2 ms·
I think proactivity and a knowing a good platform to ask questions in, helps. For some of us, the difficulty is often whether to ask on StackOverflow or to open
by nevi-me 8y ago
I think proactivity and a knowing a good platform to ask questions in, helps. For some of us, the difficulty is often whether to ask on StackOverflow or to open a bug on Github.
An overflow of Github issues can lead to some issues going unnoticed (depending on whether issues are regularly triaged by a team).
For example, I have a Flutter Android app that has an infinite notification loop bug [1]. I've struggled for many hours (spread over a few weekends) trying to figure out what was wrong. I eventually opened an issue on GH after giving up.
After a few days, I decided to post a question on SO, and someone from the Flutter team saw it. The issue might have gone unnoticed for a bit if I hadn't done so (especially given its obscurity). I believe so because popular projects on GH get a lot of issues, and yours might get buried if someone doesn't pick it up immediately.
The other thing is that when we open issues, and project maintainers try to work with us, we often delay in responding to them. That over time, leads to issue clutter, and probably developer fatigue.
So perhaps we can be better citizens by being more proactive, given that someone who's often paid a lot is taking time to help us with issues.
---
I agree with your point on RN and in-house development.
[0] https://play.google.com/store/apps/details?id=za.co.movinggauteng.mytrain https://play.google.com/store/apps/details?id=za.co.movingga...
[1] https://github.com/flutter/flutter/issues/18524 https://github.com/flutter/flutter/issues/18524
- mythz 8y agoI'm not referring to technical support questions, I'm referring to actual issues with verifiable repros which works everywhere else (inc. their debugger). They dismiss issues with core functionality like their W3C fetch API not being able to maintain long-running HTTP connections which breaks solutions relying on them like SSE [1]. They originally closed it making it known they have no intentions on ever fixing it and have since locked it preventing anyone else from being able to comment on it either. RN enables a fast development model but you'll continually skate into an open mine field of issues where you'll end up wasting time on putting out fires and diagnosing issues only to discover they're known issues they have no intentions of fixing so you'll be spending time on implementing workarounds mitigating broken UX experiences in your App instead of focusing on delivering end-user value. Given the 1000's of issues that have received similar treatment I suspect React Native is just an architecturally flawed experiment and why we'll continue to see more users abandoning it. Hopefully they wont have wasted as much effort and resources as Airbnb's multi-year investment who despite having a large enough dev team to maintain their own fork was still unable to make RN work for them. [1] https://github.com/facebook/react-native/issues/12912 https://github.com/facebook/react-native/issues/12912