Those are the number of actions per second per each service.
I am surprised that for an audience such as hackernews that no one knew to read the source
However, when building an infographic, the information is meant to be displayed in a useful and interesting way. Shouldn't be a puzzle to be deciphered.
I'm a little surprised I have to explain this, but the point of the visualization is for you to experience the scale. Did you scroll all the way to the end of the page? Then you felt the data, which was the author's goal.
It's one thing to know how far it is through the Hundred-Mile Wilderness, but it's another thing to experience it and to know for yourself how far that is.
But before/after you've experienced it, you should know how far it was, otherwise all you know is subjective, with no actual way to gauge it. "That felt really damn far" is kinda worthless unless you can say, "Wow, 100 miles felt really damn far" because then you can judge other distances.
I assumed that was a bug. Especially since the text says "Reddit votes cast" which implies "X Reddit votes cast" instead of "Reddit votes cast [visualization of a number]".
Even that was difficult because there was no thousands separator.
It normally takes me a moment to understand that 3697893 is three million and not three hundred thousand. It's even worse if that number keeps changing. (Perhaps it's just me because I'm mildly dyslexic.)
There's a number on the left below the name and a time you've been there at the top, but I don't see an actual number for likes/second. You either have to calculate it or count the pics.
They do a straight multiplication so the number doesn't appear to be pulled dynamically from any of the sites. I.E. You can take the total and the time you've been there, do a little division and get the number per second without having to count.
EDIT:
Also if you refresh the page near skype, reddit or any of the others and wait 1 second it'll show you the rate before it's modified by the time you've been on the page.
To summarise, email - that insecure, yet open and mature standard - still beats, hands down without breaking a sweat, the latest fad websites' cool metrics of performance.
It would be very interesting to compare this to whatever is current in 10 years time. I seriously hope we have at least started to consider replacing email by then.
@2. But it's not easily harnessable, being essentially invisible, so it doesn't get as much rep. Facebook, Google, etc. are like Continents and Countries. Emails...that's the Ocean.
1. Show you how open standards are important. This might be an argument for Bitcoin vs. Banks.
Eh? Speaking for the US, the Federal Reserve Bank is the US equivalent of a standard protocol. Before a central bank existed, states and colonies had their own currencies which created many problems, which makes your argument for open standards quite good.
Which is why I don't think any paid service will ever beat a protocol for communication. Email may eventually diminish (I see no evidence of this forthcoming anytime soon, though), but I don't think it will be replaced by anything other than a better protocol on top of which independent services will be built.
Keeping in mind 80-92% of it is spam, depending on studies (wikipedia numbers from 2010 apparently, if anyone has a more up-to-date source that'd be grand).
I think these graphs are based on the reported number of queries from the companies themselves.
But I think Google has always publicly under-reported the number of search queries they get. So I suspect the number is higher than we're seeing on this page, perhaps by quite a bit.
(Google says "Every day Google answers more than one billion questions" [1]. But that number hasn't changed in years. And technically ten billion questions a day is more than one billion.)
Hmmm... Now I don't know if they just added it or if I just missed it. But thanks, that'll do the job!
edit: judging by the 6 upvotes I've got so far for the original comment, it's either just been added (great) or it was already there and there's a UI issue.
The most surprising thing to me was that, according to this site, there are more YouTube videos being watched every second than there are Google searches made. Does anyone have a source to confirm this?
Sounded implausible to me too. Perhaps its the number of videos currently being viewed such that e.g. viewing a 60 second video would effectively count for 60
This. Comparing "the amount of google searches conducted in the past minute" and "the number of Youtube videos playing during the past minute" is a bit meaningless because one is a (basically) instantaneous event and the other is a continuous event that generally lasts longer than a minute.
Would much prefer to be able to see the actual number of each thing, even if it's at the bottom.
I instead found myself trying to extrapolate the number by staring at the "since you loaded this page" number and observing the difference as it changed each second.
Seems to me the internet began over 40 years ago, in 1969. If you don't count the beginning as the first connection between two nodes, how do you mark the beginning?