people did schedule a chat or even a voice call with me. During this period, some designer contacts even joined a design team meeting. It was emotionally taxing to first get excited about the idea of what seemed to be a game-changing increase in contributors and then realize it was just an illusion. And here is the current incarnation of my response to basically any inquiries. This is the beginning of the email. So it goes, our dogs are pretty good and the work can be self-starring but I would still like to interview you. We can talk in text mode over IRC. When would be a good time for you? I am in Helsinki, Finland and available during blah blah blah. Please use this time zone converter to find the most convenient time for you and me. So the background for this is one day in September last year I was discussing the frustrations related to the situation with long time contributor Drew Jensen and the interaction with Drew was really beneficial as it gave me the idea for a better structured approach. I decided that instead of being optional, orienting interviews would be required. If a person never replied to my request for an interview, I would not need to speculate if there is something wrong with the information I provide. At first I offered IRC chat or GC audio calls as the platform for the interview. Over time it became apparent that people really preferred text chat over audio calls so I could simply stop mentioning GC. As most of the contacts live in the United States and I did in Finland scheduling the interviews became interesting. I settled on mentioning my available local assignment referring to an online time zone converter. So the interviewing is a good method to fill gaps in understanding basic concepts and they get used to a crucial tool that we use to chat and I'm able to make sure they know what to expect of the work. Of course convince them that it's a great idea to work with us. It should be obvious that tech documentation cannot replace human interaction. Meeting on location would be optimal but remote mentoring is the only sustainable solution because we cannot have a thousand mentors on standby spread across the major cities of the world. So typically the interview goes like this I'm getting guidance from our skilled experts I tell them about our user base, the contributor community and the events Directed via email or chat to ask for assistance or ask in public in the chat. Some questions they might ask me is how many hours should we expect them to work per week and if they can get a recommendation letter. So I just say the hours are entirely up to them. I promise to sign a recommendation letter but so far no letters have been signed. After starting about 50% of the contacts have agreed to be interviewed so that's kind of a big filter effect there. So it's very rare to get an explanation on how someone did not start working and here are a couple. So our quality assurance reply after an interview I don't think this will work. I thought I had the time and I doubt. A person replied and never interviewed. After reviewing all the info you provided I do not think I am the right fit for this position. So I think the first reason is much more common than would appear. It might explain much of what we see here with the amount of contacts because people are hopeful about their amount of free time. And for the second one I would have loved to get more details but unfortunately did not receive anything more. While preparing this presentation I ran into an article on the volunteer match website that reveals our experience is actually quite normal. This is a bit of a relief as it confirms the optimizing strategy has been the right approach. So they say we have found that many volunteers who use our service make connections with multiple opportunities and organizations. So they like going shopping for what's the best organization. And on average 25% of connections result in successful ongoing volunteer relationships. So what did we get out of this? Our mission contributor revealed two chapters of getting started guide. And three quality assurance contributors did meaningful work checking 30, 80 and 420 reports. So one of the quality assurance contributors has done significant work and wants to get into automation next. And some random interesting things. We got two inquiries from students who were in the USA on an F1 visa. They have completed their studies and needed to find some work. Even volunteer work to continue their stay in the country. Or they will be kicked out. This happens in the context of a thing called optional practical training. Unfortunately TDF does not have any branch office in the United States so we were not able to help them. Well I did try to help them anyway. So volunteer match is not the only game in town. There are other platforms, many of which do not operate internationally. And recently Judenek Sremanek has added Libra office to a check volunteer platform called Omsen Untam. I look forward to working with Judenek and helping with this effort. As recruiting continues to be such a challenge I have started wishing we have science to guide us. It turns out there is a genre of scientific literature about contributors in free and open source software and how to recruit them. The researchers are doing helpful stuff and I hope to somehow work with them in the future. One such researcher is Anne Markholm who is pursuing a PhD at LERA, the Irish Software Research Center at the University of Liberty. This year she released a paper called Why do episodic volunteers stay in free and Libra open source communities? It is based on a survey to test a proposed model of factors associated with keeping contributors in free software projects. So even though the paper was specifically about irregular contributors I found it useful that it considered different motivations in a fine, great way. The paper concludes that social norms, satisfaction and community commitment are all positively associated with an intention to remain. In the results of this particular survey contributor benefits were not related with intention to remain. What are these contributor benefits motives? They include that I want to be recognized for my contributions I want to receive a tangible acknowledgement of my contributions I volunteer to get a reputation in the free software scene I volunteer to improve my job opportunities to make money, to learn and develop new skills. Based on what I have observed it would appear that among the people coming to volunteer match contributor benefit motives are more strongly represented. In an event conversation Barcom said benefit motives might have more of an impact in recruitment as opposed to retention. Thanks.