Sesse__

88 ms·

871.
by Sesse__ 4y ago
You don't need a table for a simple sine oscillator. You just compute sin(a) and cos(a) on note-on, and that's it. You don't store the output in a table, you just apply that 2x2 matrix every time. Tables can definitely have t
872.
by Sesse__ 4y ago
Also if you don't care about accuracy at all. my_cos(2 * M_PI) = 0.9965. my_cos(4 * M_PI) = -2917.7144 (!).
873.
by Sesse__ 4y ago
It's not a Taylor series. It's a series, but a minimax approximation, not Taylor. Taylor series are essentially “Give me a polynomial p(x). At x=a, I want the value to be b, the first derivative to be c, the second derivative to b
874.
by Sesse__ 4y ago
If you're doing a synth and want a sine wave, you usually won't need to compute cos(x + at) from scratch every time. There's a cheaper and more accurate trick. First, figure out how much the phase (in radians) is going to inc
875.
by Sesse__ 4y ago
SRT is transport-agnostic. I've streamed AV1 within MKV within SRT.
876.
by Sesse__ 4y ago
SRT is definitely “something better”. It doesn't support codec negotiation, but it's more generally much more reasonable than RTMP for video ingest.
877.
by Sesse__ 4y ago
If you already have a receiver, pretty much anything these days will come with a room measurement system to create that FIR filter. You don't get the coefficients, though.
878.
by Sesse__ 4y ago
Shift them to the right one bit :-)
879.
by Sesse__ 4y ago
0.2% really isn't free. If you have a data set of 10 TB (one small-ish HDD right now, or one large SSD), that's ~20 GiB in page tables alone. If you can use 2MB pages, then sure, but I would assume that also limits your page-in gr
880.
by Sesse__ 4y ago
Also, generally higher overhead, more memory use for very large files (the page tables are not free), and no way of handling errors.
881.
by Sesse__ 4y ago
They don't really seem to be using a lot of CPU here? I did a quick tracing of the flying toasters screensaver, and the Chrome profiler says that a random frame I picked out used: 0.23 ms in recalc style 61 µs in pre-paint 17 µs in pai
882.
by Sesse__ 4y ago
Coreboot sets up cache as RAM through memory mapping tricks, to have something a normal C compiler can work with. (DRAM needs to be trained and set up properly before you have any RAM; before that, you have only registers, which is pretty p
883.
by Sesse__ 4y ago
That rule, while sensible at the time, has aged pretty poorly. n used to be small in the 70s, but now people have terabytes of data and run UNIX tools over them.
884.
by Sesse__ 4y ago
> In worst case complexity of styles resolution is O(nS*nD) where nS is a number of style rules and nD is a number of DOM elements. That's for current flat style table used by CSS. You can optimize bits and cases but it in general p
885.
by Sesse__ 4y ago
The flag you're looking for is --enable-blink-features=LayoutNGSubgrid, I assume.
886.
by Sesse__ 4y ago
Contemporary CSS engines use bucketing, so a rule like div[name="some"] will only ever be considered for something that has (depending on your browser implementation) either a “name” attribute or is a <div>. In other words,
887.
by Sesse__ 4y ago
Dan Luu talks a bit about this here: https://twitter.com/danluu/status/1115707741102727168
888.
by Sesse__ 4y ago
Very, very roughly: OLTP is for dealing with one row at a time (TP = transaction processing; think “handling a sale”). OLAP is for combining many rows and extracting useful information from them (AP = analytics processing; think “figure out
889.
by Sesse__ 4y ago
> In the case of OpenBSD itself, the popularity of the PCEngines APU2 has definitely provided motivation for improving the performance of PF. And Linux' packet processing now pushes 100Gig and beyond, which is not motivated by low-e
890.
by Sesse__ 4y ago
It does not. Slow hardware _can_ give you a certain amount of _motivation_ for speeding up your software, but this is by no means a given (if software running slowly on your own computer automatically made you optimize it, there would not b
891.
by Sesse__ 4y ago
If you think 2M pages are extraordinarily huge, what would you call 1G pages? :-)
892.
by Sesse__ 4y ago
FWIW, man-db index updates got a lot faster in the latest versions (roughly 40x).
893.
by Sesse__ 4y ago
The biggest problem is that it would completely kill all of the precomputation style engines do to make CSS matching fast. When you style an element, you don't actually go through the steps of matching it against every rule; but if the
894.
by Sesse__ 4y ago
> But this solution clearly doesn't require infinite lookahead. It clearly does? There can be an infinite number of tokens before you see the semicolon and know what you're parsing. The page contains examples of this, or you ca
895.
by Sesse__ 4y ago
Once you start looking at the suffix to disambiguate what the first token means, you're already in the more-than-one-token lookahead land, which is what we're trying to avoid in the first place. CSS property declarations already n
896.
by Sesse__ 4y ago
That was option 1, and it was voted out for being unneccessarily verbose.
897.
by Sesse__ 4y ago
It's not clear at all what that strict syntax would look like that would disambiguate with less lookahead. The fundamental problem is that colon is used both for declarations and in (pseudo-)selectors, not that people are somehow inven
898.
by Sesse__ 4y ago
Yes, the latter. At least Chromium-based browsers support it behind a flag (ie., you can't use it on the web yet, but you can experiment with it and file feedback). I don't know if Firefox does, too, but they very well might.
899.
by Sesse__ 4y ago
Have you looked into @scope?
900.
by Sesse__ 4y ago
If nothing else, it would strongly increase maintenance and testing costs. You also have the usual problems of cut-and-paste between examples and stylesheets not necessarily working as you'd expect. Then there's the subtle issue l

More ›