Have you ever written software? Do you know what the acronym WFM stands for? Do I cringe every time I shrug sheepishly at the person in charge of customer service? My company's software is used by millions of people and man do they run into some of the most esoteric shit that we cannot produce in the lab. Once I got lucky and my sister's off-brand Android was displaying an edge-case bug we weren't even sure was real. Side-loaded a severely over-logging build and finally found the sucker.
Special iTunes build just means it dumps logs faster than... Never mind. The images that just popped into my mind weren't where I thought that was going as I typed it.
Exactly -- the broader the user base, the more likely it is you'll run into people with really bizarre and difficult-to-replicate issues. I've seen our main product grow from nothing to an industry standard, and darned if we aren't still seeing bugs crop up that had gone undetected for years.
People have been grumbling on blogs about this for some time -- google trends for "apple music deletes library" spike for about the last year or so. Wasn't it posted on HN?
Given Apple's original, incredulous response to Vellum (see links elsewhere in this thread), its 'known issue' status, the seemingly tepid response up-until-now, and the amount of time they have had to solve this most arrogant of bugs, my money is on the their appearance being some bright-eyed PR fix. I'd bet Apple's management doesn't give a shit about your files when they compete against Apple's pay service.
Edit: and good lord people! Empathize with me long enough to think about how, possibly -- just possibly -- someone might have a different interpretation! This is about tardiness and arrogance, not an insult to engineering. Sheesh
We get crash reports for our games on PS4 - there's one type of crash we got a few thousand reports for(out of a few million users, so statistically it's almost nothing). The callstack on that crash seems to be literally impossible, so we contacted Sony about it - they said it can only happen if the user's hard drive is damaged/corrupted. But there is no way for us to tell, so as far as the customer is concerned, it's our game that is crashing.
A similar problem occurs even when you control all the hardware.
I work for a company with tens of thousands of servers, and the number of strange things we run into is crazy. The exact same software running on the exact same hardware, and sometimes it just behaves differently.
We like to think that computers behave perfectly predictably, but so much randomness goes on.
Two threads, on one processor, out of dozens, out of hundreds of thousands of machines. Those two threads would do exp() in the wrong direction, somehow, but only with a certain byte pattern. A byte pattern that corresponded to some floating point number we used in a few places.
On that machine, everything using that library would stall. But only when it ran on that processor. Everywhere else, it was fine.
I've had bugs relating to EMI issues crop up in testing before. The device in question had a compressor mounted inches from the touchscreen. It turned out the EMI from the compressor would generate phantom touch input on the display. It looked like a bug until I added a bunch of debug logging to the device.
Even if the computer is behaving predictably, the environment can do a wonderful job of making it go haywire.
We do this for customers all the time, customer has a special hardware/software setup, so we will send people out on site to diagnose odd hardware problems.
What would you rather them do? Folks can't have it both ways. In one corner, people bitch because apple isn't doing anything, and in the other, they start bitching because they did do something.
This stuff gets old after awhile. The older I get the more I hate the Internet.
Me too. How about Apple not fucks up in the first place? How about people put out software that works and not this MVP bullshit that passes for a finished product? How about respecting the user by default?
Except even NASA projects written to those standards had and have bugs still.
Apply Mission Critical/Life or Death practices to software can reduce the amount of bugs, but it doesn't take much searching to find medical software and other space software that has shipped with bugs, some with dire consequences.
I don't think the consumer software industry is prepared to take on those costs to still end up with bugs anyway.
I think thats mostly an illusion, and the costs for that "last mile" of quality is huge.
Think of it this way, if the software works for 99.9% of people, that's still 800k people that can experience undesired behavior. You hear "everyone has this problem!" but in aggregate, its not. Yet 800k is still a lot of people! Yet its 0.1% of iOS devices sold!
In short, what you are asking for is easy to say with words, but in reality its pretty hard to do, even if you have "more money than a nation" or whatever comparison people are using.
You're not wrong, definitely. It's just that criticizing iTunes feels like picking on the mentally-challenged. As rarepostinlurkr points out, even the "best of best" practices can't really guarantee bug-free software. And Apple is very far down indeed on any list of respected software development institutions.
The only rational approach to iTunes is to assume it will either delete your music or otherwise render it unindexable, and prepare accordingly.
The deletion thing has been talked about for months, but no-one has been able to actually replicate it. I've never seen the problem, no-one I know personally has, and I'm sure if Apple could see it happening in-house they would be all over it.
That's the key point. Apple should never be deleting stuff from users' devices. Never. Disk space is cheap. IP status is too ambiguous, and none of their damn business.
And a special version of itunes? Does nobody over there collect mp3s?