but the use is massively frustrating. Swipe controls
don't work, the app sucks, pairing with multiple devices
is a nightmare.
This is a great expression of one of the primary issues facing the software (and technology) industry today: it seems like software is fundamentally just too hard to do right. Basic competency and non-broken operation is regarded as a hugely valuable luxury feature. It's a cornerstone of the luxury technology brand!
But... is software really too hard? If so, how does Apple do it, are they just better engineers than the people who work at Sony? Maybe, but that seems unlikely to account for the difference.
And of course I don't really know, as I'm just a person on the internet, but I have a hunch: I believe Apple can do better because they have so much vertical integration, i.e. real control over much more of the technology stack in their product.
Let's say Sony has some in-house devs, but they hire a team on contract for the app's actual UI. Sony's just going to write the API layer for the UI to talk to. Sony will also make some of the hardware, but they're using a prefab chip and some antenna they bought, and they'll buy some proprietary off-the-shelf firmware for the lowest layer.
But, maybe it turns out the antenna and the chip don't play nicely under some key modes of operation. They can't redesign either of them, because they're just buying those parts. They shop around changes, but it would exceed reasonable costs. Maybe the firmware should be able to compensate here, but it's closed-source, and the company that wrote it shut down and sold their properties to someone else that won't turn any request around in under a year. And then, Sony sells the division that was working on the API, and they end up finishing the job out as a contract. This leaves all the API<->UI planning to contractors talking through an intermediary who still works at Sony. So it goes. And we all end up with another expensive piece of crap that just doesn't work.
But on Apple's side? I bet way more of it is in-house, and what isn't is locked in on deals that make Apple by far the most important customer. If they need a change, they get it. Apple's teams just... write the software that needs to get written! Maybe they physically meet each other. The firmware gets fixed when it's needed. And Apple ends up with a 'luxury product' built atop "hey look, it actually works!"
Is my point that everyone should be like Apple, or SpaceX? It's clear that not every firm can have tight vertical integration. Most have nowhere near the size, power, time, or budget to do this.
So my point is that if software and hardware were open by default, at least we could fix the damn things.
Sony maybe made some decent hardware there. It need not be shackled to some crayola software joke, if only they (and all of us) weren't so darn proud of our intellectual property that we could never dream of allowing someone to take a wrench to it.
Things being closed is so built-in that it's hard to even imagine this: how would we fix the headphone software? We'd probably need (expensive, hard to use) ROM-flashing utility hardware, decompilers, and the ability to load arbitrary code onto Apple devices! Impossible! Well, it's all hard because we want it that way. It's safer, probably.
There's a simple reason. From the top down, the culture at Sony is, "it's good enough for 90%". And getting that extra 10% isn't going to be worth the investment. Those folks can go buy Apple.
I have dealt with SO many (and worked for some) software companies that live by this axiom. Get an MVP to market, who cares if its riddled with bugs and hobbles along with duct-tape and paperclips holding it together. "We're just going to throw it away and rewrite it later". Devs spend all their days fixing little bugs with more band-aids - dreaming of the day when they can re-arch it properly. But the truth is, that day is never coming. Why? Because think of the expense vs the gain. It literally is not worth it to have a better experience. The crappy version is already making money. Customer Service agents are cheap. Achieving Apple-level polish is simply not worth it to most companies. That's why you have Chrysler mini-vans dying after 5 years vs Honda mini-vans lasting 15 years.
This is also an important factor in why public-sector and government solutions cost so much more. They have to serve everybody, regardless of disability, the tech available, etc. etc. A company can just dis-regard the 20% of the market that is expensive to serve; governments, the DMV, etc. can't.
That points to the other part of the issue. It is not enough to insist on 100%. It is also necessary to develop those solutions efficiently and to develop the right things (have good taste.) Both Sony and the public sector cannot achieve superlative results, for different reasons.
The magic comes from proprietary protocols. My airpods pro switch devices intuitively, but when I pair them to my windows laptop with bluetooth things don’t work so smoothly anymore and they are just as cumbersome as any bluetooth headset. I think the core problem of bluetooth headsets is the bluetooth spec itself makes it impossible to deliver a good UX.
This isn’t just apple either. My logitech mouse is paired to my laptop via their dongle, and to my desktop via bluetooth. On the laptop the connection is rock-solid, on the desktop it loses the connection once or twice a day for a few seconds.
I’m convinced the problem is bluetooth itself, not the device makers.
There's probably a lot of truth to this. It does go against the common sentiment here that open everything is always better. Well somebody will improve it themselves, right? So why hasn't anybody improved the open standard Bluetooth yet to meet those same standards of quality?
I think lately openness is massively overrated. This stuff is hard to get right, and it takes a lot of high-end hardware and tight collaboration between full-time engineers to do it. Random people working part-time in their garages and collaborating over Github will never do it. It works sort of okay for a few particular types of projects, but fails massively for many others, particularly things involving hardware. Only big corps can manage the budget and coordination required to do it right, and they'll only do it if it's closed, so they know they'll get the revenue from customers who want it done right.
I think this is exactly the crux of the issue, but I see a big difference: Bluetooth being "open" doesn't mean much when the kernel and device drivers and electronics are all closed. What would them being "open" mean -- specs are available? No, it would mean they are built in a fashion that allows them to be modified. It would mean that tools are easily available to the lay to allow modification.
I fully agree that just publishing the spec of a chip does not enable random people working part time in their garages or collaborating over github to do it, but creating software and hardware from the ground up with modifiability in mind would enable that kind of ad hoc work.
Would it solve all problems? Would it fix every bug? No, of course not! I'm not saying that open source is the answer to the ultimate question of life, the universe and everything -- I'm saying that a world built out of proprietary, competing, closed-off little gardens at the least guarantees that that which was created sub-standard cannot ever be brought to par.
Let me fix my own damned watch, and I really might.
It's funny because your comment would make sense if the Bluetooth spec wasn't a product of design by committee by a bunch of big companies called the "Bluetooth Special Interest Group", but was instead the fault of someone hacking from home without pants on.
While in reality, we could pool together some money to buy pizza for a couple of weeks for some experienced embedded developer with some RF knowledge and odds are they'd produce something better. Without pants on.
> I believe Apple can do better because they have so much vertical integration
That is a big part of it. But how did Apple do well enough to vertically integrate? They were not the crazy powerhouse they are now for most of their history.
Another big aspect is culture. Apple chooses not to ship things they don't think are good enough. Put aside that they're not always correct. How many other shops do that? How many places stick to the schedule relentlessly, because shipping is all that matters to them? Convince themselves they'll fix the software problems after release?
A big part, I think, is a combination of knowing when it is 'good enough', combined with the courage, patience and money to keep building until you get there (or spike the project because you missed the market window).
Didn't Apple start integrated? The Apple 1 and Apple II were largely designed by them, the OS was written by Apple, they licensed Basic from Microsoft. The Integrated Woz Machine [0] ran the floppy drive. It seems like it's part of their DNA, they always try to do as much as possible themselves.
I am only guessing that Sony outsources software development to developers who don't even use Sony. Apple developers love Apple product, look around the third party software on Mac, their developers love the operating system, and use it every day. It makes a huge difference. Elon Musk drives a Tesla, so he knows what it feels like to be a customer. I doubt Sony executives use Sony headphones in everyday life.
I definitely agree with this. It's quite absurd how many devices have a broken experience in just less than ideal conditions.
About 2 or 3 years ago I was looking for some wireless headphones who don't suck as much and there's an absurd defect that is shared by too many of them, some of them even on the higher end; the loud beep when the battery is getting low, some even had a recurring beep when the battery still has enough battery to keep listening. It makes some kind of sense with how batteries work, but it just breaks the experience so much you have to wonder why nobody thought of simple ways to avoid this behavior, but then you realize that is quite likely that the product was only tested with full battery and nobody on the team uses it on daily basis.
They don't let engineers design and implement UX/UI's but instead have them create properly designed UX/UI by UX/UI people who know what they're doing? It does make sense, I have no idea where does the notion of engineers doing UX/UI came from, would we want UX/UI people to actually code the software delivering their designs? ;)
If I were to hazard a guess, it would be a culture of bugfixing and software updates post-launch. These headphones no doubt share a lot of code with Airpods. Code which has received numerous bugfixes throughout it's life. Coupled with their ability to actually push out firmware updates to their users without needing them to install a separate (buggy) app, it leads to the evolution of a very stable software platform over time.
I'm just speculating, but how much embedded code do you imagine Sony throws away and re-writes for every new headphone release versus how much is retained.
But... is software really too hard? If so, how does Apple do it, are they just better engineers than the people who work at Sony? Maybe, but that seems unlikely to account for the difference.
And of course I don't really know, as I'm just a person on the internet, but I have a hunch: I believe Apple can do better because they have so much vertical integration, i.e. real control over much more of the technology stack in their product.
Let's say Sony has some in-house devs, but they hire a team on contract for the app's actual UI. Sony's just going to write the API layer for the UI to talk to. Sony will also make some of the hardware, but they're using a prefab chip and some antenna they bought, and they'll buy some proprietary off-the-shelf firmware for the lowest layer.
But, maybe it turns out the antenna and the chip don't play nicely under some key modes of operation. They can't redesign either of them, because they're just buying those parts. They shop around changes, but it would exceed reasonable costs. Maybe the firmware should be able to compensate here, but it's closed-source, and the company that wrote it shut down and sold their properties to someone else that won't turn any request around in under a year. And then, Sony sells the division that was working on the API, and they end up finishing the job out as a contract. This leaves all the API<->UI planning to contractors talking through an intermediary who still works at Sony. So it goes. And we all end up with another expensive piece of crap that just doesn't work.
But on Apple's side? I bet way more of it is in-house, and what isn't is locked in on deals that make Apple by far the most important customer. If they need a change, they get it. Apple's teams just... write the software that needs to get written! Maybe they physically meet each other. The firmware gets fixed when it's needed. And Apple ends up with a 'luxury product' built atop "hey look, it actually works!"
Is my point that everyone should be like Apple, or SpaceX? It's clear that not every firm can have tight vertical integration. Most have nowhere near the size, power, time, or budget to do this.
So my point is that if software and hardware were open by default, at least we could fix the damn things.
Sony maybe made some decent hardware there. It need not be shackled to some crayola software joke, if only they (and all of us) weren't so darn proud of our intellectual property that we could never dream of allowing someone to take a wrench to it.
Things being closed is so built-in that it's hard to even imagine this: how would we fix the headphone software? We'd probably need (expensive, hard to use) ROM-flashing utility hardware, decompilers, and the ability to load arbitrary code onto Apple devices! Impossible! Well, it's all hard because we want it that way. It's safer, probably.
But it boggles the mind.