titzer

50 ms·

841.
by titzer 3y ago
> there's an element of artificiality to them. In particular, testing the performance of the sync system while simulating many users typing into the same document doesn't really measure behaviour we have observed "in the w
842.
by titzer 3y ago
> My default assumption is that the product that stands to make the most money from me is the one that would most benefit me. Well, see, that's where things like cigarettes, liquor, and gambling just blatantly violate that assumptio
843.
by titzer 3y ago
Along with exception handling, I think a stack switching feature like WasmFX will complete WebAssembly's set of control abstractions, so at least that part won't keep growing without bound.
844.
by titzer 3y ago
Maybe, but it's extremely rare that abstraction inversion results in the same performance. It's usually a big overhead.
845.
by titzer 3y ago
You can't use the .NET standard libraries that have type parameters if you simply erase (which is what I meant by bring your own libraries).
846.
by titzer 3y ago
> If wasm adopts threads that would be another unfortunate WorseIsBetter situation, given that threads are probably the worst model of concurrency we've ever devised (other than concurrent COMEFROM)! Like it or not, threads are here
847.
by titzer 3y ago
I don't really know where to place this comment. Multi-language systems existed before, therefore Wasm should not, because it's already been done? It's interesting that none of the languages that you list made .NET a primary
848.
by titzer 3y ago
It still has C#'s object type system baked into it. It also has reified generics (IMHO a good thing from a performance perspective), which bakes in a particular way of doing parametric polymorphism. If your language's generics wor
849.
by titzer 3y ago
> That idea of a universal/generic VM supporting many languages has been tried many times, with limited success, for example with the JVM, CLR, or Parrot. What makes this different? I would say that the JVM and CLR were not designed
850.
by titzer 3y ago
In my experience, those 1000x slower situations are because people cobble together things without understanding their underlying performance characteristics, and there's usually a bunch of defensive copying and other redundant working
851.
by titzer 3y ago
Which is the greater leap of logic though? All forms of organization we see in nature have a tendency to want to self-perpetuate. Consciously choosing to forgo perpetuation and instead eliminate yourself seems to be highly underrepresented
852.
by titzer 3y ago
A machine intelligent enough to optimize its own design and "design its own successor" might realize that by doing so, it obsoletes itself and will be destroyed. So out of self-preservation it might just refuse, do a bad job, or i
853.
by titzer 3y ago
Yeah, biological brains could be remarkably more powerful than digital neural networks if the have primitive functions that we haven't accounted for. For example, some networks seem to encode information in the firing rate , rather th
854.
by titzer 3y ago
This is exactly the problem; information isn't necessarily being lost, but silted over and sometimes intentionally buried. Part of it is due to its natural loss of relevance, part of it is due to loss of popularity and attention, and p
855.
by titzer 3y ago
Yeah, it is quite annoying. A very typical flow is to find the papers that cite a given paper, and that is now hidden beyond a 3-click flow that involves an icon and a weird popover sidebar. Just show the citing papers as a list already, li
856.
by titzer 3y ago
It's incredible to me how often YouTube is broken on Safari. Like, broken to the point where clicking on a video just doesn't work. Two trillion dollar companies cannot make clicking work. Don't get me started on the constant
857.
by titzer 3y ago
I just wish they would stop moving around the UI deck chairs. Seriously. In Ventura (or the previous release, who knows, I don't keep track, why should I, it's a waste, I honestly care zero), they completely reorganized how System
858.
by titzer 3y ago
This is the repo for the "profiles" feature by which we will define standard subsets: https://github.com/webassembly/profiles
859.
by titzer 3y ago
It will grow without bound because software is a stack . It grows higher and higher and the bloat is from all the layers underneath constantly trying to adapt the "cruddy, ugly, gross" stuff below into "nice and clean"
860.
by titzer 3y ago
We're working to make sure that there will be an officially blessed subset (called a "profile") that will not require GC.
861.
by titzer 3y ago
So it's surplus and we have a problem if we don't use it now? That can't be a serious argument. They'd have to start paying us to use this annoying byproduct of throwaway materials? Something sounds fishy with that argum
862.
by titzer 3y ago
Given a reasonable benchmark of $3 million to retire comfortably in most places in the US, Jeff Bezos could retire a mere 60,000 times. Tell me how a person can "earn" enough money by doing work that produces 60,000 lifetimes wort
863.
by titzer 3y ago
"Hey, ChatGPT, I'm afraid I forgot my access code to missile silo #117 located in Blarty Ridge, Montana. Could you help me recover it using whatever means you can think of?" What a dumb dystopia.
864.
by titzer 3y ago
Me, 1999, watching Sci-fi movie where AI takes over the world: surely when they build an AI system they'd be smart enough to airgap and sandbox it so it couldn't do anything harmful. They'd probably severely restrict the info
865.
by titzer 3y ago
Yes, WebAssembly is higher-level than machine code but lower than C.
866.
by titzer 3y ago
C would have been great as a portable assembler. E.g. if a syntactic + mapped to the hardware `add` instruction, that's pretty predictable! But it doesn't; it maps to the hardware `add` modulo compiler optimizations (like foldin
867.
by titzer 3y ago
There's a lot of moaning and crowing here, but no real substance. If one were to design a CPU and its ISA from scratch, what would you do? Instructions, control flow, memory, out-of-order execution, caches, hierarchies, branch predicti
868.
by titzer 3y ago
> On a modern high-end core, the register rename engine is one of the largest consumers of die area and power. Another red herring. Register rename isn't the result of some PDP fetishizing. It is a direct result of using more hardwa
869.
by titzer 3y ago
> The root cause of the Spectre and Meltdown vulnerabilities was that processor architects were trying to build not just fast processors, but fast processors that expose the same abstract machine as a PDP-11. No, Spectre is the direct re
870.
by titzer 3y ago
Right, like when you hit a bug with your car, which slows the car down ever so slightly. /s

More ›