I think they all got rid of QA, and developers don't have enough time to actually care about what they're building/fixing.
GitHub/Microsoft is incredible at this. The feature for selecting and closing multiple PRs was broken for weeks even after I and others reported it. I saw duplicate "A user you've blocked has contributed to this repository" warnings (two of them, one above the other) for weeks too before that was fixed.
I've stopped reporting issues as well, as it seems to have no effect on how quick it takes, so I don't think it's an issue of not being aware of them.
I don't have any insider information here, but it's likely that the creator of this menu is long gone (it looks old!) and this feature is in maintenance mode. Someone--or a bot--just trying to keep the build green may have disabled the test without understanding that it was pointing to a real issue.
Sometimes I wonder what testing these giant companies do for features.
A week ago Facebook's language translation broke. Results come back in the original language or a mix. How could no-one notice?
> testing these giant companies do for features
I think they all got rid of QA, and developers don't have enough time to actually care about what they're building/fixing.
GitHub/Microsoft is incredible at this. The feature for selecting and closing multiple PRs was broken for weeks even after I and others reported it. I saw duplicate "A user you've blocked has contributed to this repository" warnings (two of them, one above the other) for weeks too before that was fixed.
I've stopped reporting issues as well, as it seems to have no effect on how quick it takes, so I don't think it's an issue of not being aware of them.
Google typically has very, very thorough tests.
I don't have any insider information here, but it's likely that the creator of this menu is long gone (it looks old!) and this feature is in maintenance mode. Someone--or a bot--just trying to keep the build green may have disabled the test without understanding that it was pointing to a real issue.
> trying to keep the build green may have disabled the test without understanding
> Google typically has very, very thorough tests.
Yeah, sure sounds like they have figured out this whole testing thing perfectly well, if that's the case :)
"and we increased test coverage by 10% on Q3" doesn't get you promoted unfortunately
No, but if you found real issues along the way, it's enough to get a decent review. Search takes their five nines seriously.
Or they noticed but cannot be bothered to fix.
They probably also sent AI on bug hunting quest there too like they were bragging for Chrome...