Comment by jrochkind1 | Hacker News Reader

Comment by jrochkind1 | Hacker News Reader

Parent Full thread

It would be very easy to contribute by writing documentation to most open source projects, they'd be happy to take it.

Coordinating teams and planning releases.... I think a lot of open source projects could _use_ help planning releases (release management is incredibly important), but I'm not sure they'd _take_ it from a stranger they didn't trust (and they might be right). "Coordinating teams" not as sure how it would even look or would be needed. But anything that looks like being a "boss", people are going to be resistant to, esp from a newcomer stranger, for obvious reasons.

But I bet people will welcome documentation contributions with little hesitation. For the rest... if you want to try, I recommend taking a product you're familiar with (as a user, not a developer), and offering your services being very careful to come from a real service perspective. Your job is to figure out what they need help with and how they want it done, and help them do it -- not to tell them what they ought to want to do or how they ought to want it to be done, not at first even to helpfully _suggest_ it. At least not until you've built up a lot of trust.

After writing documentation, I think the thing you mentioned people would be most welcome to and least resistant to would be 'triaging issues' -- but I think you'll have to build up some trust (say, by writing documentation!) before people will be willing to trust you with it.

This is how I've thought of it as well, why I've been hesitant to do much more so far, and why I'm polling a forum with lots of different maintainers rather than a single person at a single project already.

Because of my skills, I'm not going to stumble onto something by chance where I'm contributing upstream to scratch my own itch and then sticking around. I'm expecting I would need to spend time to find a project that seems like a good fit both ways, and become an active user intentionally where without my side agenda I'd be better off using a larger OSS project or a paid service.

I'm cool with putting that time in to help something grow and to help kindred spirits out. I don't want to put in that time just to be a source of annoyance or frustration to someone who's already done me and the world a solid by putting something cool out into the world.

Documentation is where it seemed to me like the best first step, and what people are saying here as well. As a followup, what's the general consensus on the least bothersome approach? Editing/prettifying existing documentation; appending to existing documentation (adding diagrams or examples); or creating "newbie docs" (how do I install this, how do I compile this on Windows, or aggregating a central FAQ from closed issues); or perhaps something else?

> Editing/prettifying existing documentation; appending to existing documentation (adding diagrams or examples); or creating "newbie docs"

I think any of those could be good, whatever the project needs. If it's got existing docs that need editing/prettifying, that's def a natural first step.

>But I bet people will welcome documentation contributions with little hesitation.

People should treat documentation issues the same way as critical bugs IMO. Can anyone point me to projects where this is done? e.g. documentation category on a bug tracker, open issues requesting additional documentation.

well not treating them as critical, but there are many projects that do so. One scala Project for example: https://github.com/playframework/playframework/labels/docume...

Unfortunatly once they are there not many PRs will address, these, only if their are actually critical. (Well the problem with play is actually that you need to read the README.md of the documentation carefully to make documentation and to understand a little bit of sbt and scala.)

Also it's actually not easy to write understandable documentation for many users. Some people are just too different, so that some actually need a way more detailed view, while for others this would be disturbing and useless noise, the golden middle is hard, especially if you contribute documentation to a framework which includes different libraries, it's really hard to keep the documentation simple.

P.S.: gitlab also treats documentation very seriously. (just to have at least two examples)

I remember noticing Vuejs using the issue tracker for documentation in the lead up to their release of 2.0.

FreeBSD also does this, which will surprise no one who has ever read the FreeBSD docs.

Also, one of the reasons I asked this question in response to antirez is because the Redis docs are excellent and have been since I first set eyes on the project three or four years ago. I don't know what he's doing or how he does it, but whatever it is, it's working.

Hello, I write the doc myself for the most part, since I believe the docs should be only written by persons that wrote the code they are documenting, or that fully understand it in the subtle parts. I happen to also love to write, so it's quite fun for me to do documentation, while I understand there are a seizable amount of programmers that don't like to write documentation a lot.

However to return to your original question of what is possible to do to help, issues verification/triage is a HUGE thing because the issues are of hugely mixed quality. Moreover Github does not help by making users not able to label the issues in order to make them into categories: only the collaborators can label issues.

Another problem I face is high quality QA, that is, a person that explores certain parts of Redis to understand where there is a quality problem and how it is possible to fix it.

Code reviews are also crucial. For a few months I had a contributor, Sun He, that prevented a number of bugs in Redis by reading each commit and by understanding non trivial bugs just reading the code. Impressive. Eventually he got a job after the university and now contributes only from time to time, but I'll be thankful forever ;-)

Reply on news.ycombinator.com