4 ms·

That's pretty much how it happens, in my experience. Contributing to OSS projects generally happens when you need it, and often stops when you don't / it isn't

by STRML 12y ago

That's pretty much how it happens, in my experience. Contributing to OSS projects generally happens when you need it, and often stops when you don't / it isn't your job anymore. My most significant contributions to projects always happened when I was using them in my day job. It's inevitable that a project will have bugs, or you'll want a feature it doesn't have yet. I love it when people contribute instead of just opening issues; we're all busy, many of us maintain a dozen or more projects and it's hard to keep up when there is family, a day job, and so on. I usually merge PRs with passing tests within an hour - without tests, it could take weeks, and more open-ended "should we do X?" issues can be open for months or nearly a year. I see this all over the OSS ecosystem in so many projects. The best way to keep a project going is to give a few people who are using it every day commit rights or encouragement to submit PRs as often as possible. If the main collaborators stop using the project, it can easily stagnate and die. It's okay; generally, a project will eventually be picked up by someone else. All I ask of project maintainers is that they are willing to give the reins to someone else when it becomes clear that they no longer have the time for the project. Most maintainers are very good about it; some seem to have disappeared off the face of the earth. So it goes. You can always rename the project.

tnorthcutt 12y ago
I love it when people contribute instead of just opening issues I understand the sentiment, but please keep in mind that the set of people who use a piece of software and are able to file a useful issue does not always (or probably never does) overlap completely with the set of people who are able to contribute code to a project. Well thought out issues can be valuable to an OSS project, and should IMO be thought of as a form of contributing. Same with documention.
- jghn 12y ago
  Sure, but that doesn't mean the GP has time to deal with those properly. I maintain a fairly small time open source package, and generally the users aren't super savvy in terms of coding, I understand that. It's an R package, and most of my questions I receive are really basic R questions, not about my software. C'est la vie. But while I understand that doesn't mean I have time to deal with it much anymore. Responses tend to come in large batches once every couple of months just because that's when I'll finally get a chance to devote a decent block of time to the thing. OTOH if I get a non-horrible PR I handle it relatively quickly. To your point, a well thought out issue is appreciated by me, but it's just not going to get handle anywhere near as fast as a PR - that's not a judgement on the submitter, just a reality of free time
- sanderjd 12y ago
  I totally agree with this. I would hate to think there are a bunch of issues with something I've made lurking out there but unreported because the people who discovered them didn't have the time or ability to fix them and felt bad about just filing an issue.
gingerlime 12y ago
> It's okay; generally, a project will eventually be picked up by someone else. All I ask of project maintainers is that they are willing to give the reins to someone else when it becomes clear that they no longer have the time for the project I tend to agree, but this is actually the only thing I can think of that doesn't work as well as I wish it did. Or could be improved. There's quite a lot of abandoned projects, or ones that seem active but in fact lack this steering / leadership. You can submit a PR to one of those projects and never hear back, or bump into long delays that have nothing to do with the quality of your own code or its place within the project. In those cases, renaming the project or just forking it doesn't really create a replacement. Nobody knows which fork is the one to follow from now on. Even if lots of people +1 a PR, it's still "hanging there". And this gets exponentially more tricky with many different PRs from different contributors for the same project. This is more a cultural than a technological issue, but worth mentioning and thinking about. I wonder if there are some tools that can help the community as a whole have a stronger force. Perhaps some voting system, or a better weighting system to point to the "strongest" fork... just a couple of very rough ideas. * Use the fork, luke.

by STRML 12y ago

tnorthcutt 12y ago

I love it when people contribute instead of just opening issues I understand the sentiment, but please keep in mind that the set of people who use a piece of software and are able to file a useful issue does not always (or probably never does) overlap completely with the set of people who are able to contribute code to a project. Well thought out issues can be valuable to an OSS project, and should IMO be thought of as a form of contributing. Same with documention.

jghn 12y ago
Sure, but that doesn't mean the GP has time to deal with those properly. I maintain a fairly small time open source package, and generally the users aren't super savvy in terms of coding, I understand that. It's an R package, and most of my questions I receive are really basic R questions, not about my software. C'est la vie. But while I understand that doesn't mean I have time to deal with it much anymore. Responses tend to come in large batches once every couple of months just because that's when I'll finally get a chance to devote a decent block of time to the thing. OTOH if I get a non-horrible PR I handle it relatively quickly. To your point, a well thought out issue is appreciated by me, but it's just not going to get handle anywhere near as fast as a PR - that's not a judgement on the submitter, just a reality of free time
sanderjd 12y ago
I totally agree with this. I would hate to think there are a bunch of issues with something I've made lurking out there but unreported because the people who discovered them didn't have the time or ability to fix them and felt bad about just filing an issue.

gingerlime 12y ago

> It's okay; generally, a project will eventually be picked up by someone else. All I ask of project maintainers is that they are willing to give the reins to someone else when it becomes clear that they no longer have the time for the project I tend to agree, but this is actually the only thing I can think of that doesn't work as well as I wish it did. Or could be improved. There's quite a lot of abandoned projects, or ones that seem active but in fact lack this steering / leadership. You can submit a PR to one of those projects and never hear back, or bump into long delays that have nothing to do with the quality of your own code or its place within the project. In those cases, renaming the project or just forking it doesn't really create a replacement. Nobody knows which fork is the one to follow from now on. Even if lots of people +1 a PR, it's still "hanging there". And this gets exponentially more tricky with many different PRs from different contributors for the same project. This is more a cultural than a technological issue, but worth mentioning and thinking about. I wonder if there are some tools that can help the community as a whole have a stronger force. Perhaps some voting system, or a better weighting system to point to the "strongest" fork... just a couple of very rough ideas. * Use the fork, luke.