Ask HN: Who has had a successful PWA product?

They definitely clear localstorage. But they also are honest about doing this under disk pressure — can’t find the link right now though. My old company had a Cordova app and we stored the client token in localstorage and we found some 16G phones losing their tokens and it turned out to be this. We changed to store tokens in the keyring. From a real web app you can’t but I guess if you would use a cookie this might just work.

Yes, a full local reset. Its an issue that plagued us for almost a year sporadically across several thousand customers. After instrumenting our app with very detailed remote logging, we were fortunate enough to catch it on one of our in-house local test iPhones. The logging showed an app with rich local state, the app close, then several minutes later the app open with no local state. Everything was wiped. This lined up with the feedback we had been receiving. Specifically for us, the push notification subscription tokens were being erased/revoked from the browser, meaning the customer couldn't receive notifications anymore. But the logging showed all local state was reset, including IndexedDb. If you dig through the Apple developer forums and StackOverflow, you will see many others complaining about this. You will also see links to resources claiming that Apple ITP does not do this for PWAs, specifically:

https://webkit.org/tracking-prevention/#intelligent-tracking... "Home Screen Web Application Domain Exempt From ITP". This is 100% bullshit though.

My personal belief is that Apple is purposefully nuking PWAs from orbit in a non-deterministic and "buggy" way because they are a threat to the app store business model, and I suspect if they frustrate enough developers, people will stay away from PWAs.

I am not sure that I would be happy with a 'local only' experience actually backing up data to the cloud.

You can definitely say that about Tesla and most people don't care about substack

> Are you saying the Safari browser on iOS intentionally cripples web applications by randomly erasing / deleting browser data (cookies, localStorage, etc)?

I would absolutely make this claim.

This isn't even a PWA thing. We had a really simple series of boring web forms that suffered from this. The first iteration of this had the server swap an initial tokenized link for a cookie when clicked.

After some feedback from our customers, we discovered that iOS users were having a lot of trouble getting all the way through the desired workflow. We were able to replicate this on our iOS devices but it seemed completely arbitrary.

After abandoning cookies in favor of maintaining a tokenized URL, we found we were getting 100% success rate for iOS end users.

My theory is there is a malicious bit of code in iOS Safari that is applying some insane cookie policy over "unpopular" domains.

When do legit bugs turn into active negligence?

No, that also gets deleted sometimes.

But on the plus side, random deletions also happen on Android!

The unfortunate truth is that for too many businesses this is equivalent to "stop running your business"

Genuine question - not defending Apple/google - does your pwa not use a dedicated server / db for user’s data? And then localstorage just as a cache if needed?

Easier with Android. Apple requires many hurdles, at least for US. You'll also then have to revenue share.

Depends on your audience on whether this will actually be worth the investment. Being on the app store doesn't necessarily make you visible anyhow. It's just easier to install for apple users.

If you publish the app in the App Store you will have to give Apple a chunk of you revenue as well, not just the yearly fee

Discoverability was a good argument at one time, but much less so with the shitshow the iOS app store has become.

Are you referring to builds of a PWA wrapper shipped to the App store? If you don't mind sharing, how did your approval process go? Any hangups?

In my experience that's always been the web, which is why PWA is so attractive. Especially now that Android allows you to basically publish your PWAs to the Play store.

I'm aware of the differences; I was encouraging the OP to consider whether those differences are worth the increased pain that comes with PWAs, particularly on Apple platforms.

Not OP, but the source code can be found here: https://github.com/excalidraw/excalidraw

In Safari hit the share button then scroll down and hit Add to Home Screen.

Thanks!

The algorithm is described here: https://FreeSolitaire.win/strategy I came up with it by myself.

“The strategy presented here is the basis of the computer algorithm of FreeSolitaire.win, which is used to select finishable deals, offer hints, and detect dead-ends. You also see it surface when you are told that The computer could do it in N moves at the end of a game.

This strategy/algorithm is quite simple. It does not require to “count cards”, i.e. keep track of which cards were seen in the stock. It doesn’t require to make guesses on where such or such face-down card must be located. And the procedure is the same all along the game; there is no special treatement of the opening or ending. It’s stateless, a programmer would say. […]”

AdSense.

Oh, you can play in “draw 3 by 3” mode too: https://FreeSolitaire.win/turn-three ;-)

(You can also choose the mode in the New Game dialog.)

And yes, it’s indeed harder. The solving algorithm has a winning rate of 52% in “turn one” mode, but only 18% in “turn three”.

That's how I learned it too.

Thanks!

Word of mouth, mainly. Posted it on Reddit back then, things like that. It grew organically from that. And, although it’s far from being on top results on SERP (search engine result pages), some people do find it that way.

Also: because it’s so lightweight, it got included some years ago in the Moya app (https://play.google.com/store/apps/details?id=nu.bi.moya), a popular messaging app in South Africa that is “data-free” for users (it does reverse-billing). Now half the players are South Africans!

There is also the “turn 3 by 3” mode! Please see https://news.ycombinator.com/item?id=40726628, just below in this thread.

Oh thank you for that! Could reproduce that bug.

It’s because of https://bugzilla.mozilla.org/show_bug.cgi?id=1813213

I just spent a couple hours trying to find a clean workaround, without success… Will investigate more ASAP.

Thanks again!

OK, finally found a fix:

  /*
    Work around Firefox Mobile getting the hit target (clickable area)
    of the game control buttons wrong, when the (bottom) navigation bar
    gets hidden (when scrolling down).
    
    See https://bugzilla.mozilla.org/show_bug.cgi?id=1813213
    and https://stackoverflow.com/questions/70893056/mozilla-android-addressbar-conflict-with-sticky-button-on-bottom
    and https://news.ycombinator.com/item?id=40733062 ← thanks @ttymck for the bug report!
  */
  if (/Mobile.+Firefox/.test(navigator.userAgent)) {
    self.addEventListener('resize', frameThrottle1(() => {
      document.getElementById('metagame').style.position = (window.outerHeight > window.innerHeight)
        ? 'sticky'
        : 'fixed';
    }));
  }

Thanks for the feedback!

I understand your will to disable vibrations. But thing is, right now I can’t think of how to offer that option properly. For instance, adding a Setting menu+icon just for that would be lame.

I’ll think about it more, tho.

You can put PWAs in all the app stores today with a thin wrapper, I know, I do it for all my apps.

I have a different take. As a person reliant on accessibility, PWAs almost always deliver a worse experience for me. So after a while, I learnt to prefer native apps, simply because PWAs typically deliver a worse UX for me.

The user doesn't care about the delivery tech. They care about the installation and usage experience. They want to go search the app store, click the button, and get an icon on their homescreen.

> But some of us also want support tech that fulfills the goals of the open web. You don't seem to care about that, but others do. Telling everyone its basically a lost cause is a silly and pointless endeavor, because PWAs are the long game, and you're playing the short game.

Get off the cross, we could use the wood.

Spare me. I love the web, in fact all my “apps” are built with web tech and I offer them all as websites/PWAs in addition to apps but standing on principals of only using the web as some kind of middle finger to the app stores AND your customers is just silly.

So I bite the bullet and wrap my web apps in a thin wrapper and shove them into the App Store. You can call it principles but I just call it stubborn and dragging your customers into a crusade they don’t care about.

At the end of the day I care about the customers of my products more than I care about trying to wage a losing war against something they actively want, apps from stores.

Life is way too short to tie up your identity in things like this. I want to spend my time making products people enjoy using. If you actually care so much about PWAs then stop writing PWAs that your users hate using and go be a full time advocate for that instead. Right now shipping PWAs is a worse experience so the only one you hurt is yourself and your customers when you double down on them.

More likely, users want something that works, is a good experience, and don't care about apps _per se_ at all.

If Apple wasn't hamstringing PWAs, they might be viable for more categories of software.

Every single “write once run everywhere” technology has sucked compared to native apps.

Developers complained when the first iPhone came out and when Steve Jobs called web apps a “sweet solution”. Even Gruber called the solution a “shit sandwich”.

If it’s only Apple holding PWAs back, then why do most of the same companies write Android apps instead of just producing PWAs that work on PCs and Android devices?

That’s not a good example, I’ll bet 99% of users of that website couldn’t tell you it’s a PWA or what a PWA is. All they know is that the only way to get this “app” is add it to the home screen and if it was allowed in the App Store they’d all get it from there instead.

That’s not customers choosing PWA over native, it’s customers forced to use PWA since native isn’t an option.

How do you make messaging apps inoperable and guarantee end to end encryption? How do you add features to your messaging app without having to wait years for a standard body?

On my phone? Yes, a combo of Octal and HACK, but I often fall back to the web since the apps don't expose everything.

On my computer I always use the web. I prefer the web on my computer but on my phone apps can provide a nicer experience (larger touch targets, swipe gestures, etc). But in general I find the phone (and tablet) limiting and if I need to do any actual "research" I immediately move to my computer where I can open multiple tabs and move between them easier.

I find your experience rather relatable.

I did figure out there is some sort of spaced repetition component to it, but that’s probably because I’m already an Anki user and aware of the concept.

I quite like the eccentricity of it, although I’m sure it could be toned down a notch for wider appeal.

Feature-wise, having thesaurus definitions immediately accessible is very useful.

I appreciate the kind words. A lot of painstaking work went into aesthetics.

The Stripe integration launched a few days ago, so the app is very much in the red. Paid advertising since the initial release six months ago has shown promising signs from prospective customers though.