3 ms·

> “History of incompetence” really needs some citations Here’s a big one: https://arstechnica.com/gadgets/2007/11/be-cautious-of-that-serious-data-loss-bug-in-

> “History of incompetence” really needs some citations Here’s a big one: https://arstechnica.com/gadgets/2007/11/be-cautious-of-that-serious-data-loss-bug-in-leopard/ https://arstechnica.com/gadgets/2007/11/be-cautious-of-that-... The article contains few details. More insight is at the discussion on https://www.cableforum.uk/board/showthread.php?p=34430700 https://www.cableforum.uk/board/showthread.php?p=34430700 Briefly, when implementing file moving in (closed source) Finder code, someone at Apple who was apparently a beginner programmer or student intern made an elementary error that reliably destroyed files. The system already had a battle-tested mv command in its BSD subsystem, but they didn’t use that code. The point of this story (and, yes, there is a large handful of similar stories) is what it tells you about Apple’s culture and practices around testing, care, and concern for real soundness and quality (as opposed to appearance).

Klonoar 2y ago
Something from almost 20 years ago doesn’t quite work the way you think it does. Apple does take this stuff seriously. Implying otherwise is just absurd.
acdha 2y ago
So you’re saying that having a data loss condition 17 years ago which only affected a subset of users who lost their storage connection in the middle of a move operation tells us something about their products today? Does that also mean that I shouldn’t use Linux because ReiserFS lost data on local storage in some cases back then, or btrfs did so more recently? What policy are you saying we should use to evaluate software? In particular, what’s left if we’re saying it’s no longer good enough to have learned from mistakes? I supported a lab full of computational scientists who heavily used their Macs with external storage at that time and this caused no problems for anyone, so I also am looking for something more quantitative like whether a bug which only affected handful of users disqualifies usage for everyone.

Klonoar 2y ago

Something from almost 20 years ago doesn’t quite work the way you think it does. Apple does take this stuff seriously. Implying otherwise is just absurd.

acdha 2y ago

So you’re saying that having a data loss condition 17 years ago which only affected a subset of users who lost their storage connection in the middle of a move operation tells us something about their products today? Does that also mean that I shouldn’t use Linux because ReiserFS lost data on local storage in some cases back then, or btrfs did so more recently? What policy are you saying we should use to evaluate software? In particular, what’s left if we’re saying it’s no longer good enough to have learned from mistakes? I supported a lab full of computational scientists who heavily used their Macs with external storage at that time and this caused no problems for anyone, so I also am looking for something more quantitative like whether a bug which only affected handful of users disqualifies usage for everyone.