@arj Shades of https://devblogs.microsoft.com/oldnewthing/20110310-00/?p=11253

@benjojo Perhaps these are the same people who added the command-not-found package to Ubuntu's default set. (Not exactly the same I suppose, but usually the first thing I torch on any new installation I encounter.)

@regehr Amsterdam might be promising: https://social.v.st/@th/113843212405715721

@david_chisnall Bowie was looking sleepier than usual this morning.

@jk @voltagex Then all you need is a way to introduce some controlled clock skew and presto, you could even adjust the duty cycle!

@cliffle Pets are great thermal photography subjects; unfortunately my already sort of camera-shy dog gets way more so when I pull out the Flir, so he's a bit hard to capture, but I did manage this one a few months ago.

@swetland I frequently find myself thinking "okay, that's it, we've saturated the scale, the world cannot possibly get any dumber than that", only for it to find a way to outdo itself mere moments later.

@jzb It seems to do that for me regardless of the console state (or anything else). It's very odd -- I see it documented, and the tooltip that pops up if I hover over the reader mode button at the right of the URL bar clearly says "Ctrl+Alt+R", but pressing those keys triggers an immediate browser restart no matter what else is going on.

[...pokes around a bit more...]

Ah -- after starting an instance from a terminal and triggering it, I see:

Crash Annotation GraphicsCriticalError: |[0][GFX1-]: RenderCompositorSWGL failed mapping default framebuffer, no dt (t=6.3471) [GFX1-]: RenderCompositorSWGL failed mapping default framebuffer, no dt

So I guess it's crashing in some special way that leads to a restart -- usually when it crashes it just dies as you'd normally expect. Clicking the button works fine though. (Shrug.)

@jzb Is this configuration-dependent or something? For me (Firefox 128.3 ESR, on Linux) it seems to just restart the browser entirely, which was...slightly surprising. (Amusingly reminiscent of classic "press alt+F4 to ..." pranks though.)

@cks @jschauma While there have probably been occasional other problems I've encountered, refusal to complete an arbitrary filename where I want one (i.e. what @jschauma described) is the most common by such an overwhelming margin that I can't think of any others offhand. FWIW, plain filename completion is bound to M-/ (\e/ in bash/readline parlance) by default, so you can usually use that to manually override obstinate tab completion when needed.

@rk I don't think I have anywhere near enough emoji in my fonts to qualify as a cool kid, but I've been running Void on my desktop & laptop(s) for about 8 years now and am generally pretty happy with it, FWIW.

@GeoffWozniak Can I ask what registrar you're going with? I'm considering transferring mine as well and would welcome any recommendations.

@GeoffWozniak Same here; the terminal performance hype of recent years is a bit baffling, though I guess not hugely surprising:

• performance is fairly easily quantifiable
• widely-used, established implementations such as xterm and gnome-terminal weren't really designed and implemented with performance as the #1 priority (because even a straightforward, non-performance-focused implementation isn't going to be anywhere close to a bottleneck in real-world use)
• implementation is complex enough to be challenging and more than a few days' worth of work, but small enough to be tractable for a single person to take on (unlike, say, a compiler competitive with gcc or clang), so it can be (or at least start out as) a fun little project to hack on

So a motivated individual can take it on and relatively quickly feel pretty chuffed at outperforming the "major players" by a substantial margin. Sprinkle in some self-promotion and that subset of the population that was doggedly running Gentoo with -funroll-loops 20 years ago and PREEMPT_RT kernels on their desktops 10 years ago and it's the pointless performance obsession du jour.

I often wonder what fraction of the userbase would actually notice if you were to sneakily replace their {kitty,ghostty,alacritty,whatever-else} binary with a build that artificially slowed it to xterm-level performance, and I suspect the answer might approach denorm territory if expressed as a float.

@rk Nice. My alcohol-infused NYE party conversation topics last night included IEEE-754 double precision floats. I think I was the only programmer there.

@rk My parents have a void of a different species with whom I have a similar relationship...in recent years there are some grey cracks starting to appear in the darkness though, so a few stray photons at times escape.

@mxshift @cliffle @artemis I'm guessing this one of those things where most of the time it works fine, but then under specific impossible-to-predict sets of microarchitectural circumstances will just hang the core or something? (I hit a bug like that involving branch prediction and instruction prefetching across page boundaries on the ARM1176JZFS core circa 2009; reporting it to ARM and convincing them of its realness was a fun process.)

@david_chisnall @jpm @mos_8502 I wonder if that's something @kwf might be able to speak to...

@GeoffWozniak @thegibson Oh man, I'd forgotten about that line...C3PO may be kind of annoying as a character, but damn, wise words.

@rk @yvanspijk Are we counting only as written, or including as pronounced? ("Barthelona" comes to mind, though perhaps someone more familiar would consider it a distinct phoneme?)

@regehr @pervognsen @dougall @jfbastien Perhaps also @cliffle given that he was coincidentally just posting about it earlier today: https://hachyderm.io/@cliffle/113601595674347578

@mxshift @weirdunits At my last job I had plans for a little project that would have achieved speeds in that range -- simplex networking over a power cable from a server to its power shelf, transmitting bits by modulating power draw via fan control PWM settings. It would have been a great opportunity for case-sensitive unit jokes ("yeah, it achieves over 100 mbps!" "what?? megabits per second via fan speed changes?" "no, not Mbps silly, haven't you ever heard of millibits before?") but alas I never got around to implementing it.

@monsieuricon Given the last line I feel like it needs a keyboard event handler, but what to do to live up to that description isn't obvious...maybe for something on the subtle end just stop the cursor blink?

@rk My dog is strongly of the opinion that the desirable side of a shut door is whichever one he's not on. This makes things interesting when I want to keep my office door closed for a meeting or something.

@rk Memetics aside, I can think of no better use-case for "¿por qué no los dos?"

@oh_that_courtney @rk I try (if somewhat half-assedly) to keep it all...it's currently 6.1MB (287K lines, half of which are timestamps) dating back to 2016-09-30...I've got a commit to my bashrc from 2013-02-30 adding HISTFILE="$HOME/.bash_hist.zev" to avoid accidentally truncating it on the occasional run with --noprofile/--norc, which I figure I must have added after a big loss, but I no longer remember what might have destroyed whatever transpired in the 3.5 years in between the two...

@rk @oh_that_courtney Or perhaps "see my life flashing before my eyes" (and flashing likely in more ways than one, considering what it'd probably do to my terminal...)

@oh_that_courtney @rk Some years ago (2017 if the timestamps in my .bash_history are to be believed) while playing around extracting real URLs from twitter's t.co mandatory-shortener obnoxiousness I discovered that sending it a User-Agent header including a vertical tab would trigger a server-side error...and it looks like it still does:

$ curl -i -A $'beep\vboop' https://t.co
HTTP/2 520 
date: Mon, 25 Nov 2024 22:44:09 GMT
content-type: text/plain; charset=UTF-8
content-length: 15
x-frame-options: SAMEORIGIN
referrer-policy: same-origin
cache-control: private, max-age=0, no-store, no-cache, must-revalidate, post-check=0, pre-check=0
expires: Thu, 01 Jan 1970 00:00:01 GMT
server: cloudflare Pingora-Origin
cf-ray: 8e85295fbde62838-SEA

error code: 520

So who knows, perhaps they're using it for some in-band signaling roughly along those lines?