3 ms·

Is this Apple demonstrating nefarious behavior (ie, a battle against webapps), they don't care about webassembly (who knows why, lazy?) or are they incompetent?

Is this Apple demonstrating nefarious behavior (ie, a battle against webapps), they don't care about webassembly (who knows why, lazy?) or are they incompetent? There may be another option, but in my experience, they want all apps to go through their store and have been almost belligerent to web devs in their mobile Safari/iOS development. Edit: Or is this article inaccurate? The browser as a system update is such an obviously bad MS style thing to do, how was this repeated by Apple? How come so many people aren't completely baffled/angered by this decision? Edit 2: It looks like a comment on the article says it's fixed in an iOS beta update coming soon...

sp332 9y ago
Given the timing, it looks like a Spectre mitigation accidentally broke it.
- wrs 9y ago
  But how could they accidentally release it? WebAssembly has no tests at all? I don’t get it.
  - weego 9y ago
    Maybe it was a pragmatic decision to quickly fix something that effected everyone at the short term cost of a feature practically no one uses.
  - icedchai 9y ago
    Perhaps it is the opposite. It has tons of unit tests, but no integration tests.
wereHamster 9y ago
> or are they incompetent? Never attribute to malice that which is adequately explained by stupidity. (source: https://en.wikipedia.org/wiki/Hanlon%27s_razor https://en.wikipedia.org/wiki/Hanlon%27s_razor)
robterrell 9y ago
It's ridiculous to think they would go through the multi-year effort to implement webASM only to intentionally break it, yet leave it in the release, after just a few months of release.
btown 9y ago
I can imagine that there are a few developers, and perhaps managers, at Apple who absolutely love the promise of WebAssembly... and they are constantly being accosted by their colleagues who recognize that the success of WebAssembly is a knife right at the heart of the App Store's current revenue model. After all, why pay Apple a cut when you can have a native-performance web app that has its own in-app purchases? And of course, the developers fight back with the idea that iOS at a platform can't fall behind if the rest of the industry is moving towards WebAssembly, but that's by its very nature a long-term vs. short-term argument. So one can imagine that these heroic devs might attempt to get a better fix into a high-priority patch release - perhaps they even implemented and tested it in whatever passes for 20% time at Apple - and all it takes is one person up the chain to veto its inclusion in favor of a "just kill it to be safe" fix. It's incredibly easy to justify that "kill" in a scare-easy time such as post-Spectre, with everyone's eyes on major platforms, and nobody's going to expend that political capital. And eventually the fix lands in the next minor release, 11.3, at a time when doing so doesn't require those developers to overplay their hand. WebAssembly retreats from battle to fight another day. And the sun rises in the morning.
vbezhenar 9y ago
What percentage of users depend on WebAssembly? I thought that this technology in the very early state and Apple probably doesn't care too much about braking it for a few monthes.

sp332 9y ago

Given the timing, it looks like a Spectre mitigation accidentally broke it.

wrs 9y ago
But how could they accidentally release it? WebAssembly has no tests at all? I don’t get it.
- weego 9y ago
  Maybe it was a pragmatic decision to quickly fix something that effected everyone at the short term cost of a feature practically no one uses.
- icedchai 9y ago
  Perhaps it is the opposite. It has tons of unit tests, but no integration tests.

wereHamster 9y ago

> or are they incompetent? Never attribute to malice that which is adequately explained by stupidity. (source: https://en.wikipedia.org/wiki/Hanlon%27s_razor https://en.wikipedia.org/wiki/Hanlon%27s_razor)

robterrell 9y ago

It's ridiculous to think they would go through the multi-year effort to implement webASM only to intentionally break it, yet leave it in the release, after just a few months of release.

btown 9y ago

I can imagine that there are a few developers, and perhaps managers, at Apple who absolutely love the promise of WebAssembly... and they are constantly being accosted by their colleagues who recognize that the success of WebAssembly is a knife right at the heart of the App Store's current revenue model. After all, why pay Apple a cut when you can have a native-performance web app that has its own in-app purchases? And of course, the developers fight back with the idea that iOS at a platform can't fall behind if the rest of the industry is moving towards WebAssembly, but that's by its very nature a long-term vs. short-term argument. So one can imagine that these heroic devs might attempt to get a better fix into a high-priority patch release - perhaps they even implemented and tested it in whatever passes for 20% time at Apple - and all it takes is one person up the chain to veto its inclusion in favor of a "just kill it to be safe" fix. It's incredibly easy to justify that "kill" in a scare-easy time such as post-Spectre, with everyone's eyes on major platforms, and nobody's going to expend that political capital. And eventually the fix lands in the next minor release, 11.3, at a time when doing so doesn't require those developers to overplay their hand. WebAssembly retreats from battle to fight another day. And the sun rises in the morning.

vbezhenar 9y ago

What percentage of users depend on WebAssembly? I thought that this technology in the very early state and Apple probably doesn't care too much about braking it for a few monthes.