[jurassic]

A few years back I was in a role where it was my job to maintain a small but non-trivial open source project (~1500 github stars, 5 core contributors). I burned out after about two years in that role. People repeatedly demanding help in the github issues without even providing repro steps. Hate mail to the project email list from users angry we didn't implement the feature they want in the latest release. We had people second-guessing all the time if we didn't ship the things they wanted or they disagreed with the decisions we made. Some people just like to get in touch to let you know your work sucks.

The worst was the CEO of a small competitor company repeatedly hassling me over slack DMs and in github issues threads shamelessly trying to get me to prioritize free work that he wanted for his clients. Separately, over a period of several weeks, this same person asked me a ton of support questions and then compiled my answers without edits or attribution into a "white paper" he used as a resource to get people into the marketing funnel for his consulting shop. I came to hate that guy, such an asshole, but I couldn't totally block him because his company was a member of the open source foundation governing the ecosystem of tools my project was under.

Honestly, it was hell. The most stress and irritation I've ever had in any job. When I started the job I thought it was a dream getting paid to do open source, but now I consider that an anti-goal for any future role.

[morgante]

> People repeatedly demanding help in the github issues without even providing repro steps.

This is the biggest annoyance for me. There's a very clear issue template on my repos. It provides concrete, reasonable requirements (for example, post the actual config that broke!) yet people frequently expect support without filling it out. Even worse, they will make drive-by comments/demands on random PRs instead of opening an issue.

It's kind of amazing how entitled people are.

[smashah]

I literally print out a mostly pre-filled github link with most repro context, all they have to do is fill the rest of it in. They still don't do it.

Nowadays a non-templated bug issue is autoclosed.

[danielheath]

I am 90% of the way to adding "You have been banned from this issue tracker for wasting maintainer time. In future, make some attempt to do your share of the work." to the autoclose.

[jamie-vesoft]

Autoclosing is a good take.

[account42]

Depending on the userbase (e.g. end users from non-OSS ecosystems) and what you want to achieve you might just have to accept they are more used to talking to corporate support agents than filing bugs and learn to work with them. Autoclosing immediately puts the user into an adversarial position - better would be to have an automated reply for incomplete bug reports stating what is missing and that the information is needed for you to help the user to make it clear that filling it in properly is in their best interest as well.

[seer]

Do you think if more people provided positive feedback would have helped? I remember when working on customer relations roles in startups (devs wear many hats there), we had lots of bad responses, but some were just outright praise.

Like "thank you, you've changed our live for the better" kinda stuff, and it always made the work of wading through the muck more bearable.

[jurassic]

Positive feedback is nice, but it doesn't negate the amount of assholes and entitlement I needed to deal with on a regular basis.

[raxxorrax]

This is a problem of socialized open source on platforms like github. At some point you have to just say 'no'. But yes, some "consultants" just shamelessly abuse your work for themselves.

Of course if you make it your job, it will come with all the negatives a job can entail.

[wintermutestwin]

Sounds like OS developers would benefit from Product Managers who could shield them from the users (among other PM tasks). Do many OS projects have PMs? Are there any "standards" around non dev roles in OS projects?

[morgante]

Unfortunately my experience is that most PMs are not sufficiently technical. They will often add more work by injecting useless processes and try to "demonstrate value" by coming up with new ideas without actually understanding the problem domain.

The best approach I have found is simply to add more automated gates/bots. If people have not done what the bot asks, there's no need to waste time/energy on them.

[outsomnia]

> Unfortunately my experience is that most PMs are not sufficiently technical.

Yeah. And I eventually came to realize: lazy.

[kubanczyk]

> > Unfortunately my experience is that most PMs are not sufficiently technical.

> Yeah. And I eventually came to realize: lazy.

Lazy is just one side of the coin. If someone's career is focused on unsexy hard work instead of halfheartedly "demonstrating value", the core problem is not that they cannot slack off. The core problem is that others can hold them hostage.

[buovjaga]

Bigger project here, but over at LibreOffice, our shield is a strong quality assurance team. 20-30 active triagers. The number has stayed constant over the years and we actively recruit. How I do recruiting: https://discourse.sustainoss.org/t/how-i-recruit-and-mentor-...

[roastytoasty]

I don't understand, if it's open source why don't they just add whatever features they want.

[SQueeeeeL]

Laziness/lack of understanding