3 ms·
As a maintainer of dozens of fairly small OSS projects on GitHub, I learned early the importance of 'No'. Also, unmentioned here, you have to be willing to igno
by geerlingguy 10y ago
As a maintainer of dozens of fairly small OSS projects on GitHub, I learned early the importance of 'No'. Also, unmentioned here, you have to be willing to ignore the 'me too' present in a lot of PRs and feature requests; what's most important to me is 'will this new code/feature make the project easier or harder to maintain?'
If harder, it had better be a stunning new addition that makes the extra maintenance worth it... And often, it had better include an automated test!
- peatfreak 10y agoOh yeah, and even in regular work too. I've noticed a lot of peer pressure in many different environments, sometimes team members feel obligated to provide a "me too" or "thumbs up" or whatever, out of fear of being the nail that sticks up and gets hammered back down.
- btown 10y agoAsking the author of a PR to include an automated test is a favor to the larger community - you're building momentum for this practice. As a maintainer, you shouldn't feel bad about introducing this sort of hurdle. (And if you don't have the infrastructure to run those automated tests, then it's a good excuse to get it up and running!)
- haikuginger 10y ago> Asking the author of a PR to include an automated test is a favor to the larger community Agreed. Once this practice is in place, it's much easier to keep going - we have 100% coverage on urllib3 and require any new contributions to meet that bar.
- Spearchucker 10y agoI imagine a similar rationale is used by big vendors to select features for products like Windows, Android, Safari, Outlook... The business proposition is sound, but to the user "Microsoft just never learns", or "Apple need to start bringing their hardware quality to their software", and so on.
- spdionis 10y agoMy thoughts everytime I hear about criticism about features of various big software products.