haberman

53 ms·

571.
by haberman 6y ago
> Things may have changed since, but AFAIK the C++ implementation would always allocate on the heap for nested messages This is no longer the case if you use arenas: https://developers.google.com/protocol-buffers/doc
572.
by haberman 6y ago
> Google's implementations, at least C++ and Java, are a bunch of bloated crap (or maybe they're very good, but for a use case that I haven't yet encountered). As someone who has been working on protobuf-related things for
573.
by haberman 6y ago
The author of this piece is wanting to stream possibly multi-MB messages with sub-linear memory use. See more in our Twitter conversation here: https://twitter.com/richardstartin/status/134406813297822105...
574.
by haberman 6y ago
> Beyond all this, without understanding the underlying cause of the behavior you mention, it's hard to say from a risk perspective if the car made the right choice. What argument could there be that suddenly slowing down on the fre
575.
by haberman 6y ago
The car is comfortable, responsive, stylish, fun, and feels great to drive. The cruise control is fantastic in stop-and-go traffic. I always look forward to driving it. And then every once in a while does something that drives me absolutel
576.
by haberman 6y ago
That's quite reassuring to hear. It's surprising to me that the software wouldn't have been stereoscopically merging cameras already, but adding that capability does sound like the kind of thing that could actually lead to a
577.
by haberman 6y ago
I love my Model 3, but I cannot overstate how little confidence I have that Tesla is anywhere near to full self driving. Just this weekend I drove a 600 mile round trip. It was cold weather with some snow on the shoulders, but the actual r
578.
by haberman 6y ago
I'm skimming your link trying to understand the design. It sounds like there is a global flag for whether the whole program is in evented or blocking mode? > during compile-time, it’s possible to inspect if the overall program is i
579.
by haberman 6y ago
Perhaps it is just a defect in my hardware, but on mine it is nearly impossible to reliably "soft" press in either direction. Every motion either registers as nothing or hard press.
580.
by haberman 6y ago
Somewhat ironically, the physical controls for the turns signals are by far the worst part of the whole Tesla UI. I've had my Model 3 for over two years now, and I still cannot reliably cancel a turn signal, or "soft push" t
581.
by haberman 6y ago
> Toll tells his patients that those designated as most at-risk by public health officials should get it first. Let us hope this is true. Because it appears that public policy experts are racializing the question of who is an "esse
582.
by haberman 6y ago
Thanks for the tips. I'm unable to replicate my previous experiment re: -arch. When I compile a wrapper program that does an exec() of clang, it is able to build arm64 or x86_64 even if the wrapper is built as x86_64. The "arch&qu
583.
by haberman 6y ago
The same thing happened to me the first time I tried to benchmark my code on M1. In my case I was building using Bazel. Bazel was running under Rosetta because they don't release a darwin_arm64 build yet. I didn't realize the resu
584.
by haberman 6y ago
Interesting, it looks like there was a move from V8 to QuickJS more recently than that: https://frida.re/news/2020/10/28/frida-14-0-released/ EDIT: and then the default moved back in https:/
585.
by haberman 6y ago
Wow, that is very interesting. The docs say it is injecting QuickJS though, not v8? > Frida’s core is written in C and injects QuickJS into the target processes, where your JS gets executed with full access to memory, hooking functions
586.
by haberman 6y ago
> You can redraw it with ^L, or disable the tui while it's running. I often run into a problem where switching to TUI after a crash disables local echo for some reason. I can't see my typing and TUI's output is totally bus
587.
by haberman 6y ago
I think the classic essay "How to Deconstruct Almost Anything" from 1993 is a counterexample: https://www.info.ucl.ac.be/~pvr/decon.html
588.
by haberman 6y ago
> Packrat parsers are notably faster than recursive descent parsers I think this needs to be qualified. I don't think a packrat parser is going to beat a deterministic top-down parser. Maybe the packrat parser will win if the recu
589.
by haberman 6y ago
Interesting, are you suggesting that a large part of the M1 performance advantage is thanks to the weaker ARM memory model? Is the 20% perf hit of TSO mode that you cite an ARM vs. ARM comparison? If so, that would be pretty damning. Is th
590.
by haberman 6y ago
I am trying to wrap my head around whether ARM's looser memory model is a fundamental performance advantage or not. I had always assumed that the looser memory model must have a performance benefit. But this comment from last week argu
591.
by haberman 6y ago
> The problem with using a permissive license such as Apache is, anyone can fork the project exclusively under xGPLv3. If that fork becomes more popular, there's no way for the Apache-licensed upstream to merge any patches from the
592.
by haberman 6y ago
There is also a separate article from this site that speculates that Tesla's easily observable quality problems with Autopilot are actually the result of Tesla intentionally hobbling it so drivers don't become complacent: https:&
593.
by haberman 6y ago
My understanding is that developer mode on Android and Chromebook will unlock the bootloader. Software wise, that's about as user-hackable as you can get. It's true that the Oculus Quest is not as open. I don't expect or ev
594.
by haberman 6y ago
"Developer Mode" exists in many new devices, eg. Chromebooks, Oculus Quest, Android phones.
595.
by haberman 6y ago
"What is the counter to this article?" does not deserve downvotes.
596.
by haberman 6y ago
What is the objective bright line between reasonable observation of an employee and "workplace surveillance"? Is it "workplace surveillance" when you can view the calendar appointments of your coworkers? Or see stats ab
597.
by haberman 6y ago
I mentioned that I ran into two problems with this: - it only runs on Windows (I was trying to install without having Windows already). - it creates an installer that cannot install to a GPT partition.
598.
by haberman 6y ago
> Why is Windows a nightmare? I won't call Windows a "nightmare" overall, but I had a nightmarish experience trying to install it last night. It turns out that creating a bootable USB stick that can install Windows 10 is a
599.
by haberman 6y ago
Retiring memory operations earlier must have some benefit though? If data is being passed between threads, earlier retirement of a store means the data is available earlier to other CPUs. For a mutex, it means the mutex is released sooner
600.
by haberman 6y ago
> Then you specify that the source files depend on the makefile. Then you gratuitously rebuild everything whenever the Makefile changes, even if you only changed a comment. Also this scheme is incorrect if the Makefile was written to all

More ›