Favourite Album: Marco Arment and Phish →
I invaded Inquisitive’s Favourite Album series this week to explain why I like Phish so much and why you don’t, even though you’ve probably never heard any of their music.
I’m Marco Arment: a programmer, writer, podcaster, geek, and coffee enthusiast.
I invaded Inquisitive’s Favourite Album series this week to explain why I like Phish so much and why you don’t, even though you’ve probably never heard any of their music.
Apple Music, Beats 1, and why you may not want to force TRIM on your SSD.
On an infinite timescale, is Safari obligated to implement all web standards?
It’s been almost two weeks since I’ve had trouble seeing my home network’s shares, printing to the network printer, or screen-sharing with my home server, or since one of our computers or Apple TVs gained an erroneous “(2)” after its name.
It’s been almost two weeks since my Apple TV refused to see my iMac’s iTunes share, or since I had to restart iTunes and reboot the Apple TV or disable and re-enable Home Sharing to get them to (maybe) see each other.
In fact, it’s been almost two weeks since I’ve seen nearly any of the annoying, recurring problems that made me write, back in January, that Apple’s software quality had lost the functional high ground.
Not coincidentally, it’s been almost two weeks since Mac OS X 10.10.4 was released, which replaced the problematic discoveryd with the older, more reliable mDNSResponder. These system processes are responsible for tracking the network’s computers, names, and shared services, and discoveryd’s unreliability in these tasks caused erratic network problems like mine for a huge portion of Apple’s customers.
The entire decline of software quality that I felt in January wasn’t all due to a single buggy network-lookup service — but, unbeknownst to me at the time, a lot of it was. And that huge swath of problems that annoyed me every day disappeared instantly and completely as soon as I updated every computer on the network to 10.10.4.1
It’s still scary that such a critical component could ship in such a buggy state, breaking common tasks and tarnishing Apple’s reputation, without being fixed or reverted for almost a year.
But fixing this is major progress. And seeing the direction and priorities of iOS 9 and El Capitan, I have high hopes that we’re moving in the right direction again.
discoveryd persists on the iOS devices in the house that aren’t running the iOS 9 beta, and the Apple TVs, but upgrading the Macs has fixed every noticeable network-related issue in my daily use. ↩︎
The Verge did a blind test:
The results were very, very surprising to me. It was generally random across the board, though Spotify fared slightly worse than Apple Music and Tidal overall. In roughly 29 percent of the tests, subjects couldn’t tell any notable difference at all. Tidal — which wants you to pay more for lossless quality — most definitely didn’t take the crown, and in several cases, subjects actually identified it as the worst-sounding of the three.
Granted, it’s extremely hard to do a proper blind test of audio quality,1 and it’s arguable whether the “pick the best” method or an ABX test is more relevant. And I can quibble over some details of the test: the sample group was very small, they only tested three songs, and I find the Sony MDR-7506 headphones to be good for under $100, but not very good overall, and very poorly suited to this role.2
But even if you disagree with my nitpicks or have plenty of your own,3 it says a lot that in a fairly decent test, three different encodings (including lossless) weren’t reliably distinguishable.
If you need to try really hard to identify whether there’s any difference at all, it probably doesn’t matter.
Invest your energy and money in what matters so clearly and obviously that nobody needs to strain to hear the difference: great headphones or speakers fed by great recordings of great music.
The hardest problems: making sure each source is playing the same recording/mastering of the songs, and making sure the volume level is properly matched between them. If there’s even a slight, nearly imperceptible difference in volume levels — some say the difference can be as small as 0.1 dB — people will subconsciously favor the louder one. ↩︎
Most decent modern headphones resolve far more detail than the 7506, and its harsh midrange and rolled-off bass and treble can mask a lot, good and bad. Many of the specific complaints about the sound in The Verge’s video — “crunchy” cymbals, etc. — were identifying common flaws of the 7506, not the music’s encoding. This 24-year-old headphone is well-suited to monitoring cameras and sound boards cheaply, comfortably, and durably, not critical listening or enjoying music. ↩︎
Audiophiles will find much more about this test to complain about, like using an iPhone’s built-in amp and DAC and not using overpriced placebo-quackery cables, but none of these matter for this test. (Or ever, for the most part.) ↩︎
New iPods are released, sort of, and I compare web apps to Visual Basic.
Great perspective from a former Target manager on who’s still buying iPods in the age of seemingly ubiquitous smartphones.
I wouldn’t have guessed a lot of this.
There’s no hole count on Twitter.
The first Bluetooth wireless headphones I’ve found that are good enough to be included.
Did you know that The Verge delivers you to around 20 companies for advertising & tracking purposes? I didn’t. That might foul up your mobile web experience a little bit. Maybe we should try something different.
The Verge is an especially egregious example, but this is incredibly commonplace.
I’m interested in running a content blocker not because I don’t want to see ads, but because I feel the need to fight back against being opted in, without my knowledge or consent, to third-party collecting, tracking, and selling of my personal data just by following a link.
And if such blocking becomes a big problem for publishers, it’s up to them to switch to ad delivery methods without these privacy invasions.
The B&O H6 has topped my portable, closed headphone mega-review for a while, so I had very high hopes for the H8. I hoped it would be just like the H6 with Bluetooth, but it’s really not.
That place is great. Nice staff, casual atmosphere, good food.
You didn’t like it? Really? Why? It’s great.
Oh, you got the fish? Rookie mistake. Don’t order the fish, it’s terrible. But everything else there is good!
* * *
Jim Dalrymple’s loss of his music library was painful to read because, as much as I use, rely on, and mostly like Apple’s products, we all know that there are some toxic hellstews that are best avoided.
We’ve all heard (or said) the blanket statement that Apple isn’t good at cloud services, but that’s not universally true.
I’ve been using Apple’s services to sync contacts and calendars since they were called .Mac, with very few issues and zero data loss over the years.
As both a user and developer, I’ve never had a single issue with push notifications, probably Apple’s largest-scale and largest-volume web service. It’s rock-solid, fast, and reliable.
Even the new Photos app and iCloud Photo Library have been rock-solid not only for me, but for everyone I’ve talked to. The sole complaint I’ve heard is slow performance with very large libraries. I’ve had no issues with sync performance, data loss or duplication, or any confusion about what’s where.
But the iTunes Store back-end is a toxic hellstew of unreliability. Everything that touches the iTunes Store has a spotty record for me and almost every Mac owner I know.
And the iTunes app itself is the toxic hellstew. iTunes has an impossible combination of tasks on its plate that cannot be done well. iTunes is the definition of cruft and technical debt. It was an early version of iTunes that demonstrated the first software bugs to Grace Hopper in 1946.
Probably not coincidentally, some of iTunes’ least reliable features are reliant on the iTunes Store back-end, including Genius from forever ago, iTunes Match more recently, and now, Apple Music.
iTunes’ UI design is horrible for similar reasons: not because it has bad designers, but because they’ve been given an impossible task: cramming way too much functionality into a single app while also making it look “clean”.
iTunes is designed by the Junk Drawer Method: when enough cruft has built up that somebody tells the team to redesign it, while also adding and heavily promoting these great new features in the UI that are really important to the company’s other interests and are absolutely non-negotiable, the only thing they can really do is hide all of the old complexity in new places.
With the introduction of Apple Music, Apple confusingly introduced a confusing service backed by the iTunes Store that’s confusingly integrated into iTunes and the iOS Music app (don’t even get me started on that) and partially, maybe, mostly replaces the also very confusing and historically unreliable iTunes Match.
So iTunes is a toxic hellstew of technical cruft and a toxic hellstew of UI design, in the middle of a transition between two partly redundant cloud services, both of which are confusing and vague to most people about which songs of theirs are in the cloud, which are safe to delete, and which ones they actually have.
Even Jim’s follow-up piece, after meeting privately with Apple in PR-damage-control mode, is confusing at best about what actually might have happened, which is completely understandable because it sounds like even Apple isn’t sure.
I have plenty of plausible theories on why iTunes didn’t get the iCloud Photos treatment — why Apple Music was bolted onto this ancient, crufty, legacy app instead of discontinuing iTunes, dropping its obsolete functions, and starting fresh with a new app and a CloudKit-based service. (Engineering resources, time to market, iPods, Windows, and people with slow internet connections.)
But that doesn’t make iTunes better or less confusing, and Jim’s still missing those Ozzy tracks.
The safest, most sensible course of action for users is to just keep their music libraries away from iTunes Match and Apple Music. We’ll all just know not to order that fish, and many of us won’t use Apple Music at all because its integration into our local libraries feels too unsafe.
And that’s too bad for everyone, because Apple Music is pretty great when everything works and you can figure out where everything is.
Rootless in El Capitan, the sad state of the Mac App Store, and a bit more on Reddit.
https://marco.org/2015/07/31/apple-music-matches-files-with-metadata-only
Great investigating by Kirk McElhearn.
This is embarrassing. No wonder people have had so many problems and so much data loss with Apple Music’s cloud-library features.
It’s as if nobody who made this implementation decision had ever encountered remasters, re-recordings, clean versions, live performances, or the many other extremely common reasons why two very different audio recordings might have the same artist and title.
What’s even more puzzling is that iTunes Match did this much more correctly using acoustic fingerprinting, and for whatever reason, Apple Music chose not to. And if you’ve activated Apple Music, it seems to override iTunes Match similarly. (Update: This may have been “a glitch”.)
I highly recommend disabling iCloud Music Library and migrating away from any reliance you have on iTunes Match. You can still listen to Apple Music without them. Don’t let these cloud-matching “features” anywhere near your music collection.