I think, with the evidence presented above, you should maybe re-evaluate your claim that these groups are "very talented".
-
-
Another possibility is that they are actually talented, but the foundation (the web, browsers, etc) is too awful to create high-performance software on.
1 reply 0 retweets 1 like -
Replying to @shwa_programs @Jonathan_Blow
skia is just 2D canvas lib, it's not browser-related at all - it's also used by sublime text, flutter, sciter, ...
1 reply 0 retweets 0 likes -
Replying to @cztomsik @Jonathan_Blow
I see. I use Sublime Text every day and it runs pretty flawlessly. I will admit I am unfamiliar with Skia, but if it's a 2D rendering library I struggle to think of any reason for it to be huge (though I don't know how huge you're talking here).
3 replies 0 retweets 0 likes -
Replying to @shwa_programs @Jonathan_Blow
by HUGE I mean last time it took 45mins to compile it has lots of edge-cases for different hardware, also it has high-quality antialiasing and font rendering for example (you don't care about these in games but you do for webapps)https://github.com/google/skia
2 replies 0 retweets 0 likes -
Replying to @cztomsik @Jonathan_Blow
If you think we don't care about high-quality AA and font rendering in games you simply have no idea what you're talking about
2 replies 0 retweets 1 like -
I refer you again to the Battlefield V footage above. Your ideas of how powerful computers are are incompatible with any AAA game, and are in need of some revision
1 reply 0 retweets 1 like -
Replying to @shwa_programs @Jonathan_Blow
I am not talking about AAA hardware, I am reacting to a nonsense like that it's easy to do high-quality, antialiased vector graphics on a low-level GPU
3 replies 0 retweets 0 likes -
Replying to @cztomsik @shwa_programs
It is way easier than doing high-quality, antialiased 3D animated meshes with dynamic lighting and complex reflectance functions, which is now done routinely at way higher frame rates. I don’t know how else to communicate this. Look at a 3D game, how do you think it happens?
2 replies 0 retweets 8 likes -
Replying to @Jonathan_Blow @shwa_programs
Sorry, I wanted to say low-end. Yeah, I don't know? Do you tesselate it? or SDF? those are common approaches, are there other tricks? Every one of those has trade-offs. So most libs just do glyph atlasing.
1 reply 0 retweets 0 likes
There are many other tricks, usually involving directly rasterizing the glyph from the boundary representation (or part of it) in the pixel shader. There are many, many papers on this topic from the past decade.
-
-
Replying to @cmuratori @cztomsik and
(Although I'm not sure what that has to do with this discussion? Most of a web page is static, so no matter how you did it, even if it was entirely CPU-side, it would be much faster than the observed performance.)
0 replies 0 retweets 6 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.