Hacker News new | past | comments | ask | show | jobs | submit login

There’s one issue, which is that of “Sherlocking” a third-party solution with a first-party implementation. Then there’s another, which is having an open enough system to support such innovations in the first place.

Something like Growl, or f.lux (mentioned down-thread) could never have come about if macOS had been as restrictive as iOS. I have little doubt that we’ve missed at least a few such innovations over past decade, especially on the iPad, due to this.

While the Mac will likely never, despite some tireless predictions, go fully locked-down, little things like the deprecation of kernel extensions will chip away at this from the Mac side as well. Market-driven innovation and platform control are a difficult balance, but they are ultimately a zero-sum game.

Well, yeah. Important things like tethering would never have made it into iOS if Apple hadn't been dragged there by the jailbreakers.

This has always been the domain of the carriers, not Apple.

I’m not from the US, could you explain this perspective to me?

You give me internet on device X. How is it up to you how I use it?

Same way it is up to me how you use books you bought (kindle) or songs you bought (any streaming crap nowadays).

If you're shitty enough person (good businessman) you can profit off of anything.

DRM on books is about the rights holder, not the person selling it. I can assure you that Amazon doesn’t actually give a shit what you do with your books outside of having an obligation to the rights holders.

Easy access to a huge library of books is one of the main selling points of the Kindle. DRM is good for Amazon.

Yeah, because they wouldn’t be able to offer it otherwise. If they could offer the same content without DRM at the same price, I’m certain they would do it. Apple did with the iTunes Store.

They wouldn’t, simply because their position as the dominant player in ePublishing gives them a massive incentive to lock-in users. Which is what their DRM does: you cannot move books to non-Amazon devices, so you’re stuck on Kindle forever. That this also satisfies publishers is something of a side-effect.

They did the opposite with music (they actually sold simple standard mp3s, unlike what you get from iTunes) because they were challengers, not incumbents, so the priority was to attract users.

iTunes has not sold music with DRM since 2009.

I assure you that they care much more about you buying books from them than about "lock-in" from you using their apps and buying their devices. Proof: their apps absolutely suck and their devices are not much better. The devices exist to get you to buy more books from them. It is absolutely not a side-effect that their DRM satisfies publishers.

In my experience in Canada from several years ago, the business teams in wireless telcos don't see it as providing you internet on device x. The ip connection on a device is a gateway towards a service, and the telecoms want to be providers of services, and the differentiator from competitors is having better and different services.

In my opinion the biggest fear of those business development teams is to be treated as or thought of just as providing internet on a device. As being just the pipe is not considered a lucrative market, and open to lots of competition where the race is to the bottom on price. This is also where you get lots of the anti-competitive behaviors, like not investing the effort to make a competitors service optimal within a network, as the carriers would rather launch their own service instead of making someone elses service they don't generate revenue off of work great.

To be fair, in the older gen access technologies, all internet access isn't equal. The access network was optimized for the usage patterns of a mobile browser, in the way it scheduled and idled the radio (I'm glossing over a large number of details). So connecting a computer via tethering that behaves differently and is far more chatty on the network, is actually a much costlier device to support. The current gen techs like LTE do operate much differently, to support chatty devices and apps. With a better optimized network and devices loaded with chatty apps, this difference is probably disappearing.

So if you put yourself in the shoes of a business dev person at a wireless carrier. They'll look at this and say, well we can have a $40 plan for just internet access. Or we can have a $35 plan for just mobile browsing, since that's more efficient and allows us to be more competitive when using the phones browser, and a $15 add on for tethering for those who want it and cost the company more. There are entire teams dedicated to just figuring this stuff out.

And this leads to some of the stuff I would get thrown from time to time, like the $60k bills (no longer allowed in canada, a rule was created to prevent surprise bills, but used to be a thing) when someone bought unlimited mobile browser for $7, and thought that meant unlimited usage for their tethered computer and started torrenting all day which was pay per use.

An unlimited $7 plan only works in the context of the way a user can use a mobile browser, and that's why the carriers think they can dictate how you use the internet connection. They have departments to tailor make plans that only work if they're able to dictate how the internet is used.

I'm not saying I agree or disagree with this model, I'm just trying to explain where I believe this perspective comes from based on my experience. I've been out of the industry several years now, and my opinions are my own.

This jives with my own experience. Great comment, thanks.

One group was trying to monetize services which were already free. Like make a list of data provided by stock mobile runtimes (Android, iOS), such as location data, and then create a developer portal for accessing that same data thru web services.

I'm having trouble remembering the details because I didn't even understand it at the time. My buddies were contractors, loved this kind of work, the money was good, and you could never fail. They thought there were helping me, get me some cheddar too.

I was briefly on the "mobile app services" team, it took me about 3 months to figure out what we were even doing. But there were lots and lots of meetings. Convened by empire builders and corporate climbers. Expensive suits. And lots of agile, kanban, scrum masters, velocity. To build something biggly impressive for which there would never be customers or revenue.

I assume there was at least one other group trying to "accomplish" the same thing.

The carriers would print money forever if they simply fired everyone unrelated to the actual network (cables and towers) and charged everyone a flat rate. And some overhead of the basic infrastructure services like provisioning numbers and 911.

No metering, no discounts, no custom phones, no carrier add-ons. The actual effort to create additional income streams was ridiculously wasteful.

It was crazy making. The actual features that would be awesome were never part of the conversation. For instance, I'd love to hear my voice mails on my laptop. I'm sure enterprise customers would love some tools for managing fleets of phones. Etc.

Maybe the carriers got high from the SMS profiteering, didn't realize that was a one-off bonanza, and have never recovered.

Much of what you said rings true with me and brings back chills of having a meeting to plan a meeting to discuss actually doing work, except for one piece.

> The carriers would print money forever if they simply fired everyone unrelated to the actual network (cables and towers) and charged everyone a flat rate. And some overhead of the basic infrastructure services like provisioning numbers and 911.

In my experience with these projects, I'm not sure it's safe to conclude that they add up to enough to significantly change the economics of the network. I know working on the projects feels like a huge waste of time and energy, but we're talking a million dollars here and there, when the cables, towers, and core people are spending 500 million plus per year. I'm sure it adds up, but all businesses have overhead, and I don't see it as a game changer.

As for charging a flat rate, I'm actually of the opposite perspective. I realize it's probably unpopular for this community, but for a long time I've been an advocate that mobile usage should be metered by use. The problem I have with the flat rates, is the users who use less always end up subsidizing the power users who use significantly more. Much of the network investment goes into supporting the top end users, but it's everyone else who has to pay for it. For wireline access this might be a bit different roi calculation, but for mobile wireless I think this is a real economics problem, and flat rate is not an incentive for a carrier to support or retain a power user who wants to use lots of data.

So personally I prefer a model that works more like electricity usage or filling a gas tank, where you pay for what you use, and you naturally get a feel for watching 4k video all day costs more. This of course needs the tools in place to understand where the usage is coming from, not to surprise anyone, etc, etc, so it's not a perfect model, but compared to years ago and the rates that could be offered, it seems atleast to me like a more natural model.

I'd be more okay with metered billing if:

- The fees weren't disproportionate with respect to any infrastructure investments. Why should my internet bill triple over the span of a few years while speeds decrease unless I call, wait on hold for a day, and ask to cancel service?

- There weren't an aspect of double dipping -- why are we being charged for peak bandwidth (with absolutely no guarantee of reaching those speeds) and _also_ being charged for exceeding bandwidth caps equivalent to a few dozen minutes of full use? Why is that "extra" bandwidth more expensive than buying several extra full internet plans?

- It didn't open the door to metering different content sources differently based on the ISP's monopolistic whims.

There are physical bandwidth limits, and those need to be allocated somehow, but the status quo isn't great.

Thanks for this detailed explanation. It’s really strange and I’m so glad this practice didn’t take hold where I’m from. Granted, mobile data is really expensive here so it probably wouldn’t make sense, anyways.

> This has always been the domain of the carriers, not Apple.

Maybe in the US. In Europe carriers never had any say about what functionality a mobile device offered.

All they provided was a SIM-card for that device.

Above what the sibling said regarding branding and pre-installed crap-ware, they still make a distinction between "on device" use and "modem use".

Source: read the fine print on an offer from my provider, Bouygues Telecom.

My regular plan has 40 GB, which I can use on the phone or tethered. This summer I was spending a lot of time at my parent's and their land internet connection was spotty. At the same time, Bouygues ran a campaign where I could get an extra 20 GB for next to nothing, so I was considering that. Read the fine print which basically said "the allowance doesn't apply to tethered use". I didn't care enough to challenge them on this so I passed.

Edit: Just checked the available options, this is still the case. Can't give a direct link for some reason, so here's a screenshot: https://imgur.com/a/rQgoz0m

Basically this is an option for "unlimited internet on the weekends". The originally hidden disclaimer says, among other things, "except for modem mode".

That's not true. Based in Germany, my mom still had, and I believe still has, a SIM in her phone that doesn't support tethering.

They did have a say, before iPhone and Android happened, devices have been branded, sim-locked, locked down etc by the carriers. I think they mostly stopped doing that in recent years.

I’ve been living in Norway since I was born back in the 70s.

We’ve never had carrier-branded phones. Not one. Only thing sold has been generic phones which accepts a generic SIM.

And that’s how the market is supposed to work. Free competition on devices. Free competition on service. Customers can combine as they like.

Granted you could buy carrier-locked phones rebated through a contract, but the carrier lock was time-limited and reversible and the phone was a generic, international model.

Carrier-branded phones was definitely not a EU-wide phenomenon.

If anything the introduction of the iPhone in Europe (launched using the very confusing US carrier-model) was what started pushing carriers into attempting to making new restrictions on how people were allowed to use their (formerly unrestricted) subscriptions.

So you got it pretty much 100% backwards.

You said "in Europe", when you must have meant "in Norway".

In the UK there were DEFINITELY carrier branded phones, tethering was disabled by many carriers, and you couldn't even use a regular SIM card in a non-phone device - you needed a "data sim".

I travelled around Europe for 2 years using local SIM cards - and also encountered carriers which disabled tethering.

There were even android apps specifically to work around these tethering restrictions, by making the phone act as a proxy.

Same for Austria, which interestingly was considered a "test-market" for international carriers, thus we always had the cheaper contracts and some novel business models quite some time before the rest of the EU (unfortunately not network generations or coverage in general).

The answer is money. Tethering was usually not allowed but you could buy in. You get phones for free, but only if you pay 40€+ a month for the next 2 years for something you actually don't need. E.g. some unlimited services (streaming) while your general data is capped.

I haven't been in the market for such contracts for quite a few years, this has changed a lot in recent times due to "contract-less" cheap providers gobbling up the marketshare. And these packages always disappeared over time and became standard. I don't think tethering is not allowed anywhere anymore.

This is how the free market should work. Providers tried to extract more money by forcing contracts, other providers swooped in with contractless plans and captured a significant market share.

I am pretty sure carrier-branded phones were a thing in Central Europe (Slovakia, Czechia, Austria, Germany etc) for many years. As the owners of carriers were also French (Orange) or Spanish (Telefonica/O2), I suspect they present there too. So no, not backwards at all.

They can put default APN configurations in the SIM card that (by default) would have different fares for mobile usage and tethered/hotspot usage. They cannot block you from changing the configuration, but I think in theory they could terminate the contract if you did so. In practice 99% people got used to buying both a "computer SIM" (which is typically sold with a cheap Huawei USB modem) and a "phone SIM", so the providers that do differentiate that way don't bother going after violations.

> In Europe carriers never had any say about what functionality a mobile device offered.

Not so, unfortunately. I tried putting a phone SIM into an iPad, and soon got a message from the mobile company saying the SIM card wasn’t intended for this use and would be disabled until it went back into a phone.

(This was in the UK a few years ago with a Three PAYG SIM).

Three is a special kind of nasty anyway.

They used to proxy all traffic and the only way to get out of their stupid slow proxy was to go to the shop with ID to get on an "18+" list.

They said it was because of some UK law but I was in Ireland. So not applicable. And other UK based providers like Vodafone didn't have this stupidity.

> In Europe carriers never had any say about what functionality a mobile device offered.

not really. a few years ago, before the eu roaming was made cheap, i bought a sim from wind operator in italy, and it was blocking my tethered traffic. i had to "fix" my ttl for it to work

Who do you think Ericsson and Nokia talked to when they planed their extensive and differentiated portfolio?

Oh, it would have made it, they just needed to figure out the best way to charge for it.

Was it ever Apple who charged for tethering? I always thought it was in the hands of the networks and their various plans. Today it seems like it's not even a feature anymore, just something you can do with your phone (or, at least for me it is: I'm on one of the cheapest cellphone plans in my country and tethering just works)

In the US tethering is fundamentally different (or at least the carriers want you to think that so they can upcharge it)

Yes, it is the networks that demand extra payment for tethering. That is still a thing in the US; you have to pay extra for a plan that includes tethering.

Surprising from a European perspective. I have used tethering since before phones had a HTTP/HTML browser (~ 2001, remember WAP?). The question has always been do you have packet data or not, what is your bandwidth and volume. Nothing else is the operator's business.

Not really. They still separate on-device use from modem use. At least Bouygues Telecom in France does.

See my other comment: https://news.ycombinator.com/item?id=25243710

hors modem, you are right. So France is different. I'm 100% sure that has never existed in Finland (well, the most common packages are unlimited anyway) and 90% sure it's not common in Germany either (they use throttling after you reach the limit)

So technically, how do they differentiate the traffic?

Most of the time, tethering is allowed in the base package. What happens afterwards depends. They can either cut you off or throttle. The only experience I have with this is Orange, and they throttle to the point where you basically have no connection (SSH is unusable).

Not sure though how they differentiate the traffic. There are "access points" set up which I presume are used for either connection. In my case they are the same but I seem to remember on an older phone they used to be different. It was also a different provider. I'm also not sure how they would detect that they're changed.

Screenshot of my setup: https://imgur.com/a/xH88Itp

Tethering and hotspot traffic is still marked with a different APN type ("dun").

Does APN type go out over the network?

I thought that's an Android internal thing to decide which of several APNs to use.

Also the first 2 references that Google returns say dun is obsolete and not in use with today's devices. I believe that belongs to the ATD *99# era

(Havent't worked in the field for 10+ years, could be wrong.)

No it doesn't go out, but the APN itself does go out on the network and can be used to apply different billing to tethered vs mobile browser data usage. https://news.ycombinator.com/item?id=25244052 for more info.

The networks demand it, but Apple provided them the ability to demand it.

Actually the same thing did happen on iOS with jailbreak tweaks being copied by Apple devs. (E.g. wifi sync, quick reply, predictive keyboard...).

Admittedly the pool of developers was restricted to those who were able and cared to jailbreak...

It's interesting to think about the fact that, now that Macs are starting to be differentiated more by hardware (again, due to Apple Silicon) than software, Apple might actually allow more tweaking/low-level access to do interesting stuff.

Performance of A14X is almost the same as M1, but it only became a sensation when the CPU was allowed to run macOS.

Think about it: Apple already had an awesome CPU, but hardly anybody noticed, because iPad OS is so limited.

Not true, there is no A14X (yet). Closest thing would be the A12Z in the 2020 iPad Pros which are impressive but ~25% slower than the M1.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact