4 ms·
Should WebAssembly consider compatibility with the Open Web Platform important?
- rektide 5y agowhile i feel there's a lot to consider & balance, @dcodeIO has advocated in the past very strong stances for very strongly Windows quirks positions. personally i havent valued compatibility with the niche technical issues @dcodeIO attests very highly. just because windows decided to go to some freak utf16 weirdshit doesnt raise alarm for me that all else ought. i see a ton of very very high quality contributions from dcodeIO, they are a very good voice generally, but these kinds of bizarre character encoding complaints keep arising from them, having heavy windows biases, & they have long felt over emphasized & out of place. imo theres a strong perspective distortion field trying to apply itself here. with a history of false umbrage, for pokey irrelevant unimportant minor factors affecting very few.
- olliej 5y agoutf16 is not a windows-ism. It is the encoding of strings in the DOM and JavaScript. That you think it is a windows-ism indicates a lack of familiarity with the web. More specifically, JS is UCS-2 rather than utf16 so strings can also validly contain unmatched surrogates. If WASM is expected to being compatible with the web - which I would have thought was a given, given that browsers are the primary execution engines for wasm, and the V8 javascript engine seems to be the primary non-web wasm runtime - then WASM has to accept UCS-2 strings. I don't deny that the post seems biased over specific features, but I want to be super clear that the string issue specifically is not a windows-ism and is a critical element of the primary WASM host environments.