When I lived in London, people waiting for the tube were leas than an arm’s length away when it rushed in. Blew my mind. Eventually I got used to it and did the same.
The difference is that the tube driver (assuming there even is one) can't just decide to take a slight turn from the planned route, while the car driver really might.
It does happen, it happened to me — a driver was distracted (because they were texting while driving) and they veered from the traffic flow and I was hit at 40 mph. The EMS worker who responded to the scene told me two things: 1. They almost never pull people alive from this kind of accident, 2. This kind of head-on, no brakes, distracted driver accident was happening more and more. This was 10 years ago.
Yeah but look at how normalized it is. All you need to do is slightly misstep, out of the thousands of steps you take, to bump your head on a passing train. You think it’s not possible (part of the trance) but all the parameters are set up so that the slot machine eventually hits the jackpot. You’re 12 inches away from a projectile, and you are to repeat the loop daily for a lifetime. The odds are the odds.
On iOS you can swipe with two fingers to select multiple rows. One of the more hidden features. Mentioning it to show that we didn’t lose it everywhere.
As an indie app developer, this design update discourages me massively. The previous, minimal design gave the impression of being a platform, even though it was always mostly Apple stuff in Apple land.
The new design is so visually overwhelming that I think the only way for users to deal with it is to reduce complexity. I read a statistic that said the average user had 21 apps on their phone. I think that will reduce to 15 now, or less.
As for my app, this basically throws my whole design system out the window. I don't want to add glass to all my UI elements. Remember the visual noise that translucent window borders introduced in Vista? Why would I do that to my UI?
I like the fact that the new design introduces a sense of hierarchy, and that it has more animations. I also like that transition animations are now interruptible by default (watch the "What's new in UIKit" video for that). But that could've happened without the glass nonsense.
It was hard to feel excited in previous WWDCs, but I just took it as a sign of platform maturity. This year, on the other hand, is outright discouraging.
The author mentions code signing as a tricky ecosystem thing. Well, wait until the app has to get new assets to be resubmitted with bug fixes for new versions of iOS.
And if they don’t keep updating it, it will stop working and the buy it for life idea commits the dev to maintenance that isn’t paid for either upfront or through subscription.
There are folks who make and give away a lot but some learn their lesson quickly and find ways to get people to put a reasonable amount of recurring payment into the app to make it even remotely sustainable beyond a hobby.
FWIW, AI may make this cost so much lower that this kind of thing can make sense now. Something to consider, I suppose.
I think a lot of what Apple has added to their App Store in the name of security is really just to get developers to move to recurring revenue models, which in turns makes Apple a lot more money. An old app that hasn't been updated in a couple years isn't inherently dangerous, but if Apple can convince users that it is, then devs will have to rethink how they do business. It's a shame to watch Google copy this move too on the Play Store.
There are changes that can be valuable to users. Like the privacy attestations. I was not publishing at that time but I presume that was part of the mandatory dev march.
eg Apple publishes a new ios rev; debugging on it requires upgrading xcode; and that requires updating your OS.. Good thing you don't mind wasting a full day or more reinstalling every other devtool you may happen to use. Or xcode just doesn't connect to your ios device because reasons. (The reason is apple writes shoddy slapdash software because monopoly.)
So now you're spending another $1500+ on a studio so you can do all this in VMs and see how bad the damage will be to avoid blowing up your main devbox. etc etc.
This sounds like Python development....and React.....and Svelte....and oh, remember Angular 2 to 3? Or people who invest time in Clojure? Were you a fan of the MEAN stack a decade ago? Build something on it? How's that app doing?
What you're describing is not unique to Apple. It's a regular occurrence for anyone who's not writing for a enterprise SAAS company with a largely legacy codebase and a dozen DevOps guys mostly obscuring that stuff from you.
I can easily do those on my Late 2013 MBP running Catalina, but not iOS development for the App Store, because it has a minimum Xcode version requirement, which in turn requires a newer OS, which requires buying a new Mac.
> This sounds like Python development....and React.....and Svelte....and oh, remember Angular 2 to 3?
I can open a webpage written in Angular 1, or written in year 1990. I can run a program written for windows 95 on my new PC with windows 11. It’s normal to keep compatibility for compiled/finished ebd user software.
but, can you run npm install on your angular 1 codebase from 1990? or was it bower install ? remember grunt?
my 2012 mac hardware works perfectly fine, even the battery is OG, apple stopped supporting it, chrome won't give update on the last supported os.
software is fragile.
i constrain myself to html and plain vanilla js. if i have to use deps, they are local or hosted .js lib files, minified. d3 is a great example of this pattern.
I miss gulp, things seemed so much simpler back then in retrospect and the nature of JavaScript fatigue seemed to be the FOMO kind instead of this abstraction over abstraction that abstracts that other abstraction but you still have to the understand what it abstracted away kind we have going on today.
Will TypeScript go the way of CoffeeScript in the future? Who knows.
Nonsense. Upgrading python using standard tools does not regularly require OS upgrades. asdf, rbenv or various other tools will happily pull in new releases from their ecosystems.
Go requires macos from 2016 or kernel 3.2 which I think is over a decade old. I can't find any limit for jvm 25, though I'm sure there is one. No competently-built tool requires OS upgrades like xcode.
I had an idea years ago for an app but came to the same conclusion after some market research and a risk assessment so didn’t bother. I do not regret that decision for a moment.
I either have to put enough time into the idea to do it full time or do a shitty job. I can’t win either way without incurring massive risks so I will continue to part time two jobs and invest the earnings from those wisely instead.
There’s no code signing process or update review. There’s no $99/year registration fee for the privilege to have your code signed. There’s no up to -30% hit on your fees.
So yeah, there’s an easier way of avoiding the Apple specific pain points. If you think that I felt it would be easier to program the actual app, then you’re clearly just trying to be a hostile commenter and insinuating something about me.
I wasn’t talking about any of that. I was talking about the ease of earning a living as a solo developer on the platform. On iOS, it’s true that there are hoops you have to jump through and you have to share 15% of your fees (if you get to the point where you owe 30%, you are doing well!), but in return you get access to a fairly large audience of people willing to pay for apps and services.
The web is the Wild West and I don’t really have much of a feel for how difficult it is to find an audience that will pay to use the software you build.
It’s not just that. Companies who are anywhere between unethical and outright complete scams have discovered how incredibly easy it is to get people to sign up for a free trial for something that milks them of tons of money they don’t realize.
Despite the fact that Apple tells you when your subscription will renew it doesn’t seem to help enough people. So they buy a scientific calculator app because they don’t realize that it’s built into the one on the phone and then end up paying five dollars a week for it. And even if they find out after the first renewal that means the app developer got five dollars (minus fees).
There’s tons of subscriptions out there that are just completely out of whack with their prices. And Apple just doesn’t seem to care.
So Apple gets $99/year PLUS 30% app revenue share?
So at a $2.99 "fair" price point as mentioned in the post, how many copies does he have to sell to break even (assuming 4 days of development priced at $2,500/day contractor rate)?
That aside, 1) the author is not an experienced iOS dev 2) hourly rates != cost, and 3) you can certainly get the same app built for under $1000 by a freelancer.
You also seem to have accidentally used a $0.99 price instead of $2.99.
I thought it was 15% for devs selling under $1M per year.
It also seems like a super-weird analysis angle to both pay yourself a (very generous) full day rate AND then expect upside on top of that and conclude that making just the $10K for 4 days’ work is somehow a loss for the solo entrepreneur.
We're talking about an app made as a weekend project to solve a problem the developer themselves has. Most normal
humans would count their time as being free. I pity the person who considers the economic value and tax implications of their leisure time projects. So the calculation is just how many downloads do they need to make up for the $99 developer fee which breaks even at 47 downloads at 30% or 38 at 15%. Definitely not unrealistic.
$2500 a day?! I will do anything, including clubbing baby seals to death, for half that rate. Is there any place where you can make that kind of cash doing contracting work?
At 3 days to develop, the author could make 100 of these apps a year. Of course, they will probably spend additional time developing and fixing bugs, IF the app gets any traction. And if it becomes something more substantial, they could up the price or release a preium in app purchase to upgrade with new features.
true to that. App Store is flooded with junk. it is hard to get to users even with legit free good app. and marketing is unbelievably expensive. users trust into random apps on App Store is very low too. getting iOS apps to work is very very hard.
it should go without saying that development cost, and rest of operational cost (design, legal, etc.) should be considered as zero. meaning do yourself with no pay. with all money goes into marketing. only then there is slight hope to break even.
Yes, making money on iOS is an uphill climb, many times more so if you’re not playing the TikTok ads and subscription model.
I’ve been making iOS software independently for almost 2 years now (https://heliographe.studio) and am about ramen profitable.
A few notes in case OP (or anyone interested in making some money in the App Store) is reading:
- you have to make the app free to download, and quickly demonstrate value then show a paywall if you want any purchases. Paid upfront just does not work unless you’re an already recognized product.
I had some apps that were paid upfront, and would mostly get $0 days. Switching to free to download immediately brought me to a slow but steady trickle of daily downloads, and from there you just have to work on your conversion rate.
- but that's still going to be pretty low, if you want any meaningful user acquisition, you're going to have to go look for the kind of people who might be interested in your product. The broader your potential audience is, the harder that's going to be (but that's why TikTok ads can work so well). In my case, choosing to focus on a somewhat niche area (tools for photography) is helpful; there's a strong photography community going on Threads and regularly posting on there yields good results (for now...)
- $2.99 is dramatically underselling yourself, especially if you offer a quality product that you put time to craft to your standards and has no tracking, no subscription, no ads, etc. You should play with pricing to see what the sweet spot in terms of conversion is, but in my experience it's always worth it to start at least at $4.99/$7.99 for these sort of utility apps. Of course, the design of your funnel/paywall will make a huge difference (ie you'll likely sell more of an app marked as $4.99 at 50% off, than just $4.99)
- learn about what makes for good App Store screenshots, descriptions, how keywords work, etc. Ariel from App Figures has some good videos on YouTube about what they see and what seems to work based on their data.
The days where you could make a little app, chuck it on the App Store for $.99, and have it just blow up are well over. If you want to make any money on the App Store (even if to just pay back for your Apple Developer membership), you have to put as much effort, if not more, in the marketing and promotion of your product than you put in the design & development of it. It's a grind for sure — and don't count on Apple to help you in any way (by and large they seemed more interested to promote games and dating apps with $49.99/mo subscriptions than small indies doing interesting things).
LCOE assumes that electricity generation is comparable. However, renewables have a high variability, which puts a much higher load on the grid.
The grid investments are sizeable. You not only need to add a lot of batteries, you also have to make other investments, for example to add moment to the grid, because unlike big turbines like nuclear, water or gas, solar or small wind turbines have almost no moment of inertia, which was one of the problems behind Spain's power outage.
This isn't new stuff, it's all solvable and countries already do this; the power outage of Spain would've been impossible in Germany for example. It's just important to highlight that with old-school power plants, you don't need a lot of that stuff to stabilise the grid. You need to include the grid costs when calculating the true LCOE, which most of these charts, including the Wikipedia one, don't do. Wikipedia isn't lying about that; they outline this very fact as one of the key weaknesses of the LCOE metric.
When grids set up markets to let people compete to provide those grid balancing services batteries totally dominate. Which suggests that is just another area that modern renewables win and reduce costs.
On the other hand Nuclear LCOE generally assume they can sell a high proportion of their power for the next 40 years.
So really the big hidden assumption is that solar won't eat half their market in that timeframe. And then solar plus batteries eats into it further. Which would drive up their cost, letting solar plus batteries win more business in a vicious cycle.
With the recent Iberian power situation half their nuclear was offline because they were already in a huff because they weren't being paid enough money.
exactly that is why we need to put electrolyzers into mix, that way we can have load on nuclear and on renewable at same time, we need insane amounts of hydrogen for chemical industry anyway.
also chemical industry needs big investment because they have to change mindset about procuring carbon and other inputs in net zero economy.
(similar thing as what some datacenter companies say they do with buying nuclear capacity)
i think most people do not want to understand, they want socialized grid payed by citizens instead of putting real energy price into pricing for goods/services.
hydrogen as energy carrier, not as transportable comodity. hydrogen as MEANS not as a goal. we need iron + water to have from 20 sec up to seasonal storage of energy in megawatts per meter cubed...
99.9999% of people do not understand that point about inertia, you are one of them. and no it was not problem with spain, problem with spain was badly selected and configured inverters. if someone says otherwise he is liar.
" with old-school power plants, you don't need a lot of that stuff to stabilise the grid."
you just need proper sizing of renewables inverters + firmware update...... so no you do not need to have inertia of huge mass in turbines. also 99.999999999999999999999999999999999999999999999999999999% of all problems stem of peoples need to regulate grid to flat line for nonsensical reasons, IF you have slight artificially made "fluctuations in grid" which are generated by all inverter synchronized and planned ahead, there is no problem. grid has to have "pulse". THAT is decentralized / new grid. what you are describing is Stanley/Westinghouse grid. so mixing is resulting in nonsense.
This is why carbon tax is so important. The moment you tax carbon emissions, natural gas no longer wins.
Assessing the right amount of tax is non-trivial, but approximations for the costs of carbon emissions exist, and even conservative estimates push natural gas out of the profitability zone.
I disagree. The "right amount" is pretty trivial. Structuring it such that it's politically favorable, precise, and robust is the actual issue.
There's a general tension between precision and robustness in law, and finding that tradeoff is often the difficult part once public opinion is secured.
> right amount is the amount that drives net emissions to zero
On what timeframe? Put it too soon and you risk social upheaval crashing the whole project. Put it too late and you cause unnecessary damage to the climate.
That's debatable. I haven't looked at the numbers, but surely there's some amount of net co2 emissions that doesn't have a significant impact on the climate.
The current global CO2 emission rate is about 200x the average emission rate during the Permian/Triassic mass extinction. Just slowing down these emissions buys time, which is useful and valuable, but is not itself a solution. Net CO2 emissions have to be driven effectively to zero or we're just delaying disaster.
Also, "net emissions" that include CO2 being absorbed into the oceans must take into account that oceanic absorption declines as surface waters become more saturated with dissolved CO2 and the pH declines. The rate of carbon flux to the deep ocean (from descending organic detritus) is just a few percent of annual emissions.
German has a word for second hand embarrassment. Fremdschämen. Comes in very handy here. If Sam continues like this it won’t be long until it becomes part of the regular English language like other German words such as Kindergarten.
And I’ll be happy that I don’t have to explain Fremdschämen anymore. Everything has its upsides.
When I lived in London, people waiting for the tube were leas than an arm’s length away when it rushed in. Blew my mind. Eventually I got used to it and did the same.
reply