|
Listen to this story
|
In 2021, the open-source community had a moral dilemma. The Freedom Software Foundation (FSF), a non-profit organisation that was a pioneer in the open-source movement, had re-elected Richard Stallman to its board. Stallman, who was the founder of the FSF and a professor at MIT, was battling accusations that he was in general toxic. He had publicly defended pedophilia, had been allegedly sexually harrassing women on the MIT campus and had a ‘racist and sexist’ hacker humour.
Open-source community leaders dubious morals
It is a strange irony that technologists who are the flag bearers of free software are sometimes distasteful people themselves. Stallman isn’t the sole trespasser, Linus Torvalds, the creator of the Linux kernel (the first OS that was free) is another prominent example.
A couple of years before the Linux kernel first appeared in 1991, the Finnish developer’s derogatory behaviour had forced Linux contributors to quit the community. Torvalds, who was known for his abusive outbursts to fellow coders had defended himself saying his cursing was necessary to keep everyone on the same page.
The toxic nature of such prominent individuals in open-source software is a symptom of the disease affecting open-source developer communities overall.
Toxicity in open-source communities
Last year, computer scientists at Carnegie Mellon University and Wesleyan University conducted a study to understand how online toxicity rears it ugly head in open-source communities.
In a paper titled, ‘“Did You Miss My Comment or What?” Understanding Toxicity in Open Source Discussions,’ the researchers stated, ‘Toxicity is different in open-source communities. It is more contextual, entitled, subtle and passive-aggressive.’ And every community had a specific flavour of toxicity.
An anonymous developer discussed the Python community on an online forum saying, ‘Being forced to seek other developers’ help in this software environment pretty often, I am faced with the fact that in response to my request for help, I usually get an unfriendly, reluctant answer. Sometimes developers, who are trying to help me, have time to argue with each other, choosing the best approach to the task at hand. I’m asking myself whether this Python community is an army of snobs, or it’s just me, looking for answers in the wrong place.”
The Carnegie Mellon study was also able to create a dataset of 100 toxic issues it found on GitHub. While the comments here were milder than the unrestrained severe language used on social media platforms like Reddit or Twitter, they were more entitled and arrogant. This is because most of the toxic conversations aren’t between anonymous trolls, these are work-related discussions between experienced open-source developers.
“In contrast to arrogant, insulting, and trolling comments, entitled comments seem to be a phenomenon more specific to open-source and the dynamics of free-to-use software, with seemingly free support despite no contractual obligations,” the paper stated.
Lack of value in moderation
For sometime now, Stack Overflow, the Q & A website for developers, has also been called toxic, especially to developers who are new to the forum. In 2019, George Stocker, a long-term Stack Overflow moderator quit after writing a blog post. Stocker stated the company was in a downward spiral since 2014 saying Stack Overflow had “forgotten how to lead, how to persuade, and how to talk with the community. They no longer want to build a relationship with the community they have.”
Stocker also noted that the company was undervaluing the valuable work that moderators were doing. He said they were being treated more like volunteers than partners in building a better community. Moderation rules are the single-most important ingredient to making an online community safer. Studies have shown that platforms with stricter moderation guidelines align more with the ideas of expressive and free speech.
Solutions to fix toxicity issue
Now that we can admit that toxicity is a problem, there’s work being done to solve it. Toronto-based AI startup Cohere is one of the companies that have come up with content moderation tools. The startup has built a set of APIs that can be used by any business to use NLP, as a standalone solution or even integrated into existing applications. Because of advanced NLP, businesses can shape their moderation policies according to their own nuanced set of issues. This is useful to detect toxicity in open-source communities where the language might be harder to be classified technically.
Until recently, toxicity in open-source has often been written off as either a natural product of it or sometimes even necessary to the open-source culture. And while there still maybe people who are used to or defend interactions that are ruder, there are consequences now for toxic offenders.
