How I Fought To Graduate Without Using Nonfree Software
by Wojciech Kosior [1]As a university student, I have struggled during the pandemic like everyone else. Many have experienced deaths in their families, or have lost their jobs. While studying informatics at the AGH University of Science and Technology in Kraków, Poland, I have been fighting another, seemingly less important battle, but one I passionately feel is vital to our future freedoms. I describe my fight below, so as to encourage and inspire others.
Unethical platforms
Software freedom is a huge but hidden issue in our time. Digital communications technologies such as videoconferencing have taken center stage in our lives, and for many the use of these has been a savior. They do not notice the danger concealed in the way it works: whoever controls this technology controls our lives. Recently we have seen the power of Big Tech to subvert democracy, control speech, exclude groups, and invade our privacy.
Software freedom is a fight to return control to people. It is a fight against “nonfree” software, also called proprietary software, which imposes unjust and invasive harms on its users. In pursuit of our liberating mission, advocates of software freedom like myself insist on using libre software.
It is especially important to spread these ideals to new generations. Unfortunately, we often see the opposite trend. The default operating system found in most computer classrooms of my country is proprietary Microsoft Windows, with some universities even providing students licenses for it. At some point I came to realize this practice really only benefits the proprietary operating system vendor. Similarly terrifying is the level of dependence of course organization on nonfree Google Sheets and Google Forms.
During the pandemic we saw educational facilities hastily embrace proprietary tools such as Microsoft Teams, Zoom, and WhatsApp, pressured by the network they generate. Schools and universities then tried to impose them on students, who subsequently suffered the loss of freedom from using programs that users don't control, as well as bad security and violations of privacy.
Because I refuse to use unethical software, the complete reliance on proprietary platforms has created an ethical conflict. My aim has been to complete my university degree without surrendering to the imposed nonfree services, by convincing my professors[2] to allow me to use only free-software replacements to proprietary applications. I didn't expect to win a fight against such power, but now, through polite but firm action, I think I may have prevailed. Hopefully this story will help you resist too.
Ethical studying
Over time I've become more and more determined to avoid nonfree software. Among other challenges, that meant getting a Libreboot'ed ThinkPad and switching to GNU/Linux distros that include only libre packages. One might ask:
What about studies? Weren't you required to use Windows? Or MS Office? Or some other proprietary tools?
Actually, a majority of classroom assignments could be completed with free software. Today we have the luxury of excellent libre operating systems and libre tools for most tasks. Most popular programming languages have free software implementations. On those few occasions when some nonfree tool was strictly required, I was able either to convince the professor to let me make a substitution—for example, to complete the exercises with a PostgreSQL database instead of Oracle—or to do the assignment on a university computer in the lab. I admit, running nonfree software on a computer other than one's own doesn't fully solve the ethical problem. It just seemed fair, but it is not something I'm proud of.
I also take the issue of in-browser JavaScript (js) more seriously now. Web js runs in an isolated sandbox, which leads many to believe it's acceptable, even though it's nonfree. Sandboxing might indeed solve security issues, but the true problem with proprietary programs lies elsewhere, in its denial of letting users have control. Currently, maintaining one's software freedom in the field of web browsing is not easy. Sites frequently malfunction when js is disabled. I have had to ask colleagues to help me enter study-related data into a Google Sheet because I couldn't do that without js enabled. In addition, js code is used to implement browser fingerprinting which is used to track users.
Gentle persuasion
Without serious problems, I completed the fifth semester of my studies. At the beginning of the sixth semester, the pandemic began. Universities closed their physical facilities, so most students returned home and professors started organizing remote classes. Unsurprisingly, they all chose proprietary platforms. Cisco WebEx, Microsoft Teams, ClickMeeting, and Skype were popular choices. I could not find a free software client for any of those. Also, not realizing the problem of nonfree js, professors expected everyone to be able to easily join the video sessions using any web interface.
How did I handle these requirements? I would very politely email every single professor who announced something would be done using a problematic platform, explaining the lack of a suitable free software client. I often included a link to a popular online explanation of the issues of software freedom and universities, the “Costumed Heroes” video created by the Free Software Foundation (FSF), along with some other links to free videoconferencing programs like Jami and Jitsi Meet.
Although there are many documented surveillance and security issues on these centralized platforms, I explained that, for me, software freedom was the troubling factor. Replies urging me to “run the program in a virtual machine” or saying that I “don't need the source code to use the service,” made it clear that some of my professors didn't understand, or understood only part of the issues. Had I been studying anything other than informatics, I suspect the fraction of those who understood the problem would be far smaller.
Missing out
There were two distinct areas of concern. The first was with accessing and participating in the teaching materials; for example, in a Machine Learning class I found someone to forward on to me what the professor had said. The second was around registration and assessment. For some remote classes, presence was not checked. I skipped those. Uploading my homework to Moodle also didn't pose any issues.
The first real problem arose with the Artificial Intelligence (AI) course. It was taught by rotation. The first professor gave homework requiring the proprietary Framsticks application, but allowed me to do a neural networks exercise instead. Another professor agreed I could use Webots instead of Choreographe for a simulation exercise. Yet another one asked us to complete an online NVIDIA course that required nonfree js. That professor did not respond to my email.
One Distributed Systems homework was supposed to be submitted via WebEx, but that professor agreed to let me use Jami instead.
Uncertainty and doubt
Another issue that arose was gnawing uncertainty in the absence of a clear policy. Not knowing whether the university would recognize my principles was a cause of ongoing stress. Despite my small early victory, other rotational courses meant that three more professors would each need to agree if I were to pass, so until June I could not be sure I would succeed. In March, System Programming classes started. The professor, who didn't want to lose time connecting to a libre platform to rate my homework, gave me little hope, so again I was to live in uncertainty through the Easter and beyond.
I believe every class should at minimum be offered ways to interoperate with libre tools so that students can at least read class assignments on free platforms, and upload their answers from them. Unless universities offer interoperability, the reliance on proprietary software costs both students and professors time and headaches. At one point I emailed two professors about the use of nonfree platforms for lectures. One didn't respond and the other replied rudely. They seemed not to understand, but I suspect they were avoiding any extra work. This had a corrosive impact on my engagement and I stopped caring about lectures. Avoiding a language-specific package manager that I felt put me at risk of security and freedom issues cost me considerable time and delayed my studies. Time is precious for us all.
Friction over freedom
Although stressful, thus far things had gone fairly smoothly. But after Easter, a Software Engineering course presented the first big problem. This professor first ignored my emails, but eventually wrote a long reply and threatened to fail me if I missed one more meeting. That email's tone showed great annoyance, perhaps anger. It was suggested that I use a colleague's help to participate in the meeting. Another classmate and I connected through Mumble, through which the professor was also intermediated—not perfect, but it worked!
The issues of software freedom, which are ethical, must be separated from other concerns to which open source supporters often give priority. For example, advocates of open source refrain from bringing those important freedom issues to the table and only say that software with source code publicly available is going to achieve higher quality with the help of the community. Meanwhile, our opponents claim proprietary software can bring higher revenue, allowing the hiring of more developers to work on improving it.
The Compilers course exam was to be conducted through Microsoft Teams. Again, sticking to my principles, I thought I would fail. Funnily enough, it was Teams that failed. It could not handle dozens of students connecting, so instead the exam was conducted via email. On the other hand, during contact with my thesis supervisor in July, Jami broke during the meeting. No software is perfect. But with libre software you at least get to keep both pieces when it breaks.
It is not necessarily the functional aspects of the software that creates friction around lack of software freedom. During the summer I had to do an internship. I backed out of a paid offer after learning that the employer would make my code nonfree. I eventually did another, unpaid internship.
So after all my struggles, I finally passed the summer semester and even had decent grades. What at some point seemed almost impossible, was now a reality.
Proprietary imposed at all levels
Before the winter semester, a list of allowed videoconferencing platforms that comply with the data protection law was given to professors. It contained Microsoft Teams, Cisco WebEx, ClickMeeting, and Google Meet. You will surely see the irony here!
One professor agreed to use Jitsi Meet for all his classes and suggested that I ask the student council to recommend it to the Dean, but the council never responded to my emails. High quality software offering better data protection capabilities was deliberately sidelined in favor of commercial nonfree solutions in what seems like a case of corrupt corporate capture of an educational institution. The libre software didn't get approved and the professor kept communicating with other students via WebEx.
Misconceptions
As I mentioned earlier, despite being highly knowledgeable computer scientists and experienced in informatics, many academics demonstrated a generally poor understanding of the politics and ethics around software.
The professor giving the seminar claimed that because a libre platform also runs on someone else's server, “it cannot be safer.” I responded that Jitsi Meet allows for independent instances to be created, which eliminates the need to rely on a single company. I also noted that the lack of libre clients is the main problem with other services. It is a shame that professors at this level, who fully grasp the difference between intermediated encryption and end-to-end encryption, will teach it in their classes but not practice it in their daily profession.
On another occasion, I objected to using a Windows VM for a penetration testing exercise. The professor remarked that one would not be a good penetration tester if restricted to testing only libre servers. I gave up on responding to him, but I think proprietary platforms should be considered insecure by default due to, for instance, possible backdoors they may have.
Having resolve
At some point I had an argument with my supervisor, who gave me an ultimatum that I must use Microsoft Teams. I didn't agree and my supervisor was supposed to inform the Dean about withdrawing from supervising me. Perhaps the Dean didn't read that email? I'm just guessing. Anyway, a few weeks later I even borrowed some electronics from my supervisor—almost as if the argument had never happened.
Later, one professor who didn't agree to let me pass a course without using Teams wanted to fail me for my “absences,” despite my uploading homework throughout the semester. After a protracted argument, I was offered an option to meet online on January 8th… on Teams! I politely refused again, and reiterated my points. The professor eventually CC'd the Associate Dean in an email. In the meantime, the deadline to upload my thesis for a January defense expired. After many reminder emails, a response finally came, and through the Dean's intercession I got a grade, passed my seventh semester and successfully defended my thesis in March.
Conclusions
Looking back, I'm proud of my actions. I took the risk of failing my studies, and I would end up with lower final grade than if I had submitted to the use of unethical and insecure software products. But I am content with this. I don't think surrendering to nonfree platforms would bring any long-term benefits—only more compromises.
We can see some people are intolerant to software freedom principles, but in the end those were few and most university staff at the AGH were actually kind to me. Thanks to them I now have a proof that it is possible to study, graduate… indeed to live without relying on proprietary software. After all this hard experience, I feel more independent than ever, and I even received appreciation from the well-known RMS [3]. Hopefully, my story will help more students get to where I am.
Struggling to run only libre programs forced, and continues to force me, to gain new skills. I now know enough about web technologies to make several sites function without JavaScript. But what is best about my experience is that I will be able to share my fixes with others and eventually make a subset of the World Wide Web usable in freedom.
[1] Thanks to Andy Farnell, Andy Oram and Richard Stallman for their help.
[2] Throughout this essay, I refer to all university teachers in the vernacular as my professors, although only some wear that official academic title.
[3] Dr. Richard M. Stallman, Founder of the Free Software Foundation and Chief GNUisance of the GNU Project.