[justinschuh]

I want to be clear that I fully intend to spend my weekend with my family, and not arguing this with you. So, I'll avoid rehashing any of our past discussions regarding technical concerns with asm.js. However, as the person responsible for Chrome's NPAPI deprecation, I just want to set the record straight regarding Unity.

Unity previously had a high quality NaCl port (long before asm.js ever existed) and I fully expected that they would continue to support it. Plus, long prior to NPAPI deprecation the Unity NPAPI plugin was on our security blocklist due to a rash of vulnerabilities. So, I don't know why Unity chose to drop support for NaCl, but it occurred well after they became aware of our plans to remove NPAPI. And if you want a public record, you'll notice that Unity 4.3 was released (without NaCl support) months after our NPAPI announcement:

http://answers.unity3d.com/questions/574054/can-i-use-native...

[flohofwoe]

Here's some feedback from a game developer working on a game that runs in a browser plugin (not Unity) and also has a standalone client:

The 1.5 year notice you gave before depreciating the NPAPI in Chrome was way to short. 5 to 10 years would have been realistic to allow existing games to reach their natural end of life. Both NaCl and HTML5 are not appropriate porting target for an existing, complex PC game that's already running for several years. They are only a realistic target platform for starting new projects. What we do now is simply moving our users out of the browser into the standalone client, thankfully our users don't mind that much since most are 'hardened' PC players that don't mind the scare popup dialogs browsers put up when downloading a native installer. But it can hardly be in the interest of browser vendor to move gamers out of their platform I would think.

And for NaCl vs HTML5:

- NaCl was never allowed in the wild, only through the Chrome app store

- PNaCl simply came too late and had very long compilation times at first start, this has become better but is still a lot slower than FF's asm.js AOT compile pass

- performance differences between PNaCl and JS on Chrome are pretty much negligible, especially for a 3D game where WebGL overhead is much more important

- emscripten puts much more effort into supporting standard APIs for games then NaCl (e.g. SDL, glfw, OpenAL, ...)

- from personal experience, bugs in Chrome on the HTML5/JS side are fixed much more quickly than bugs on the NaCl side, very likely because the HTML5 side has much more resources available

From my experience, the PNaCl implementation in Chrome doesn't provide any real-world-advantages over the HTML5 way, and both platforms are a 'hard' porting target. Of all existing platforms (iOS, Android, OSX, game consoles), the web is definitely the hardest to port an existing PC game to (starting from scratch is much easier though).

[com2kid]

> However, as the person responsible for Chrome's NPAPI deprecation, I just want to set the record straight regarding Unity.

Why? You have regressed a feature that your users want. I realize that plugins aren't sexy, but Unity is something that both developers and users want. Entire development studios are based around creating online games in Unity that have active user bases who enjoy those games.

For better or for worse you have removed the ability for people to enjoy using a part of the web. But hey, an idealized vision of software perfection won out over what users wanted, so you got that going for you.

I'm a bit bitter that those in charge of browsers have en-masse decided that being pragmatic is worthless in comparison to creating idealized castles in the sky. This just as much applies to Mozilla as it does Chrome. Mozilla's response to why they won't implement Pepper is just as "pie in the sky, who cares what users want" as Chrome removing NPAPI is.

Plugins are something that users want. Ignoring how Adobe has managed to completely drive Flash's overall quality and performance into the ground, plugins are useful. They may not be useful to Developers who are doing Serious Developer Stuff at Serious Companies, but for a lot of users, my mother, my grandmother, my nieces and nephews, they are very useful.

On the other side, browser developers have content developers as customers. Those content developers prefer alternative toolchains. Flash and Unity have far better tool chains than WebGL does. Debugging problems is far easier when you don't have to go through multiple translation layers. And less layers in the stack means less places for something to go wrong. If you are writing against Flash then Flash may have a problem. If you are writing against 3 different browser's implementation of WebGL and/or ASM.js you now have 3 different sets of bugs and implementation quirks. (Not that Flash works the same across all browsers, but at least it is from one vendor!)

But all of that is thrown away in the name of "web standards".

Users care about what works. As developers we have a responsibility to make users happy, our own sense of happiness is sort of not the primary concern.

[tedmielczarek]

> This just as much applies to Mozilla as it does Chrome. Mozilla's response to why they won't implement Pepper is just as "pie in the sky, who cares what users want" as Chrome removing NPAPI is.

This seems like an unfair comparison. "Remove existing functionality that we no longer want to support" and "support a large technology with a single implementation and no specification" are not even in the same ballpark.

[azakai]

Not sure what I did to deserve that hostility. I respect you and your work, and I agree that deprecating NPAPI is a good thing.

I don't know anything about why Unity stopped supporting NaCl.

But the fact remains that Chrome has left Unity no good option. Chrome's asm.js performance on Unity (and large asm.js codebases in general) is lagging. NPAPI is gone (again, a good thing by itself). NaCl exists, but would require Unity to maintain support for a platform only for a single browser.

[justinschuh]

It's not hostility. It's just that you're a very passionate Mozilla employee and a very passionate creator of asm.js. And from past experience I expect that you'll be willing to continue a debate longer than I can. Plus, you already know that I have reservations about asm.js, and it's just not worth rehashing that debate (accepting that WebAssembly appears to be heading in a direction that addresses most concerns, so I should offer well earned kudos on that front).

Getting back to your point, pragmatically speaking, maintaining a port is a distinction with less difference than implied. First, any reasonably complicated NPAPI plugin needs to handle a range of browser-specific quirks, meaning that significant porting is unavoidable even if it's bundled into one binary. Second, Unity supports ActiveX in IE, so they're clearly already comfortable maintaining a browser-specific port that doesn't offer any of the security or platform portability benefits of NaCl.

Outside of that, there's just the one-sided nature of the argument. As a creator of asm.js you are obviously a huge proponent. However, you must appreciate that not everyone perceives it the way you do, and not everyone agrees with the path you chose for adding support into Firefox. So, the technical burdens on another browser are just not the same, and in the specific case of Unity in Chrome have to be balanced against the fact there was already a tremendous past investment in a high-quality implementation.

So, no, I completely disagree with the framing that "Chrome has left Unity no good option". Independent of any work in Chrome on better supporting asm.js, the fact is that Unity had at least one other good option. So, I'm sure Unity had their reasons for the path they chose, but it was and continues to be a very explicit choice.

[azakai]

I think I see where you're coming from. Overall, I think it's hard for us to estimate how much effort it would take for Unity to support another platform. Perhaps I was overly pessimistic when I said supporting a platform just for a single browser is too burdensome; perhaps you were overly optimistic in saying that maintaining their existing port would have been a good option. Only Unity knows the answer. So I admit you might be right on that point.

But the more important thing is that I disagree on the effort it takes to optimize asm.js. As you can see in those forum posts, out of the 4 major browsers, mainly Chrome is presenting a problem to Unity developers. Yes, it takes some work to optimize huge compiled codebases well, but Google has talented engineers and massive resources. I refuse to believe that Chrome cannot match the results of the other 3 browsers - it would be insulting to the v8 devs to assume so. Since the ability is there but the results are not, I can only guess that the Chrome developer's focus is elsewhere.

And of course there are many other important things to optimize on the web. But Unity and other high-quality 3D games are very important too. Chrome is holding back this part of the web, right now - I hope not for long.

[justinschuh]

I think you're reading something different than what I intended. I don't work on V8, and what I know about their priorities is no different from what they've stated publicly. That is, the V8 team wants to improve asm.js performance by improving overall JS performance, and not special case behavior just for asm.js. I know that's not the route you took, but that was my point about "not everyone agrees with the path you chose" with Firefox. And regardless of what either of our personal views are, it's hard to objectively argue that the V8 team is wrong in focusing on areas that positively impact the overwhelming majority of their users.

[azakai]

Yes, and I think it's fine that v8 took that route - it's actually the route that I recommended that Firefox take. It is also the route that JavaScriptCore is taking. So I completely agree with you that the v8 team's approach is reasonable.

But they haven't optimized it well enough yet. The main issue Unity (and other big asm.js codebases) are seeing is OOMs and crashes in Chrome. The v8 and JavaScriptCore approach can avoid those, by not compiling everything at once, more efficient in-memory data structures, and so forth. This helps overall JS performance as well. It's just engineering work that needs to be prioritized and done.

(AOT, as in Firefox and Edge, does have an advantage in startup speed that I think v8 and JavaScriptCore will have a hard time approaching, but that is not a blocking issue the way that crashes and OOMs are.)