I’m glad we browser vendors had good sense to resist WebCL. What a debacle that would have been…
-
-
Replying to @fugueish
Yeah. It’s better (at least, some of the proposals are).
1 reply 0 retweets 0 likes -
(To elaborate: I like the proposals that don’t involve creating an entirely new shader language. SPIR-V exists, let’s use it.)
1 reply 0 retweets 4 likes -
I don’t get that position at all. “It exists let’s use it” is what got us the Web SQL debacle.
2 replies 0 retweets 1 like -
Well, it wasn’t that SQLite existed, it’s that it wasn’t spec’d. SPIR-V is spec’d, no?
1 reply 0 retweets 0 likes -
A spec is all PNaCl needed? It wasn’t a *formal* spec, but it had one. It could have been made extremely good. I doubt they would have sufficed. NaCl had a pretty simple spec too: the validator was *trivial*. I don’t think a spec is sufficient.
2 replies 0 retweets 2 likes -
Well, there are other issues at play too. Roughly: (1) is the existing tech spec’d? (2) does the existing tech meet the needs of the Web? NaCl/PNaCl meets (1) (though NaCl for x86 means importing x86 semantics so arguably not (1)) but not (2). SPIR-V meets both…
1 reply 0 retweets 0 likes -
Replying to @pcwalton @jfbastien and
With the caveat there are some issues around UB (IIRC) that will need to be addressed in SPIR-V to make a Web-SPIR-V. If they can’t be addressed then we might need something new.
3 replies 0 retweets 0 likes
Although now that I think about it, maybe using wasm as the IR (extended with intrinsics for GPU stuff) might not be a crazy idea?
-
-
0 replies 0 retweets 0 likesThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
