3 ms·
You can make 90% of people afraid of XYZ, if you write a sufficient amount of text "explaining" how dangerous XYZ is.
by IvanK_net 7y ago
You can make 90% of people afraid of XYZ, if you write a sufficient amount of text "explaining" how dangerous XYZ is.
- rubbingalcohol 7y agoBut arguably Webassembly is also dumb and has no reason to exist, so there's that on top of it being a great way to obfuscate malicious code.
- Pigo 7y agoRunning Unreal Engine in the browser, in any browser, is pretty cool. I think there's going to be a lot of interesting use cases for it.
- AnIdiotOnTheNet 7y agoCool? Maybe if one fetishizes overengineering and abstraction layers.
- saagarjha 7y agoI suggest you take a look at actually obfuscated JavaScript; it's barely any better than WebAssembly.
- TomMarius 7y agoCalling Webassembly dumb is really dumb. Look at their github repos and consider how much very cleverly done work is there. Calling it useless in a world where most people are running JS is also weird, but that's more subjective.
- chrysoprace 7y agoSaying that WASM has no reason to exist is going a bit far. Rather than trying to list the reasons, I'm going to refer to the official FAQ: https://webassembly.org/docs/use-cases/ https://webassembly.org/docs/use-cases/ > Better execution for languages and toolkits that are currently cross-compiled to the Web (C/C++, GWT, …). > Image / video editing. > Games > Peer-to-peer applications > ... In my view it's trying to create a safe sandbox for things that we previously used unsafe sandboxes (Flash, Java, Silverlight etc.) for. A standard interface for all of these will be much better for everyone.
- jmccorm 7y agoI look at it the same way I do a new technology on our mission-critical systems. It starts with a tag of 'unsafe' or 'unstable' until it is given time to prove itself out in the real-world. Sure, we might not get to geek out over the best and latest software and standards, but we're not going to bleed over the leading edge when it fails. I'm happy to give it more time to demonstrate its various modes of failure. In a year or two, I'll revisit turning javascript.options.wasm back on in FireFox. And if by that time it proves that it is trustworthy and necessary? Yes, I'm quite happy to enable it!