Not quite relevant but: I once found a bug in apple's calculator app for the iphone (which I reported and they fixed after 3-4 months) where REALLY basic math operations like division gave different answers depending on the way you held the phone. (portrait vs. landscape).
I don't just mean the displayed precision changed, I mean the actual answer changed, fairly significantly.
In particular, if you accidentally tilted the phone in the wrong direction so that it switched to portrait then tilted it back, you'd get a different answer than if you held the phone steady.
I only found this bug while in the middle of a midterm for a graduate-level finance class...
Edit: details of one of the bugs:
Steps to reproduce:
1. Open the calculator application
2. Ensure the app is held vertically (non-scientific mode)
3. Enter a number such as 10,000,000
4. Attempt to make the number 10,000,000.1 by pressing "." and "1"
5. Notice that instead of this number, the calculator application converts the input to 100,000,001 - i.e. the decimal point was ignored.
Every time something like this happens, I'm just reminded again of how prescient Richard Stallman has been about walled garden stores and why we shouldn't spend our time and effort developing software for them. If app developers spent even a fraction of their current output on the truly free mobile OSs, well maybe Apple's psychological power over the market would be a much smaller thing.
Isn't that like arguing that you should avoid sex because you might get an STD, or avoid flying because you might die in a crash?
Just as the overwhelming majority of sex for most people will not result in an STD, and the overwhelming majority of flights will not end with your death, the overwhelming majority of mobile app developers won't have any serious problems with the walled garden app stores.
No this isn't the same. Stallman is referring to Apple's power over its suppliers (developers who sell software through the Apple store). Because Apple is a popular platform and switching platform costs are high (porting software takes time and money), Apple is able to extract value ($) in each and every transaction that could have gone to the developers or end users.
An open platform would remove 1 player from the value chain (Apple) and would increase value for end users, developers, or both.
No, most desktop users don't run Linux because it's not the default. People can perfectly well run Linux, it just needs to have apps, drivers and all the other stuff desktop users expect. In a world where Adobe doesn't port Photoshop and Micorsoft doesn't port Office, Linux operates at an enourmous disadvantage. For users that don't need those kinds of apps ported, Linux does great. Both Chromebooks and Android have sold incredibly well, for example.
I think you're right the value would change. However, holding value constant and replacing Apple with an open platform in the game would increase value captured by either developer or end user.
One could probably argue the value of an open platform could meet or exceed Apple, given enough time and money.
> One could probably argue the value of an open platform could meet or exceed Apple, given enough time and money.
GNU/Linux are at it for 20 years with money pumped from Intel, Google, IBM, Red Hat, SuSE and many others.
The only success on the desktop is in the form of Android/Linux and Google has the freedom to change the kernel for something else whenever they feel like it.
I would argue that the Web is an example of an open source and free platform. I predict that in 3-5 years most of the mobile apps will be in Javascript with platforms exposing PhoneGap-like standard interfaces.
The web is not locked down. Anyone can publish anything, and the source has traditionally been visible. And look at the explosion that it produced - some good and some useless but the huge usefulness of the internet in the last 10 years has been led by the web, not by proprietary app silos.
Like, I get that Free Software as a movement is about software being libre, not gratis. But how many examples are there where being libre has not meant being gratis? And I don't mean the Red Hat sort of case where the software itself is gratis (yes, you have to go all the way over to CentOS to get it gratis) but support costs, I mean an example of being able to charge actual consumers for a piece of libre software.
No, it's more like choosing to always fly in a single craft with a small-but-not-insignificant chance of a fatal crash, versus choosing to fly on other crafts which have a negligible chance of crashing, and are guaranteed to preserve your life if they do.
If you boil everything down to the most reductionist form, the majority of software development is an exchange of lines of code for food, clothes and shelter. There may be more freedom on other mobile OSes (although I'm not sure which you mean... Firefox OS? Is there another one?), but there isn't more MONEY. Stallman's notions of software freedoms don't put bread on the table. People are more willing to pay for software on the closed mobile ecosystems, for a lot of reasons (ease of use/user experience, hardware availability, etc.).
Actually, the median and average developer lose money writing for the app stores, even if on the other devices they gave their software away for free. registering as an apple developer costs 99$ a year, and essentially all apps on the App Store never make more than a few dollars.
"As of June, Apple has had 75 billion cumulative App Store downloads, paying out a total of $20 billion to developers, with $5 billion paid in 2014 alone. According to app-tracking site AppShopper, there are 1,211,461 apps available in the App Store, including over 1 million iPhone apps and 610,000 iPad apps."
There are literally billons of dollars sloshing through the App Store, not even counting ad revenue. Yes, most of the apps released don't see any of that. But some of them do. It absolutely IS possible to make enough to make a living selling apps on the App Store, or on Google Play for that matter. There's over a million apps on the App Store, and if only one percent of them is producing meaningful revenues, that's still over 10,000 apps, many of them likely supporting multiple developers. And I suspect that, even though the median payment is a pittance, it's more than 1% making revenue. And again, this isn't even counting ads!
What libre smartphone OS has that kind of revenue potential available to developers?
The existence of possibility does not magically grant probability, and revenue potential does not impress banks or mortgage lenders or grocery stores when the bills come due.
By the same token, your local 7-11 is not a magical fountain of wealth just because it has lottery tickets.
Any time you are told to labor for negative $100 in order to recoup a number very likely less than $100 over time that has 30% taken out of it for no apparent reason at all, you should think twice.
What are some examples of financially successful mobile software where the value lies in the mobile software, rather than a cloud-hosted server running a Stallman-esque Unix derivative?
Most of the games I would argue the value lies in the mobile software, and even for cloud-powered games there's none where the client is valueless. There's also a bunch of fitness apps, photo editing -- one of the top five is an alarm clock! Lots of apps are "financially successful" without just being a shim in front of a cloud server.
That "clock" is collecting sensor telemetry / healthkit data, which has secondary business models beyond software revenue, and definitely relies on centralized compute.
While I agree with Richard Stallman about the importance of freedom, this is more the old issue of building your business on someone else's platform, and your cash flow now being in the hands of a completely inconsistent company. This is not about censorship or monopolistic business practices - remember that Apple also makes money off each copy of PCalc.
This is the counter argument to the freedom argument: good, consistent design and UX seems to require totalitarianism. Otherwise you get a confusing hodge podge of junk.
I'm starting to see this whole area as a major 21st century problem in computing. How do we have freedom, security, and good UX at the same time?
This is what happens when your organization's focus on user experience eclipses its focus on the user. Apple has allowed a very exacting vision of Notification Center, one ostensibly made to keep the iOS experience consistent, to trump the actual user convenience of having a particular kind of tool within it.
It's this maddening, self-sabotaging dedication to singular vision at Apple that scares off developers from trying genuinely innovative things on iOS.
> It's this maddening, self-sabotaging dedication to singular vision at Apple that scares off developers from trying genuinely innovative things on iOS.
Developers do all sorts of innovative things on iOS, and minor constraints like this are unlikely to deter them. The "pull-to-refresh" action we all know and love first appeared in Tweetie for iOS -- it became so popular, Apple turned it into a first-class control. Same with the "hamburger menu" if I recall correctly.
Apple's claim that they're looking out for user experience might have been plausible in the early days of the app store but they've let so much shit through at this point it's hard to believe this is about anything more than then walling off yet another part of the OS for their exclusive use.
In any case I think anybody with half a clue can see that building a business on this kind of platform is extremely foolish.
I wonder if this is a litmus test for rejecting CurrenC from the App Store for duplicating functionality that's already provided by iOS (Apple Pay). Have a go at the little guys, one of them bites with a lawsuit and Apple gets to duke it out in a fairly low-risk legal game.
Apples wins => likely to block CurrenC (and has a legal precedent for upcoming challenge)
Apple rejects tons of apps everyday, sometimes for wrong or incorrectly interpreted reasons. Its a pretty goofy, imperfect process. Doubt they're trying to set any precedent.
The definition of insanity is doing the same thing over and over again but expecting a different result, is it not?
At the end the author asks if we shouldn't already be past this, but, what reason have you given them to change when you just keep using it despite the things you hate?
I don't like the way they do things either but at the end of the day, it is their app store.
Have you tried the Android ecosystem recently? It's not perfect either, but there's much more choice in it. Maybe it's time to give it a[nother] try?
I love PCalc and am a long time user, but the notification widget looks ridiculous and is way too big in the notification center. Is it possible that the majority of users here on HN know about the tiny edit button at the bottom of the notification screen but the vast majority of users dont? I can't recall if PCalc turns on the notification center widget by default but I recall not liking it at all and turning it off.
It is not possible to add to the notification center by default. Every user has to click edit and add the widget manually.
What makes this even worse is that apple is currently featuring this PAID app in the hand-picked app store section as a "best-of" for the new widget feature. That's brutal for both paying customers who will get a downgraded new version and for other developers with in-development projects inspired by this highlighted example.
Also, "don't do math" - how are developers supposed to understand that rule? Are all "ADD", "SUB", "MUL", "CMP" opcodes banned? How do you even do UI layout without "math"?
I wonder if this is simply a misunderstanding by an app reviewer.
It makes sense that Apple wouldn't want CPU intensive tasks to run in the notification centre so they put in the guidelines that the widget shouldn't do calculations. I.e. You can't have a bitcoin mining widget. And, unfortunately, the app reviewer has taken that literally because he or she has confused cpu calculations with human calculations.
This would be a best-case scenario, but according to the PCalc developer it seems the decision was made "further up the chain", he had already attempted to appeal.
Seems more likely Apple has decided (far too late in the game to not make some major damage) that they want Notification Center widgets to be information-at-a-glance-only...
It's also perfectly normal for people to react and condemn a business which appears to be behaving in either an irrational or bad-faith manner. Why is that so surprising to you?
"Legal" is not the same as "good idea" or "immune from criticism".
Because it's by its nature arbitrary. After all of these years (I've been developing since iPhone OS 3), they still have problems managing developer expectations.
Even if Apple is predictably bad about this, each instance of them bringing the hammer down is still a surprise, because that's the nature of the problem.
Then people should spend an additional minute thinking about what they are going to do and if that actually has reason for success.
Developing a calculator that sits in the NOTIFICATION area of an OS does not strike me as thought out. Especially seeing as how gigantic this particular one is.
Is it surprising to you that Apple has an issue with an app that makes a mockery out of their notification center by completely obscuring literally every notification in there?
And here I thought devs were smart enough to read "Notification Center with widget ability" and figure "Well, I guess I could put a notification widget in there" and not "Ooh, sweet. that's where my calculator goes"
This is kind of a surprise but most of them aren't, and are spelled out in the Dev agreement. Usually when you talk to someone at apple it makes a lot more sense than what's in the email, and the issues are pretty easy to address.
Oh, look, I don't have a single Apple product in my possession. I already voted with my money and this stupid crap doesn't even matter to me. You can do the same, too!
I honestly feel that there is very little bias in here. I downvoted the comment just like I would have if he had replaced Apple with Google or any other name.
I don't just mean the displayed precision changed, I mean the actual answer changed, fairly significantly.
In particular, if you accidentally tilted the phone in the wrong direction so that it switched to portrait then tilted it back, you'd get a different answer than if you held the phone steady.
I only found this bug while in the middle of a midterm for a graduate-level finance class...
Edit: details of one of the bugs:
Steps to reproduce:
1. Open the calculator application 2. Ensure the app is held vertically (non-scientific mode) 3. Enter a number such as 10,000,000 4. Attempt to make the number 10,000,000.1 by pressing "." and "1" 5. Notice that instead of this number, the calculator application converts the input to 100,000,001 - i.e. the decimal point was ignored.
This does not happen in scientific mode.
Bug ID# 6743558.