1. 47

  2. 25

    So basically we finally arrived at the “make your app a web page” that Apple demanded when launching the iPhone

    1. 30

      Yes, and the latest trend in web development is to render content on the server. Everything old is new again!

      1. 7

        I think it’s better this time, because phones and network are fast enough that doing everything in the browser isn’t limited by UMTS speeds.

        1. 3

          The original iPhone didn’t even support UMTS (3G), it was GPRS (2G) EDGE (2.5G). A load of mobile providers who had already rolled out large UMTS had to go and deploy older hardware to support the iPhone without it falling back to GPRS. The latency on GPRS was awful (500ms RTTs were common, making it unusable for anything interactive).

        2. 2

          I have noticed this and had the very same reaction a few weeks ago.

        3. 13

          To be fair: when Apple announced this, React did not exist, Vue did not exist, precursors like Backbone didn’t even exist, and most critically, most of the technologies and tools we use in 2021 to do SPAs, let alone offline webapps, did not exist. Hell, I think the dominant offline storage solution was WebSQL, which was never standardized and is not (AFAIK) supported in any contemporary browser, and no equivalent of web workers existed unless you had Google Gears installed. You also had nothing like WebGL, or web sockets, or even widespread contemporary CSS that would make reasonable, cross-platform styling feasible. So what Apple was offering at the time was morally equivalent to having bookmark links on the home screen.

          (Yeah, I’m very aware of the pile of old meta tags you could use to make the experience be better than that in a literal sense, but that doesn’t resolve anything else I highlighted.)

          Speaking purely for myself, I found the initial announcement infuriating, not because I didn’t believe in the web (I did! Firefox was growing! Safari was proving the viability of KHTML! IE was on the decline!), but because Apple’s proposal was just so damn far from what doing that seriously would’ve actually looked like that it felt condescending. The Palm Pre, which notably came out two years later, was dramatically closer to what I’d have expected if Apple were being sincere in their offer. (And even there, webOS, much as I love it, is more an OS that happens to have JavaScript-powered apps than a genuine web app platform in the 2021 sense.)

          1. 5

            Even at the time, Apple’s stance felt to me like, “We aren’t finished with our native SDK yet and it’s so far from ready for public consumption that we’re going to just pretend it doesn’t exist at all.” I remember talking about the iPhone with my coworkers when it first came out and everyone just assumed native apps would be coming at some point.

            Even webOS (which I also loved) ended up supporting native apps eventually, despite having a much more feature-rich platform for JavaScript code.

            Games seem to be the killer app category that pushes mobile OS vendors to support native code. They’re one of the few categories of application where a lack of native code support can make an app impossible to implement, rather than just making it a bit slower or clunkier but still basically workable.

          2. 3

            Even Firefox OS was too early in the game for that (besides other problems of FFOS).

            1. 6

              If it was timed right, Mozilla would have found another way to run it into the ground. ;-)

              1. 1

                Could not agree more !

          3. 9

            Something they miss in the article and I don’t see in the comments here either is just how much a move like this limits the app’s audience size.

            I have a feeling most crustaceans would be SHOCKED at how many “end users” would peel off and walk away when confronted with instructions to add a web app to their mobile device’s home page in favor of just downloading a native app from the app store.

            1. 5

              Why not make a “Web Store” that helps with aggregation and discoverability of web apps that will link to the website. It will give more credibility to the web app and not some unwanted pop up that screams “Download me”.

              1. 2

                Since we’ve broached the topic of iOS feature requests, why not make a “unlocked bootloader” that helps with aggregation and discoverability of apps (and daemons) since there is no longer the concept of an “app store”. It will give more credibility to 3rd party developers and not some unwanted censor that screams “it’s for your security and privacy!!!one!11”

                1. 5

                  Ah, yes, because getting rid of a centralized, built-in way to discover new apps which (generally) are known to be more or less safe is a great idea for the average, often-times tech-adverse consumer.

                  I can think of no way in which this could cause a negative impact for app discoverability, make it more difficult for users to install applications, or increase the risk of downloading and running malicious code (which, before you say it: no, the average user is not going to know how to check to see if something is malicious).

                  This sounds like a fantastic idea.

                  1. 1

                    I’d agree, but the stores would need to do a better job of curation. As of now, there’s still a lot of grift and outright scams on the iOS app store - let alone the Android or Windows one. It undermines the message.

                    1. 1

                      Very true. The stores are all quite bad at curation, but I’d wager even a half-assed curation is better than none at all.

                      1. 1

                        You took my comment the wrong way. That’s not what I meant.

            2. 5

              Jokes on them, I’ve been making all my apps web only since i started development. It’s easier to maintain anyway.

              1. 11

                Plus I can just throw something together! I don’t have to get a Mac, don’t have to pay Apple $100, I don’t have to make my app go through app review.

                1. 3

                  Exactly, and it’s a lot easier to get someone to go to your website than download your app.

                  And if you really want it later, just use ionic.

              2. 5

                Definitely understand the motivation behind the move, but I myself have always enjoyed programming native apps (regardless of platform and distribution method) over web tech. The experience of coding in SwiftUI is genuinely infinitely better than any web framework I’ve tried (React, Svelte, etc).

                1. 6

                  Swift is a really nice elegant language, so I can appreciate what you’re saying, but man I really dislike XCode, and I REALLY dislike the fact that short of some work folks have done to make developing Swift server side apps on Linux etc possible you MUST code your apps on a Mac.

                2. 5

                  I’ve been feeling negative about PWAs since Firefox abandoned desktop support: https://www.fastcompany.com/90597411/mozilla-firefox-no-ssb-pwa-support I hope Mozilla changes their mind at some point.

                  But Firefox is still working on PWAs on mobile, and I keep forgetting how important mobile is to modern computing. Maybe it’s worth developing PWAs even if you only expect it to work on phones. Desktops are less locked down than phones, at least for now, so perhaps it’s more important to be able to escape from app stores on phones.

                  1. 4

                    The trick to use the web URL to create an app icon is a trick I use very often and really like. The big downside is that it doesn’t work so well offline (which is only useful for a few apps).

                    For example the one that I use the most is this one: https://thewinniewu.github.io/aeropress-dice/index.html

                    Creating a mobile app for this is just too much compared to a website that you can use across all devices.

                    1. 18

                      The big downside is that it doesn’t work so well offline (which is only useful for a few apps).

                      I’m sure it’s entirely coincidence that offline browser support is poor on platforms owned by companies that make bank clipping the ticket on app sales. /s

                      1. 4

                        On the whole it’s not poor at all, though.

                    2. 3

                      Thank goodness. I am so effing tired of needing to download yet another terrible app just to do a simple task (this week’s example: paying for street parking in a town I’m visiting). Almost all mobile apps aren’t worth the storage space they take up on my phone, let alone the fact that I trust my browser’s sandboxing much more than native apps.

                      1. 2

                        I trust Apple’s sandboxing well enough. But I 100% agree that random one-off apps should just be web apps. Or even just web pages. Why on earth should paying for street parking be an app at all? That’s a <form>.

                        1. 1

                          To keep your payment info safe, I guess.

                          Anyway, even Apple agrees that not all app need to be installed, and created AppClips. I’m sure Android has something like it too.