Despite having just 5.8% sales, over 38% of bug reports come from Linux

Knowing a bug will be investigated would be enough, never mind fixed. Sadly I would assert that the vast majority of open source projects do not bother to respond to reports or even patches. The result of being ignored enough times is that I have become much less motivated to report bugs on projects that might actually care, unless there is clear and present evidence (openly available) that their issue tracker is sufficiently active. If I have to wait more than a week for a reply, you have lost me for good. If I can’t see evidence of the process working, I will not waste my time.

Most Windows users don't even know what a bug report is.

Great article, though he's benefitting from a specific slice of Linux users that are great to work with.

On the other end of the spectrum, try releasing a php plugin for Wordpress, OpenCart, Magento, etc. I released a mildly popular, open-source, free add on in this space. The Linux users there are decidedly less technical, but still doing it themselves.

What I got for "bug reports" was mostly terrible, few details/logs/etc. And a weird entitlement thing where people had really high support expectations over something that was open source and free. Even a few expletive-laden emails. I did receive a few really helpful reports, pull-requests, and thank you emails...but they were very much the exception.

Yep we find a variant of this with our python open source data science users who normal businesses would hate:

- Across all Python envs (Jupyter, streamlit/plotly, flask/Django, ...), they report way more bugs, and often from our free tiers, which sounds low ROI.. but our community caught all but ~2 significant bugs we fixed this year that our internal automated testing missed and would have otherwise gone to our Enterprise and gov users . They are happy to exchange a tiny bit of GPU or viz burps for new features faster, as long as we are responsive to their reports! That also pushed us to a better dev cadence which wouldn't have worked for our Enterprise users. More subtle, we have more confidence for deploying wider features/APIs which would otherwise be too underused + risky for the same reason.

- they are often doing innovative new things with our stack, which originally meant feature requests for non-repeatable sales.. but ended up stretching our flexibility into wider applicability, and they can clearly tell us areas to advance the visual, analytic, etc sides of our R&D and package it up for the rest of their internal analysts. Almost like a multiplier for conversations with our more operationally focused no-code/low-code users, and helps us see around the corner for them in the AI+API+customization sides!

- most of our technical partnership discussions now go through our OSS client repos. Rather than PowerPoints with BD people that don't have staff to do things, a product dev or sales engineer/architect or devrel blogger will think, "oh our customers would benefit from better interactive visuals on our data, let me whip something up quick with that package" and we help them take it from there

- open source data users like PyData developers and data scientists, compared to say the # of SMBs out there or general enterprise employees, is small and almost by definition doesn't pay for most software... but they end up being attached to projects that do, and influence much of their organization that definitely does!

On the other hand, a Windows user is trained not to report bugs, especially in games. There's rarely a channel for reporting things & simple bugs go unfixed unless the community can work out some hack to do it for themselves. Games are abandonware monoliths on day one.

Secondly, if a program crashes in Windows and you try to force kill it, you get blocked waiting for the 'report' to be sent do Microsoft. Users are trained to click off this, so the OS will finally do what you asked it to.