Facebook invents new unit of time to measure modern attention spans: 1/705,600,000 of a sec Video effects designers who work with C++ code have a new unit of time to work with called a "flick." Short for "frame-tick" if you're willing to overlook the absence of the letter "l" from either word, a flick lasts 1/705,600,000 of a second. It's a bit longer than a nanosecond, which clocks in at one billionth (1/1,000,000, …
I thought it made sense , more or less.
Your version says the same pretty much , if you add the x24 on the end.
I think Tom Scotts video floating point calcs would explain this better...
Having said that , I'll try the authors suggestion ....
Represent the duration of a single 24fps frame in nanoseconds....
ok 1 billion / 24 = 41,666,666.66666667 nano seconds per frame
"then multiply that by 24, the result isn't exactly a billion."
ok <presses x24 on calc> , nope that *is* a billion.
maybe my calculator is rounding off the recurring sixes at the right point ...
It's a unit in which various fractions of a second can be represented. If you want a unit in which both 1/24 and 1/25 seconds are integers, you need something that is 1/600 seconds or smaller. As you add more and more fractions you want to represent, your unit becomes smaller and smaller. At that point, you already need an int64 to store anything useful, so you might as well add a few zeroes just to be sure you'll never need more.
It's like amounts of money tend to be stored in micros of currency unit, not because you care about a micropound, but because you want to be sure never to need more precision than is available, and by the time you need that precision, you don't want to recode everything.
"IIRC even nuclear weapon engineers don't work on time scales that short."
Why would you think that? I routinely work with picoseconds and femtoseconds; nanoseconds are actually pretty long as far as physics is concerned (I'm actually in particle physics, but nuclear physics uses timescales just as short).
El Reg forgot to mention that the flick will also be able to encompass the weird legacy audio/video related rational numbers.
Factors like 44.1, 29.97, and 23.976023976230... (NTSC frame rate, last 6 digits repeating) will result in an integer number of flicks.
This post has been deleted by its author
Zuck was from a different dimension. This goes to prove it.
I am so glad that I have avoided FB all these years and have no intention of going over to the Zuck side.
Posting AC as I don't want the evil Zuck rays to find me and force me to join FB and other antisocial media junk sites.
I suspect that this "time" was designed to count anyone (all of us none users) who even thought of following a link to FB as another FB user to inflate the numbers. The only way out is to not even think or mention FB.. Oops.. my bad.. mentioned them twice, they counted 2 clicks.
Don't forget the microfortnight as used to configure VAX/VMS timeout.
just done some sums .. its not far off that:
186282 miles per second
5280 feet per mile
= 982080000 feet per second
12 inches in a foot
=11784960000 inches per second
1000000000 a billion nanos in a sec
=11.78496 inches per nanosecond
Although that margin is probably significant when considering light speed and associated distances.
"Without flicks functions written to perform accurate conversions, the losses accumulating through timing mismatches would result in a loss of precision and could throw off the execution of the simulation."
I'd say FTFY - but in fact they probably have written those functions, specifically to handle 'flicks'.
All that's happened is that someone has discovered maths, worked out the lowest common multiple of the frame rates (x 1000), given it a name, and shouted about it as though it's some kind of miracle.
Thinking about this while in the shower, I've imagined a scenario that may have played out at their premises.
* cue wavy special effect to change the location from El Reg's comments page to Programmer Bob's desk...
Programmer Bob is sat at his desk, working away - or at least using his computer for something, which may or may not be work - when he decides he needs a coffee, so he gets up to head for the canteen. His route takes him past Supervisor Bob's office, where he can see Supervisor Bob deep in conversation with Marketing Bob.
He gives them a nod as he walks past, and Supervisor Bob waves at him to come in.
"Hey Programmer Bob," says Supervisor Bob, "I'm glad I caught you. I just wanted to check something. I was looking over some of your code earlier, and I noticed something, and wanted to clarify. You've defined a constant - frick - as seven hundred and five million, six hundred thousand, and used it in various places. What is it?"
"Uh, hey, yeah, we work to a precision of one one thousandth of a frame - that's a frame tick - and various frame rates. That number is the lowest common multiple; we can use it for the math* in any frame rate without resorting to complicated fractions."
"Ah, thought it was something like that."
Noticing the blank expression on Marketing Bob's face, Supervisor Bob starts to explain.
"You know that there are various different frame rates used in video?" he asks.
"Yeah," says Marketing Bob.
"And you know we work to a precision of one one thousandth of a frame?"
"Yeah," says Marketing Bob, showing early signs of not knowing at all.
Supervisor Bob nods to Programmer Bob to continue.
"Well, the math* surrounding this can be quite complicated," he explains, "and representing frames - or fractions of frames - in a number of nanoseconds can be messy and imprecise."
"Okaaaay," says Marketing Bob, nodding, while his face looks increasingly blank.
"So I figured, let's work out the lowest common multiple of all the frame rates multiplied by a thousand. That's seven hundred and five million, six hundred thousand. Define that as a constant, which I've called frick - it's short for frame tick - and we can use that throughout our code. One thousandth of each frame rate - the precision we work to - can be represented by a certain number of fricks."
"I see," says Marketing Bob, looking at Supervisor Bob in the hope he might offer a clue about what Programmer Bob just explained.
"Basically," says Supervisor Bob, trying to make things clearer and summarising what Programmer Bob said, "Programmer Bob has worked out a number that's really useful for the math* in our code. Seconds and nanoseconds aren't accurate enough, but the constant frick is a good way to represent each amount of time we need to work to."
"Aha!" says Marketing Bob, "So Programmer Bob has invented a new unit of time?"
"Well..." starts both Programmer Bob and Supervisor Bob, hesitatingly.
"That's genius!" exclaims Marketing Bob, getting up out of his chair, "I think we can do something with that. I'm off to write a press release. Can you email me an explanation - in simple terms if possible, I don't want to confuse the people we send it to."
He goes to walk out of the door, then hesitates.
"One thing, though. Can you change its name - frick sounds a bit sweary. How about a flick? That's got huge potential!"
* 'math' rather than 'maths' because all three Bobs are probably Overpuddlians.
The idea is that nano-second results aren't converted up to a common frame rate, but that a common denominator is used from the outset. Making something a convention often does require publicising it, though a movie studio could suggest it to the digital effects houses they subcontract to - often many work on the same production.
"If you represent the duration of a single frame at 1/24 fps in nanoseconds and then multiply that by 24, the result isn't exactly a billion. And that imprecision can cause problems."
Imprecision from being close to but not exactly a thousand/million/billion boesn't seem to have hampered the storage industry (and also allowed legions of pedants to jump in with "but they're not giviing you a full GB" on a regular basis)
Developers invent their own timescales all the time. For example I once made a Unix-epoch clone starting at the year 2000. IBM-PCs calculated the time in 1/18.2065 Hz intervalls, making an overflow of the 16-Bit Number roughly every 3600 seconds.
On 16-Bit systems it's common to count seconds of the week, because that nicely fits in a 16 Bit number.
In the US television has the weird framerate of 30000/1001 Hz because when colour was introduced there it clashed with sound, and instead of slightly changing the audio subcarrier frequency, they changed the framerate. Therefore today all of their framerates for colour video are derived from that number. Of course seconds are still seconds, so you'll have seconds with 30 frames and seconds with 29 frames.
The flick joins a few other obscure time intervals including a jiffy (the definition varies), a shake (10 nanoseconds), and a microcentury (~52 minutes or a millionth of a century,
Or any other unit regularly used in conjunction with the phrase
"I'll be back in a..."
"All hail the US! We'll invent whatever flickin' units we need, and screw the rest of you!"
FTFY
Where's the "this is what style of markup we use, and it's help document" when I need it? And while I'm at it, where's the icon selection thingy? Maybe I need to let a few more things through uMatrix?
'Video effects designers who work with C++ code have a new unit of time to work with called a "flick."
'Short for "frame-tick" if you're willing to overlook the absence of the letter "l" from either word, a flick lasts 1/705,600,000 of a second.'
So would those working in C use 1/705,600,000-1 of a second? Lua people would use 1+1/705,600,000, coz that's 1 based rather than 0 based like C and C++. Python people prefer significant white space, but I'm no Python person, so I have no idea where they would put that, 1 / 705 600 000 perhaps?
There are 97 other languages on my resume, but I can't be arsed to go through them right now.
I'm pretty sure that the original whitepaper used a 'u' ... but when the print copy was scanned in a smudge on the glass caused OCR misread it as 'li'. Which may be the only universe in which I would give a f*ck about any ridiculous unit of time that Facebook "invented".
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
