Microsoft and Google collaborate to make PWAs better • Hacker News Board

Re: Microsoft and Google collaborate to make PWAs better

#181

Post by dmitriid » Sat, Jul 11, 2020, 3:38 PM UTC

osrec wrote: post #81

dmitriid wrote: post #71

Earlier quoted context omitted.

> 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. 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…

Why, then, did it take safari so long to implement ServiceWorkers? I believe they were the last of major browsers to implement it. And they landed the feature more than 2 years after Firefox! 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…

> Why, then, did it take safari so long to implement ServiceWorkers? > Also, why was Safari's IndexedDB feature released

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.

Re: Microsoft and Google collaborate to make PWAs better

#182

Post by osrec » Sat, Jul 11, 2020, 3:55 PM UTC

dmitriid wrote: post #181

osrec wrote: post #81

Earlier quoted context omitted.

Why, then, did it take safari so long to implement ServiceWorkers? I believe they were the last of major browsers to implement it. And they landed the feature more than 2 years after Firefox! 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…

> Why, then, did it take safari so long to implement ServiceWorkers? > Also, why was Safari's IndexedDB feature released 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…

What about Firefox? Is Safari's team smaller than Firefox's as well?

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).

Re: Microsoft and Google collaborate to make PWAs better

#183

Post by alwillis » Sat, Jul 11, 2020, 7:01 PM UTC

vially wrote: post #63

alwillis wrote: post #59

Earlier quoted context omitted.

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. 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 t…

> [...] why they don't allow slower, less power efficient, less secure and less private web engines available on iOS Do you have a source that shows how chromium/edge/firefox/etc are inferior to safari in this regard?

For pure speed of the web + Javascript engines: "Introducing the JetStream 2 Benchmark Suite": https://webkit.org/blog/8685/introducing-the-jetstream-2-ben...

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/

Re: Microsoft and Google collaborate to make PWAs better

#184

Post by dmitriid » Sat, Jul 11, 2020, 7:26 PM UTC

osrec wrote: post #182

dmitriid wrote: post #181

Earlier quoted context omitted.

> Why, then, did it take safari so long to implement ServiceWorkers? > Also, why was Safari's IndexedDB feature released 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…

What about Firefox? Is Safari's team smaller than Firefox's as well? 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…

> Is Safari's team smaller than Firefox's as well?

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.

Re: Microsoft and Google collaborate to make PWAs better

#185

Post by davidpolberger » Sat, Jul 11, 2020, 10:45 PM UTC

I was disappointed to learn that https://pwabuilder.com can't process the PWAs that are produced by our app builder Calcapp, erroneously claiming that our PWAs have no manifest.

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.

Re: Microsoft and Google collaborate to make PWAs better

#186

Post by nikitaga » Sun, Jul 12, 2020, 12:41 AM UTC

soraminazuki wrote: post #175

nikitaga wrote: post #155

Earlier quoted context omitted.

> respect the conventions of the platform they're building for 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.

This is a false dichotomy. People can dislike PWAs and be critical of AppStore practices at the same time.

People "can" do that but that's not what either the commenter or Apple is doing.

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.

Re: Microsoft and Google collaborate to make PWAs better

#187

Post by soraminazuki » Sun, Jul 12, 2020, 6:23 AM UTC

nikitaga wrote: post #186

soraminazuki wrote: post #175

Earlier quoted context omitted.

This is a false dichotomy. People can dislike PWAs and be critical of AppStore practices at the same time.

People "can" do that but that's not what either the commenter or Apple is doing. 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.

> People "can" do that but that's not what either the commenter or Apple is doing.

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.

Re: Microsoft and Google collaborate to make PWAs better

#188

Post by osrec » Sun, Jul 12, 2020, 1:39 PM UTC

soraminazuki wrote: post #174

the_gipsy wrote: post #153

Earlier quoted context omitted.

> 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.

> If TikTok was a PWA it couldn't have done a fraction of the tracking it was doing. 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…

Misleading how? You have one platform that already allows crazy privacy violations natively, and you have another platform that you suspect might some day allow similar levels of privacy violation, yet you champion the privacy stance of the first platform.

Makes no sense to me.

Re: Microsoft and Google collaborate to make PWAs better

#189

Post by soraminazuki » Sun, Jul 12, 2020, 3:49 PM UTC

osrec wrote: post #188

soraminazuki wrote: post #174

Earlier quoted context omitted.

> If TikTok was a PWA it couldn't have done a fraction of the tracking it was doing. 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…

Misleading how? You have one platform that already allows crazy privacy violations natively, and you have another platform that you suspect might some day allow similar levels of privacy violation, yet you champion the privacy stance of the first platform. Makes no sense to me.

> Misleading how?

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

Re: Microsoft and Google collaborate to make PWAs better

#190

Post by nikitaga » Mon, Jul 13, 2020, 7:03 AM UTC

soraminazuki wrote: post #187

nikitaga wrote: post #186

Earlier quoted context omitted.

People "can" do that but that's not what either the commenter or Apple is doing. 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.

> People "can" do that but that's not what either the commenter or Apple is doing. 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 respo…

There's a difference between merely "disliking" something as a personal preference and insisting that other people should not have the things you dislike.

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.