If you’re a regular bug reporter, you know the pain new betas bring. A pile of email in your inbox asking you to test out all your previous issues to see whether they’re resolved in the new beta.
It often feels like a scattershot approach. This past week there was one from back in June. It involved a rather large code base in the sample I’d submitted and I had no interest in trying to re-write however many lines of Swift 0.5 code in Swift 1.2 just to test out a simple bug. I just didn’t care about it any more.
The biggest problem with radars is there’s just no incentive other than the hope things will eventually get better. Plus, there’s the whiplash effect where some things they want you to file dupes “Every radar is a vote!” and the ones where they don’t. For example, Chris Lattner has asked developers not to file duplicate feature requests for Swift.
Sure, if no one files a radar, there’s a good chance that things won’t get fixed. “But that,” a colleague points out, “is often indistinguishable from the state in which you file them.”
Many devs can simply rely on bug-style herd immunity. You can relax on faith that easily-hittable radars are filed by those more eager, less cynical, and less time crunched than yourself. At what point is “another day, another radar dupe” outweighed by the reward of a possible fix?
That Apple hasn’t instituted some kind of reward system continues to mystify me. Why not earn discounts on yearly licenses or bonus karma for use in the WWDC lottery? Even a free frozen turkey would help motivate not just initial bug filings but also those endless “was it fixed in this beta? or this beta? or this beta?” retreads that follow.
I’m feeling a bit burned out on bugs. What do you think?
As a counterpoint to this, I’ve actually had a bunch of radars I’ve filed on Swift/clang/lldb get returned to me as fixed lately. Radar can be frustrating, but it’s also the only way anything gets done.
Also, I assume you’ve seen it, but on the dev forums there’s an official post explaining why a batch of radars were returned as “is this still valid?” and apologizing for the lack of explanation: https://devforums.apple.com/thread/263089?tstart=0