 Okay, let's get started. So, my name is Evan Rishu Lip, and like most of you, I love Python. And so I have been trying for the last couple of years to change the culture of my institution, which is CUNY, to be more inclusive for free software and open development and Python in particular. And one of the battles I often fight is that a lot of computer science professors will say language doesn't matter. And I think that language actually matters a lot. And I think that in this crowd, that's not a big stretch. And I think that even more than language or ecosystem, can everybody hear me? I have a pretty loud voice. More than ecosystem or language, the thing that really matters is culture. And the thing that I love about Python is the culture. And so I assume that most of you know more about the culture of Python than I do. So let me tell you a little bit about the university of New York. So the city university of New York is the third-largest university in the system. It has 270,000 degrees, so you can see the spread over 24 campuses. And the mission of CUNY, which is up behind it, is to provide an education for all of New York. And this is legislative language, and it's a little outdated from a time when we only thought that there were two sexes. But I think it's worth saying that the city university is a vital importance of vehicle from upward mobility and disadvantage in the city of New York. The pioneering efforts of CUNY College discovering programs must not be diminished as a result of various state financial responsibilities. So part of the mission statement of the city of New York is to be a place for immigrants, for the children of a working class to get a first-class education. And for me, you can't get a first-class education in complication if you're not using three software tools. And I believe that the way they introduce students in this really good place. So you may be aware that New York is a state-new two systems, the state system and the city system. And again, there's a specific legislative reason why a city system is created. And it is the cost of affirmative action and a positive desire that the city university personnel reflect the diversity in its communities, which implies the people in the city of the state of New York. So in Queens, you may not be aware that there are over 135 different languages spoken. And on my campus, CUNY College, 90 of those are spoken. So we have a tremendous amount of geographic diversity, cultural diversity that we are, you know, we benefit from. Okay, but CUNY student demographics are specific and very different than most other demographics. So 47% of CUNY undergrad come from the household with less than 25,000 in annual income. And you can see the differences for senior colleges versus community colleges. There are 12 senior colleges. Only 23% of CUNY undergraduates come from a household with more than 17,500 in per capita income, so for a family of four, that would be $70,000. So we're talking about students that are really impressed economically. 78% of CUNY students work to pay living expenses. That's not spend money. That's red food, et cetera. 32% of CUNY undergraduates think this work doesn't affect their academic performance. And I can tell you, as somebody who has fought and has been an administrator at CUNY, that it definitely does affect their outlook and their performance. Um, you know, there's been a push, obviously, to move more and more classes online. For many of our most challenged students that doesn't work because of the limitation of broadband access at home. Um, and 43% of CUNY undergraduates have a parent who's highest degree and is high school or less. So we've done a lot of first-time college students. So there's not a lot of social capital, cultural capital, in and around higher education. Um, only 7% of CUNY undergraduates go to home or the 8% live with other students. So, you know, this is not the dorm life experience. 88% live with family. So the project that I head up is called the Tech Count Pipeline. It's an investment, $10 million industry partnership designed to support the growth of the city's tech sector and deliver quality jobs to new workers. So we take, um, basically graduating seniors in the last semester and give them a part-time internship during the school year or, um, we have another class now that we do a full-time summer internship at a New York City Tech startup. And this is like an extended interview process, right? So these are kids that are pretty trained, but, um, but, uh, they may lack certain cultural competencies for college. And so we give startups a chance to employ them at zero cost. The kids make $15 an hour. That's funded by a grant from the federal government in the city under the Workforce Opportunities and Investment Act. And, um, the companies get their work product and the students get the money and the experience. And hopefully by the end of it, um, nobody's, you know, no business wants to let those kids who've been training for four months walk out the door. Um, we have some preliminary numbers. We had over 35% of our students get offers from their host company before they start working. So we have seen that, um, and we have a bunch of other kids who run the board. So why was Queen's College selected for this? Um, in 2011, the Wall Street Journal reported that Queen's College had more computer science languages than Columbia and N1 combined. Um, we had 450 since that, um, at that time. We now have over 1,100. So people are talking about college shortage. Um, you know, there's 1,100 computer science languages at Queen's. There's another 500 at John Jay. Um, and so, um, so there's a, you know, there are people getting trained in this. Um, the reality though is that 80% of the people who code in New York who actually write code came to New York from somewhere else to write that code. And there's a tremendous, um, a tremendous diversity issue with most of the people writing code, uh, tend to be white and male or Asian male. Um, and we have a tremendous amount of diversity. The student out, some data that I can't share, share with you, what I've seen is that our, um, our students are making roughly 50% less than the average starting salary for softened-up engineers. And it's not because they're getting jobs in the, getting those jobs in the New England less. It's that most of our computer science majors after graduation are not getting jobs where they're writing code. And for those of you who do write code, you know that if I'm taking kid after, after graduating from college and I put them in a job for three or four years where they're not writing code, you may as well go back to, to, to step zero. Right? So, so this was some of the data we had. We asked them about, and none of this specifically has to do with Python. It's more, um, you know, this is more, um, more front-end development stuff and terminal, get and, and, and, and, and collaborative software development. But they all found the training course in this stuff to be more, it's useful. They're not getting this in their computer science education. Um, and so, uh, we also asked them what they thought should be part of the, um, part of the curriculum. And, you know, 68% said that having more free software was going to ruin this. Um, we saw some, okay, sorry. So we saw some, some progress between cohort one and cohort two. Um, where, um, we saw, uh, the number of people with a GitHub repo increase from, uh, of our app, our applicants went from about, uh, you know, slightly more than doubled between cohort one and cohort two. And the percentage of people with a GitHub repo actually also more than doubled. So what's, what's the problem? It sounds like we're solving all of this, right? You know, we've got this program. We're gonna, we'll expand the program. We'll get more money from the feds, right? Um, it's all happening too late to change student trajectories, particularly student trajectories for research. And one of the other things that, that, that we found in surveying the students is that the students didn't feel particularly well served by the research staff. So, um, this is the percentage of undergraduate seniors in computer science who had ever worked on an open research problem. And I just think this is appalling, right? That 80 percent of the students had never seen an open research problem. And, and so if we're, if we're continuing to hire research professors, so the, the chair of the department is extremely proud that of the last five hires that the, um, department has made, all five have won NS, NSF young investigator awards and then left the CUNY system for another bet, more prestigious R01 school. But they're clearly not involving the students who, who, who the taxpayers are paying them to serve. Um, if, if we're not getting them, if we're not getting the kids involved in research. And I feel like Python, Python has a role to play in this. One of the other things we asked them, we asked our, our, our residents was did they feel like the full-time professors taught you more than the adjuncts? So this would be people who were active in research. And so, you know, most of them thought that it was about the same. But to the extent that anybody had an opinion, they were twice as likely to think that adjuncts taught them more than the, than the research staff. So for me, I have a, I know what I'm, what I'm, what I'm at CUNY to do, but you know, people have different things. I want to amplify the voices of the marginalized. I want to use technology to create social and economic mobility. I want to give political empowerment tools to the formerly incarcerated. And I want to provide tools to run my course and give students agency over their work. And I think it's critically important to introduce students to free software and free software concepts in order to make these things happen. So a lot of academics have, have a, have a question, will this kill my career? Right? So I say publish everything, publish methods, publish results, publish collaboration challenges. There's even a place to publish teaching fails. And so to prove it, I will show, I will put up a screen of something called the Journal of Interactive Technology and Pedagogy with its teaching fails as a, and this is in, you know, all of the academic databases. It's not the highest rated journal, but at least, you know, you can get it out there. You know, a lot of people wonder if, if, if you're not doing, if you're not doing deep mathematical research, will people think I'm a phony? And I always say it's more important what you think about yourself and the work that you're doing. And so, and finally, will you find a community? And to me, free, there's nothing more that free software is about than the community in which you live. So, but free software is not one single community. It's in fact many communities. So, find one that fits you. Everybody know, or not everybody, but many people know that the Linux kernel is incredibly harsh, right? And the reason, frankly, why I prefer R, prefer Python to R, is that R is one of those harsh, terse communities where traditionally has been where the struggling student is as likely to get the message RTFM as anything else. And RTFM doesn't move you to the next place. It doesn't progress. And I prefer Python to Ruby because we have a larger ecosystem with many, many more libraries that do what academics and students want to do, and we have many more academics involved in the community. So, some practical first steps. Obviously, diversity in Python is a great resource. The Anaconda distribution, so, like many of you, I don't run Windows. I haven't run Windows in years. I don't think that there's an appropriate place in an academic institution for that corporate malware except as a demonstration of, you know, for security. But this is what our students run. And actually getting them installed and up with a development environment is very difficult because of all of the, because of unfamiliarity that most people who are developing in Python have with Windows and because of all of the idiosyncrasies that that implementation has. So continuum, the Anaconda distribution can avoid a lot of those problems for you because students, when they're admitting that they're first interested in computation and a lot of them aren't coming from computer science, they'll be coming from physics or math or chemistry or music or English or linguistics or a thousand other academic disciplines, they want to see that they're making some progress within a reasonable amount of time. And if you can get them up and running on the Anaconda distribution, you'll avoid a lot of problems and get them to a point where they're maybe doing a topic model or creating some graphs without, you know, spending three or four days debugging why NumPy isn't compiling on Windows. As far as editors go, my advice is you don't start on a Strativarius, right? You don't start playing violin on a Strativarius, you don't start piano on a Steinway, just get them going with sublime text or text Wrangler or Nano, but know that you want to move towards VI or Emacs for them long term. The other thing is everybody gets stuck, you know, the resources to point students to, stack overflow, teach them how to copy and paste their own error messages, get students on to IRC. I think, you know, everybody talks about the decline of IRC over the last five or six years, that's not true for FreeNode. So there are active channels for Python, for Django, for Flask, and you can get a lot of student questions answered quickly and it puts them into the community that will likely get them to stay. If you are, we are blessed to be in this city. There is a Python meet-up five nights a week in New York. I mean, it's unbelievable. So, you know, a lot of the people that run those meet-ups are running this conference. You know, Aaron runs Learn Python NYC, which is the first place to send a kid on Saturday or Sunday. If they're not even sure they want to learn Python, but they think there's a chance it might be useful, you can send them there. But many people don't have that luxury and they have to learn by, you know, sort of knocking through tutorials online. But if you're in New York, make sure you take advantage of the resources here. So what we did at Queens, middle of the semester in the spring, I started my job at Queens in October. We had our first cohort class to over winter break. So February came and I said, well, I don't have to do another cohort for a few months. Let's start a Python users group. And, you know, college students are famously over-scheduled everywhere. It's particularly bad within the city system where kids view college as a place to come for their class and leave. Seventy percent of CUNY students participate in zero after-school activities. Zero clubs. They're not part of the culture. And obviously, you know, free software requires a tremendous amount of training, a tremendous amount of involvement. But the colleges are starting to address this with free hour, which is an hour where professors are not supposed to schedule classes, tests, or recitations, just supposed to be clubs. So we put up a club and said, well, will anyone come? And we got people, we got students, we got students there. And what, so now once, okay, you have a Python users group, what are you going to teach kids? And so for me, I was not interested in teaching them sort of the practical web development skills that I needed to teach the seniors as they walked out the door and were interviewing at all of your companies. I wanted to teach them about research tools. And there are some great but not well-distributed resources on the internet that I want to make you aware of. So there's a systems biology course at Caltech that has about 25 Python tutorials in it. You know, I mean, it's not, none of these are 100% Python. You know, it's Python and late tech and, you know, a little bit. But it, but these are the kind of courses that students can run through, kind of tutorials that they run through, where I can then go to a professor who's got a grant and say, this student would be useful, right? They can do some basic graphing. They know what a simulation is. They can, they can write something up in late tech. They're useful. So there's also a research tools at the University of New Hampshire that was done by a Google staff scientist that's very, very good. And it goes through a series of tutorials on Emacs and late tech and Python and some are, and for students that are more interested in entrepreneurship. Okay. For students that are more interested in entrepreneurship, take a look at the Stanford startup engineering. Encourage your kids to make contributions. Keep pushing. If your colleagues are using Google Docs, show them etherpad, set up your own Git servers. Tell the professors you know to avoid the learning management systems and build your own blogs and classroom websites. Thank you. This is my information. All the links and the talk will be up on my GitHub by this afternoon. So thank you so much.