Microsoft and Google collaborate to make PWAs better
medium.com
medium.com
PWA collaboration seems to be a step in the right direction, especially since Edge is using chromium and Microsoft seems to be building great PWA apps[1].
Although I'm not entirely sure about Google's end game with PWA[2], I hope these kind of collaboration motivates industries to build more PWA apps and put pressure on Apple to properly support PWA APIs in Safari; then may be, just may be one day we'll have fair grounds for new mobile operating systems to compete without this Playstore/Appstore duopoly.
They need to stop doing this, or they will hurt themselves in the long run. As it is, most people use Safari on iOS because they are forced to - all other browsers on iOS are forced to use WKWebView, which basically means they are wrappers around Safari's view.
To this day, I can't understand why their users stand for this. In this regard, Apple reminds me of a cult leader brainwashing their followers into believing the cult's way is the best way. Makes me a little sick actually.
If the option is Android + first-class support for PWAs vs. Apple with limited support for PWAs, I’ll go for the latter every time.
As an iPhone user I actually like that all phones use the same engine. It prevents the “this website is better viewed in <browser>” pop up from affecting mobile users (and in particular it forces google to support a browser other than chrome). In other words, the engine monoculture means that making a site work for any iOS users means it will work for every iOS user, which I as a user really appreciate.
Are you sure you aren’t the one being brainwashed?
Why should anyone trust Google, since they make virtually all of their revenue from ad tech, which means tracking users across the web, even when they don't want to be. Advertisers are Google's customers, not the users of their browsers or operating systems.
And since Apple makes virtually all of their revenue from selling devices and services, who's more credible on the issue of privacy?
If we are to believe Apple cares about its users, then it makes perfect sense why they don't allow slower, less power efficient, less secure and less private web engines available on iOS.
"But shouldn't this be up to the user?" some will say ask. Yes, but to a point. There are already so many ways on the web to shoot yourself in the foot. Apple is making it clear that if you want a platform where there are no safeguards, then this isn't the platform for you.
There are certainly a tiny, vocal minority of iOS users who complain a lot about wanting the ability to shoot themselves in the foot if they so choose. But Apple knows those aren't their core users and their revenue show that.
They need to stop doing this, or they will hurt themselves in the long run.
People have been saying this for many years and now they're valued at over $1.5 trillion during an economic collapse and global pandemic. Please tell me when the lack of web APIs that would make it easier for their users to be tracked on the web is going to have any material effect on them.
Caring would mean giving people a choice. The real reason is they're pivoting into services and other forms of recurring revenue. They don't want people building platform-independent apps that don't need Apple approval and can bill outside of their 30% cut.
Do you have a source that shows how chromium/edge/firefox/etc are inferior to safari in this regard?
Anyone with a MacBook or MacBook Pro can attest to the fact their battery last significantly longer using Safari vs Firefox or Chrome. Apple has numbers on the Safari page: https://www.apple.com/safari/
You can run EFF’s Panopticlick test for what browsers block (or don't block) regarding fingerprinting. Safari blocks or disguises several features that are used for fingerprinting; Chrome blocks nothing: https://panopticlick.eff.org/
I'm not saying Google is incredible (my opinion is quite the contrary actually). However, I am saying Apple is holding the web back, in a fairly annoying way, for developers.
The fact that they are a rich corporate does not mean they always will be, or that they're conducting their business in the best way possible. There are plenty of examples of rich tech businesses that lost their market share very quickly. We don't even need to look that far into the past for an example: BlackBerry sticks out in my mind. I believe their stubbornness and resistance to change caused their downfall. I feel it may be a similar fate for Apple, and for similar reasons, unless they change their attitude towards the web.
Right. Because Google, who dominates the entire web standards spectrum, is known for creating well thought-out, well-designed APIs that take into consideration concerns and criticism, and thinking long-term of the web as a whole.
Oh wait, they don't. Up to and including the point where both Mozilla and Safari are saying they won't implement certain functionality because of glaring holes in the design, and Chrome not only shipping the features, but making them GA.
Oh, and yes. Most of the "Safari is holding back web developers" talk is bullshit, of course. It actually should read "Chrome releases APIs at neck breaking speed with zero concerns for the future of the web." See web API counts in major browsers: https://web-confluence.appspot.com/#!/confluence
Also, why was Safari's IndexedDB feature released nearly 4 years after the other major browsers? And when released it, why was it full of bugs that simple test cases would have caught?
I don't want to fuel conspiracies, but the pattern is fairly obvious. Apple appears to resist progress in Web APIs.
Priorities. Safari's team isn't as big as Chrome's.
> but the pattern is fairly obvious.
The only obvious pattern is the link I provided. And things like this:
- Chrome releasing APIs even when other vendors are against: https://twitter.com/Rich_Harris/status/1220412711768666114
- Chrome sabotaging competitors with shady practices: http://archive.is/tgIH9
and there are many many many more examples of this behaviour.
Looking at this page: https://webkit.org/team/ and comparing it to https://wiki.mozilla.org/Firefox/Team/whois would suggest otherwise.
Do you have numbers from elsewhere to back up that claim?
Btw, I'm not saying Google are right to power through with poorly designed APIs either, but when there is a standard in place, why Safari takes sooo long to implement it is perplexing (or perhaps just a rather obvious strategy).
Unknown
> but when there is a standard in place, why Safari takes sooo long to implement it is perplexing
Some standards are so bad, even Chrome ends up deprecating them (see Custom Elements v0).
Or they just go out of favor (see HTML Imports).
Or they are so poorly specified that even years after "becoming a standard" they have things like "this section is not specified yet" in their texts. And the status of these "standards" is often not above "Candidate Recommendation" (that is it is gathering implementation experience, and yes, that refers to Service Workers).
BTW Service Workers were moved to Candidate Recommendation about half a year ago: https://github.com/w3c/ServiceWorker/issues/1460 ("Double check everyone is happy with v1 going to CR.")
So yeah, I don't know what to say. Safari isn't chasing every API under the sun? Good for them. Web Devs assuming that whatever's in Chrome is the be all end all standards on the web? Bad for the web and for the devs. Apple possibly not prioritising Safari? Bad for the web as well.
I feel this is being dismissed by many people, but there's a lot of truth in this statement. Here's an example of this at play:
https://twitter.com/Rich_Harris/status/1220412711768666114
Chrome devrels and cheerleaders pile on Safari and Firefox all the time for not adhering to Chrome's agenda and frame it as "holding the web back." But the true thing that holds the web back is a browser monoculture.
The android ecosystem demonstrates that freedom isn’t necessarily creating a materially better ecosystem.
Blackberry was a different problem — they saw themselves as a messaging platform that did some other stuff.
Nope, Apple is taking care of it's customers and protecting them from Google's ad exploitation.
"I feel it may be a similar fate for Apple, and for similar reasons, unless they change their attitude towards the web"
On the contrary, if Apple allowed this web Wild West that Google is advocating, then the users would have incentive to switch. This move from Apple actually ensures that things run smoothly on Apple devices.
It is not about "power for web developers", it is web developers being lazy to self-educate and learn other stacks.
There is skill called "desktop development" and it does not entail Electron.
Luckily Apple is not gonna let that happen with the mobile apps and the PWA.
You have the likes of Tiktok on there, collecting way more data than a PWA ever could. Do Apple not care about that?
Feels like a bit of a double standard on privacy, explicitly aimed at stifling the web.
Apple is "protecting" its customers by restricting and patronising them like little children. Privacy by default is good, but many of their restrictions are clearly aimed at rent seeking.
Where customers really need protection, Apple is stabbing them in the back by aiding and abetting human rights abuses with their pointless side-loading ban and their willingness to let dictators run their own iCloud service.
We don't need this whataboutism to evaluate Apple. You're making the argument for PureOS or Debian, not iOS.
> If we are to believe Apple cares about its users, then it makes perfect sense why they don't allow slower, less power efficient, less secure and less private web engines available on iOS.
No it doesn't. They could make a better browser and, if it was actually better in every way, nobody would have any reason to use anything else. The fact that people would given the choice implies that something about choosing differently makes them better off enough to switch from the default.
> Apple is making it clear that if you want a platform where there are no safeguards, then this isn't the platform for you.
Safeguards mean warning someone adequately when they're about to do something suspect. Outright prohibiting the user from making an informed choice isn't a safeguard, it's a cage.
> People have been saying this for many years and now they're valued at over $1.5 trillion
Exxon was at one point the most valuable company on the exchange as well. What does that tell you about the ethics of how they got there, or what their future prospects look like?
If you are going to be fear-mongering based on vague assertions and appeals to business models, it would be wise to mention the 30% fee on all digital transactions that Apple extorts from all its apps. It is very clear that Apple doesn't want to lose that control (and revenue) to the Web, which despite the attempts of all corporations (including at times Google), still manages to be a little more open than our corporate overlords would like.
The privacy argument is without basis, imo. Google is terrible on privacy too, and somewhat evil, and its browser is terrible at managing memory (I say this to prove I'm not a Google fanboy!) but at least it's not stifling innovation on the web.
Woah![1]
> "But shouldn't this be up to the user?" some will say ask. Yes, but to a point. There are already so many ways on the web to shoot yourself in the foot.
That sounds exactly like what a dictator would say when asked why the citizens are not allowed to vote.
[1]https://googleprojectzero.blogspot.com/2018/10/365-days-late...
My iOS device has never once shown me "this website requires Chrome." I get to use the mobile web without needing Google. And Apple is the only reason why.
Maybe some day, web apps will become so undeniably awesome that Safari's position is untenable. But the agitation is not coming from iOS users missing out on awesome PWAs. Instead it's web developers, who are just annoyed that they have to actually respect the conventions of the platform they're building for.
Why would any PWA say that, if Safari had proper support for PWA?
>Instead it's web developers, who are just annoyed that they have to actually respect the conventions of the platform they're building for.
Say the developer is from a village in India or Nigeria who has just a Raspberry Pi to develop apps -
•What if the developer didn't have thousands of dollars to invest in new equipments just to develop for iOS?
•What if the developer wasn't able pay $99 every year for the Apple developer license?
•What if the just didn't want to give 30% cut to Apple from the revenue each time?
For all the times Apple uses 'Equality' in its marketing, the Apple platform & ecosystem restrictions clearly seems to be aimed at the privileged, pushing the agenda that 'Privileged is good' and 'Everyone should aim at being privileged'.
Less empathetic would respond, 'Well that's not Apple's fault that they aren't privileged'; Well then, we have to go back into colonial imperialism & American capitalism to prove why they aren't privileged in the first place and why Apple holds a responsibility every time it uses 'Equality, Human etc.' in it's marketing; which would be digressing from the topic.
For the same reason that far too many desktop websites say that today.
> Say the developer is from a village in India or Nigeria who has just a Raspberry Pi to develop apps
Apple is not catering to this developer demographic.
> For all the times Apple uses 'Equality' in its marketing, the Apple platform & ecosystem restrictions clearly seems to be aimed at the privileged
You're conflating developers with users. What fraction of iPhone users are writing apps? If you wanted to prioritize users, there's much tougher questions you could ask.
As an end user, I dread the days PWAs ever go mainstream. They’re horribly inefficient from a power and performance point of view, and never have I come across a PWA with a user experience half as good as native apps (especially on mobile). Until they’re indistinguishable from native apps in terms of user experience, I want nothing to do with them. Beyond that, there’s some very serious privacy concerns as well with PWAs, especially with developers abusing APIs for fingerprinting
If TikTok was a PWA it couldn't have done a fraction of the tracking it was doing. The irony is that apple is forcing you to use a much more privacy-invasive alternative to PWAs.
This is misleading when the motives of PWA is to make the web browser acquire as much capabilities as a native app. If browsers keep on expanding access to the underlying platform without any regards of the consequences, TikTok would be able to track you more without you ever having to install anything. This would kill the web as we know it because you’ll have to give websites the same trust you give to software when you install them on your computer. We won’t be able to browse the web freely as we do now.
Makes no sense to me.
I pretty much explained everything after the first three words of my comment.
> You have one platform that already allows crazy privacy violations natively,
Give me a single example of an OS for a personal computing device where you don't have to trust what you install.
> and you have another platform that you suspect might some day allow similar levels of privacy violation
No, this isn't speculation, this is the inevitable future of going the PWA route. If browsers expand access to a platform that, as you put it, "allows crazy privacy violations," then it's bound to allow similar levels of privacy violations, plain and simple.
One of the best things about the browser runtime is that it has (or perhaps, had) pretty good isolation from the host platform [1]. This is great for browsing because you can click on a link without thinking twice about it. When that isolation is rendered meaningless, you really have to stop and think about whether you trust the website you're going to visit, just like you would when you install native apps on your device. So get ready to kiss bye-bye to visiting a random blog you saw on HN.
> yet you champion the privacy stance of the first platform
You've made that one up. I explicitly mentioned that you have to trust the apps that you install.
Please note that I'm not hating on PWAs for the sake of it. I just don't want to sacrifice the privilege of being able to browse the web freely because of it.
[1]: I know, there are bugs in the implementation, but they're relatively costly to exploit in a meaningful way, especially at scale
Define "pretty good". Compare the permission model of the browser against the appstores. Think about what their manual review process actually is for 99% of the submissions.
> When that isolation is rendered meaningless
PWSs don't and won't do that. That's your whole argument and it's misleading.
It's a well-known fact. Actually, I'm pretty sure you're aware of this, considering this statement of yours:
> If TikTok was a PWA it couldn't have done a fraction of the tracking it was doing.
As for this:
> Compare the permission model of the browser against the appstores.
What are you even talking about? It has nothing to do with browser isolation.
> PWSs don't and won't do that. That's your whole argument and it's misleading.
That's an interesting assertion but you haven't left me any clue why you think so.
I also notice that you chose to quote a conditional sentence out of my argument, claiming that it was my "whole argument," but I'd also like to add that I did explain the hows and whys that lead it.
> the motives of PWA is to make the web browser acquire as much capabilities as a native app
This is misleading. To get to the point, your argument is that PWAs strive to be just like apps, but run in the browser instead of the OS. This is simply not true.
If you want back the web of old times, just documents and links, then you can pick up any one of such initiatives. Reality however is that PWAs are a successful model: just go to a URL and use it. It runs on a protocol, not some proprietary platform, which gives users enough control (e.g. tracking), while being useful.
I'm confused, because it's supposed to be PWA's major selling point. Google says as much in their documentation [1], as do many commenters in this thread.
> If you want back the web of old times, just documents and links, then you can pick up any one of such initiatives. Reality however is that PWAs are a successful model: just go to a URL and use it.
Don't get me wrong, I don't hate web apps in general. On the contrary, I happily use them on a daily basis. However, I don't like the direction PWA seems to be heading. Acquiring more "native" capabilities is one thing, though I won't be going over this again. Another important point is that it doesn't fit quite well with the "browsing" aspect of browsers. Web apps are nice, but they're hardly the only kinds of websites I visit. I visit quite a lot of random websites every day, many from HN, and I want almost none of these websites spinning up Service Workers or asking me to turn on push notifications. Web apps being more capable is nice, but I don't want them at the cost of a more spammy browsing experience.
I do want the web to be more capable, but careful design and consideration is equally as important when it comes to making the web a better place. I feel that many accusing some browsers of "holding the web back" are overlooking this point.
Re: "holding the web back" 1. Not having push notifications on iOS _is_ sadly a dealbreaker when chosing between creating a native app and a webapp. 2. Android has had this feature for over a decade I think, and it works fine. 3. Apple gets a huge revenue cut from all native iOS apps, and has a general "platform image" it defends.
So there _is_ a huge contradiction, in that apple does obviously hold some key feature back, in the name of privacy, pushing users to use their preferred much more invasive platform.
Many didn’t understand what a web app was until recently and now they want nothing else. For starters the speed and ease of deployment can also impact the user experience; people love seeing problems fixed and tweaks made on the fly in reaction to their feedback.
But more than anything they’re thoroughly impressed with the interfaces compared to other shitty desktop enterprise software they use. At the end of the day it’s the same story as with all software: quality often depends more on craftmanship than on the flavor of tool that you’re using. PWAs offer enough means to build beautiful and performant software, with full offline capabilities that can make them nearly indistinguishable from a native app.
I can’t disagree that native apps have advantages in some use cases. For the most part I wouldn’t build something like a graphics editing program as a web app - companies like Figma challenge this, but they’re using techniques at the edge of web dev that go beyond PWA capabilities (eg, WebAssembly), and unless you have an all-encompassing product strategy like they do you’d probably be better off building something like that native.
For many CRUD app use cases however PWAs can be more than sufficient. I’m very bullish about their future, and I wouldn’t be if I hadn’t seen real, happy users, because that's what matters to me the most.
Do you think native platform APIs are more restrictive?
Why would your iOS device ever need to show such a message, unless it's implemented by the website itself? I don't understand the scenario you're describing at all.
Apple only allows the Safari browser engine to run on their phone.
So Google (and other web developers) can't do what they do on other platforms which is only test their website on Chrome or block users of other browsers.
For example, Firefox on Android gets a crap Google search from like 2005 unless you spoof your user agent and it works perfectly.
By that mandate has kept complete control over the browser space from Google.
If a website doesn't work in Safari, it just doesn't work. That's it. I do know that devs have to work A LOT harder to get things to work in Safari when compared to Chrome or Firefox. It's like Safari is the new Internet explorer.
Do you mean UI conventions, or the "30% revenue cut and a gag order on even mentioning non-appstore signups" convention?
Justifying blatant anticompetitive and user-hostile practices by native UI holyness is getting kinda old.
And it's a package deal anyway. As long as PWAs do not provide viable competition to App Store on iOS, you can criticize App Store till you're blue in the face and nothing will change. So if you choose to force native UI holyness on others, at least take the responsibility for what else this brings as well.
It’s not even remotely fair to pretend that HN commenters and Apple share the exact same interests. It’s not anyone besides Apple’s fault that they are doing what they’re doing in the AppStore.
> So if you choose to force native UI holyness on others, at least take the responsibility for what else this brings as well.
Again, what responsibility? The idea that anyone who dislikes PWA has to take “responsibility” for Apple’s actions is plain ridiculous to begin with.
Another thing I’d like to mention is that there are also reasons for disliking PWAs other than worshipping “UI holiness.” Honestly, I don’t care as much for that.
The main issue I have with the pro-PWA agenda is that it’s pushing for browsers to expand access to their underlying platforms. If browsers keep on doing this, we’d have to give websites the same trust we give to apps when we install them on our devices. This would mean the end of browsing as we know it because we’d have to think twice before choosing to click on a link.
Neither the commenter I responded to nor Apple are simply "disliking" PWAs. Apple is not allowing PWAs to bolster their monopolistic moat, and the commenter is supporting Apple's decision to not allow _other people_ to build & use PWAs on iOS, seemingly just because they personally prefer native UX. Both of those are far beyond just "disliking".
This might seem like mere semantics about what "disliking" is, but while simply having personal preferences is fine, taking _actions_ that affect _other people_ or publicly expressing support for such actions necessarily comes with some moral responsibility for those actions I think. Perhaps some puny PWAs and their market effects aren't important enough for people to consider that, but the principle is all the same I would think.
Yes it is. Because you have cherry-picked a single word out of my whole argument to make your point. I dedicated an entire paragraph describing the undesired consequences of going the PWA route and you refused to acknowledge it, claiming it's only the "UX" people are taking issues with.
Not including your reason in my question does not make it a false dichotomy. I never stated that UI conventions were the _only_ reason to dislike PWAs, it was simply the only reason the commenter implied.
If _you_ think promoting walled gardens and Apple's monopolistic power is less of a problem than expanding browsers' scope of functionality, I disagree, but that's not what I came here to argue about, that doesn't have anything to do with my original comment that you have an issue with.
Let's be real though, Apple didn't enforce this to save user's the small annoyance of 'use browser X' popups. Instead it gives them full control over what web APIs are available on iOS devices, so they can avoid implementing those APIs that would impact on their App Store revenue. They have a clear financial benefit in not supporting the PWA technology that would move the 'simpler' apps off their app store and into web - where they wouldn't receive their developer license fees, a cut of paid apps and a cut of paid subscription fees (outrageous). Also, I believe it was Steve Jobs who first shared the vision of web applications on iPhone “that look exactly and behave exactly like native apps” ?
> Instead it's web developers, who are just annoyed that they have to actually respect the conventions of the platform they're building for.
Developers already get around this to some extent with cross platform tools such as React Native, Xamarin, Flutter. Native apps will always have a place, PWAs are not going to replace sufficiently complex app experiences, much like on Android or Windows.
Part of Apple's success is their extreme stubbornness. They decide something and make zero compromises. The consumer tech industry is forced to fall in line with whatever they want; they have final veto power over all new consumer tech basically.
The purpose of the App Store was to ensure that Apple got a cut of every dollar flowing through the store. Gatekeeping quality requires more than blocking developers from uploading fart apps.
Yeah, it's disgusting.
> The purpose of the App Store was to ensure that Apple got a cut of every dollar flowing through the store
I think there were competing internal forces: a desire to milk the iPhone's success, vs something akin to an Apple retail store experience. Sadly they gave us the worst of both worlds: they freely admit shit, yet pick fights with good stuff, like Basecamp Hey mail.
It makes no sense why Apple would choose to defend this hill. Either make it a Zen garden or turn it to a web-style free-for-all.
I downloaded a simple HIIT (high-intensity interval training) timer app the other day. It's just a glorified stop timer that automatically pauses at intervals so you can do exercise-rest cycles.
I settled for the second or third most popular alternative since it was the first one that didn't have a HIIT-timer-as-a-service business model (WTF... no, I don't want to create an account for a timer app).
Still a disgusting experience. Before you start an exercise there's an obnoxious 10-second video ad playing at full blast. The interface is atrocious, too.
The app was free, so I can't complain on that front, I guess. I would pay $3-$6 flat for something like this, but herein lies the problem: I can't even find such an alternative. The results in the app store search were filled with garbage (SaaS as I mentioned, irrelevant crap, bloated "fitness assistant" apps, etc.)
I could definitely conceive this app being built as a PWA and being 100x better than many of those. For this use case I couldn't care less about PWA vs. native, both could work if done well; the point is that I really don't see what Apple's supposedly strict standards have done to protect my user experience from greedy, incompetent or careless designers and developers.
At some point their hardline stance will backfire.
Firefox on android does a great job of notifying and offering to install PWAs, but on desktop seems to pretend they don't exist. The bug report for adding any PWA support on desktop has been languishing for years. And as far as i can tell, the ubuntu software center team doesn't believe that PWAs exist at all.
it seems crazy to me that there's no functionality built into ubuntu that allows me to add youtube to the app list.
Firefox is currently unusable on mobile :/
Chrome has a few nice things going for it, but i'd rather have ublock.
Major websites wouldn't work properly (Twitter) and sharing images directly from a website (memes are my bread and butter) also didn't work.
I use Chrome with NextDNS on Android.
On Android, Firefox is the MOST usable browser.
I'm car shopping and downloading specs from multiple dealers straight out doesn't work - the downloaded PDF won't open, works like a charm in Chrome.
Stuff like this is why I suspect most non tech people would revert to Chrome pretty fast.
Maybe it's because I've got an "old" phone (4 years ish), but Chrome's insisting that I must download everything a website throws at my made it extremely difficult to use comfortably. As soon as I found out Firefox had extensions on mobile, I switched both my desktop and mobile browsers, and it's been great.
Now they seem to have given up on the promise of the web as an app delivery platform and have aligned with Apple on a pro-privacy stance that seems based more on posturing than safely providing capabilities.
To me, full support for PWA across all vendors' app stores would represent a revolution in the ability for hobbyists and small development shops to publish incredible applications.
Has HN not been published as a PWA to the various app stores yet? This seems like an obvious use case and could serve as a great example for other applications.
A progressive web application is a type of application software delivered through the web, built using common web technologies including HTML, CSS and JavaScript.
- Fast (cache with service worker)
- Reponsive (no explanation needed)
- Offline page (so if the user lost connection or the web site is cached and he visite when he's no connection the app should work or at least display a custom offline page)
- Engaging (fancy word, to say it's need to behave like an app + push notifications)
- Installable (if the browser is compatible and you visite the website more than once it'll ask you if you want to install the app. This will create a shortcut and when you click on it it'll open like an app (no address bar))
But Apple is very late and don't support all of this.
Really vague concept without strict definition.
That being said, how will PWAs get around the apple and google requirements that apps downloadable from their stores where they mention apps can't just be a website wrapped into an app?
Fair question. Google is implementing a new quality bar for PWAs in their store. See this post[0] for details.
Microsoft also is doing work in this area. We're not ready to share details, but will likely in the coming months.
With PWABuilder, our goal is to help devs build quality PWAs: ones with rich offline experiences, measurably great performance, and apps that are functionally rich. This implies something more than e.g. your resume packaged as a PWA. Towards that end, PWABuilder analyzes your PWA and gives you feedback on capabilities: encourages use of our rich web components[1], helps you build a great offline experience[2], and more. Whether your app is functionally useful (i.e. not just a brochure) is a quality bar left to app stores themselves.
[0]: https://twitter.com/ChromiumDev/status/1276223634550353920
https://developers.google.com/web/android/trusted-web-activi...
If PWAs would get their own permanent storage like other apps, it would be a huge step forward.
PWAs usually store client-side data as a kind of cache: cookies, local storage, service worker cache. Caches are not meant to be permanent.
There may be room for improvement here, however. One might imagine that clearing browser data shouldn't clear data for installed PWAs. That's an argument worth further evaluation.
Not being able to reliably store data on the device means that PWAs have to send the users data over the internet and store it on an external server. I would think users rather do not like that.
I am only talking about installed PWAs here. Of course not every website should be able to avoid having its cookies deleted.
Installed PWAs is another question. I suspect users will be surprised if they clear their browsing data only to discover they have to login again to Twitter, for example.
I may raise this question to the Edge Chromium team. I'm certain it's been raised before, but with Microsoft making PWAs first-class on Windows[0], this becomes a more prominent issue.
[0]: https://www.windowscentral.com/heres-how-microsoft-making-ed...
The problem is the data. Imagine a text editor where all your text documents are gone after you cleared the data of a different app, the browser.
I wrote a fitness app as a PWA some time ago and its pretty annoying to download all the instruction videos again every time you clear your "browser data". Plus all infos about which exercises you did and when is gone for good of course.
One way to address that issue is the native file system access APIs[0] coming soon, available today in Chrome Canary and Edge Canary. There, you'd be able to save your documents to the user's file system, just like a native app would. Those files will be exempt from any browser data clearing.
> all infos about which exercises you did and when is gone for good of course.
Might be good to store that data on the server. I realize not all PWAs have a proper backend server, but that sounds like a good candidate.
[0]: https://bugs.chromium.org/p/chromium/issues/detail?id=853326
Great, so there is light at the end of the tunnel!
> store that data on the server
Well, we talked about that already:
Persistent storage can help protect
critical data from eviction, and
educe the chance of data loss.
Persistent storage is not deleted by the browser,
even if storage is running low. It will only be
deleted if the user chooses to remove it via
their site settings.What could possibly go wrong with that?
Come on, stop giving people false hope, that API is like 3rd incarnation of the same thing by Google, always shot down by Mozilla and Apple. Even today if you restrict yourself to Chrome your PWA can use the older version of the same thing, but what's the point if that works only in Chrome?
"Why did all my data get deleted? How can I get it back?"
and
"Why is it saying the browser won't let me store more data?"
The user's desires are pretty clear in both cases, but it's very unclear how the user can get what they want.
i mostly agree, but the user experience isn't there yet. browsers make it easy to delete all data, but don't have any workflows that encourage deleting the data from only a single app. at this point, it seems less about empowering users to have control over their own data and more like browsers just don't care about PWA data.
kinlan is referring to this web API[0], which allows a PWA to store data persistently.
I see from the link that Chrome is already doing an "origin trial" even if the spec itself isn't complete. So, why would it be "released more broadly". Or is it the Constructible Stylesheets again? When both Safari and Mozilla said "no", but Chrome released them anyway?
The same crowd that now complains about Google, was worshipping Chrome, talking about how its its developer tools are, the multiprocess architecture, handling hundreds of open tabs (still to understand this one),.....
So if the Web is turning into ChromeOS, they only have themselves to thank for.
One of the issues is that storage in a PWA is done via localStorage or IndexedDB... both of which aren't really available outside of the browser.
Also, part of the benefit of a PWA is being able to be in mobile devices and 'storage' there is different than on the desktop.
Perhaps the solution instead is that we need a way to distribute natively compilable apps via other means, maybe through a website but also through other app stores not subject to the Google / Apple duopoly; so basically PWA level access and downloadability, but not made with web technologies.
Then just collect the references from them. :)
Flutter supports compiling to PWAs
All that would need to happen is Apple to allow sideloading apps. Then the only limit would be developer creativity.
Since when anything in Play Store (or Apple's App Store, for that matter) was discoverable? I can't find anything even remotely relevant to anything i want using these stores, i always have to use 3rd party sites (or just google something like "<something i want> for android", sometimes even adding a site:reddit.com at the side if there are too many results, just to see what others are suggesting) and then if there isn't a Play Store link, i have to write the app's name exactly in the store's search, otherwise it will never appear on the results.
Discoverability has been a joke on these stores ever since they got more than about 1000 apps - and they still are designed as if they only have about 1000 apps.
We've trained a generation of users to look for apps in app stores. I blogged about this here[0], but having your PWA in app stores means more users for your app.
[0]: https://debuggerdotbreak.judahgabriel.com/2018/04/13/i-built...
Discoverability is an issue too. If your users search for your app keyword there then you're more easily found
I and many others would love to see Firefox supporting installable PWAs. :-)
What's missing is the ability to integrate it with my OS app launcher, though I can get around that by adding my own launcher. Something like:
firefox --ssb="https://news.ycombinator.com"If you're referring to the PWABuilder analysis page, yeah, we hear your feedback. We're going to be revamping that page in the coming months. We'll deal with push notification score when we do that.
I dug into pwabuilder-lib and found the culprit: our manifest URLs use search parameters, which are silently stripped by pwabuilder-lib. Here's the GitHub issue I filed, including a proposed fix:
https://github.com/pwa-builder/PWABuilder/issues/949
Obviously, this wouldn't have been possible if PWABuilder hadn't been open source, so kudos for that.
Pwas are for apps wherein their functionality is entirely web based, and people just want an icon for it.
* is installed
* may run offline
* may use extra system resources
Now imagine how good it would be if there were competition on the desktop market.
https://www.microsoft.com/en-us/p/surface-pro-x/8vdnrp2m6hhc...
https://www.samsung.com/us/computing/windows-laptops/galaxy-...
Is it lower friction to start up your team's blog on Medium versus the unified blog platform larger Microsoft teams tend to use?
That said, we have been mulling building our own blogging platform using our own tools and web components for the sake of dogfooding.
Open standards are meaningless when only one cares about them.
This collaboration with Google took place because Google revamped their "PWAs in the Store" story. It used to be a very basic story: a native app wrapping a web view that loads your PWA. Those worked, but didn't support some modern APIs (e.g. service worker, IIRC), didn't share cookies and login information with the browser, and other problems.
Google revamped the story here to a more first-class approach for PWAs in their Store. It's called "Trusted Web Activities."
And earlier this year, Google created Bubblewrap, a command line tool to generate a proper Trusted Web Activity for publishing in their Store. So we wanted to take advantage of that new and better approach. That's really our reason for collaboration here.
https://developer.microsoft.com/en-us/windows/pwa/
The best part was that this way WinRT APIs are directly exposed as browser APIs, no need for hybrid solutions.
This is just the next step using the Chrome based engine instead.