Cloudflare uses lava lamps to encrypt the Internet

Bryn

Doubleplusgood
Joined
Oct 29, 2010
Messages
16,252
Sounds like something stolen from Tom Scott's recent video:


[video=youtube;1cUUfMeOijg]https://www.youtube.com/watch?v=1cUUfMeOijg[/video]
 

Swa

Honorary Master
Joined
May 4, 2012
Messages
29,393
Pseudo random they mean as lava lamps aren't random. Doesn't sound like top security as radio static for instance would give.
 

LCBXX

Honorary Master
Joined
Apr 11, 2006
Messages
15,145
Pseudo random they mean as lava lamps aren't random. Doesn't sound like top security as radio static for instance would give.
But Lava Lamps looks prettier...
 

battletoad

Expert Member
Joined
Mar 10, 2009
Messages
1,432
Pseudo random they mean as lava lamps aren't random. Doesn't sound like top security as radio static for instance would give.

Others could tune into radio static I guess?
Not too sure that you can claim "lava lamps aren't random" that easily. sounds as if you're suggesting that computational fluid dynamics is an easy problem. Navier-Stokes global regularity problem springs to mind...
 

Mr Scratch

Expert Member
Joined
May 15, 2013
Messages
4,819
Pseudo random they mean as lava lamps aren't random. Doesn't sound like top security as radio static for instance would give.

What? The lamps are only to add entropy to their existing entropy pool, the static also only just adds entropy.
 

Mr Scratch

Expert Member
Joined
May 15, 2013
Messages
4,819
Others could tune into radio static I guess?
Not too sure that you can claim "lava lamps aren't random" that easily. sounds as if you're suggesting that computational fluid dynamics is an easy problem. Navier-Stokes global regularity problem springs to mind...

Impossible to predict the movements of all lamps at all times without some serious computational power, and even then, you just compromise a single source of entropy for CF - of which I imagine they have many.
 

Swa

Honorary Master
Joined
May 4, 2012
Messages
29,393
Others could tune into radio static I guess?
Not too sure that you can claim "lava lamps aren't random" that easily. sounds as if you're suggesting that computational fluid dynamics is an easy problem. Navier-Stokes global regularity problem springs to mind...
Radio static is different in every location and can even originate from the equipment you're using. It's not something that you can "tune" into because it doesn't originate from a transmitter but from background noise. It is in fact used by some of the internet random number generators.

It's not about predicting the movement of the lamps but about the fact they do have a pattern so at best the resulting keys can only be pseudo random.
 

Mr Scratch

Expert Member
Joined
May 15, 2013
Messages
4,819
Radio static is different in every location and can even originate from the equipment you're using. It's not something that you can "tune" into because it doesn't originate from a transmitter but from background noise. It is in fact used by some of the internet random number generators.

It's not about predicting the movement of the lamps but about the fact they do have a pattern so at best the resulting keys can only be pseudo random.

What would you consider to be truly random?
 

Swa

Honorary Master
Joined
May 4, 2012
Messages
29,393
I guess the answer to my question is your train of thought.
To know what "truly random" is you first have to answer what randomness is. Random it something that happens without any bounds. Like a rock on Pluto suddenly falling on my head would be random. Nothing is therefor really random, you only get a level of randomness within a specific set of rules.

If I tell you to give me five numbers from 0 to 9 you'll have to find a way to pick some random numbers. So you may look around and see your watch is on 4, next you see cd and the first number on that is 7, you look around more and see a paper you written on with the number 2, the slip that just arrived from the post office has a 48 on it. That is a random number within specific bounds.

Now instead of doing this you simply think out to yourself I'll pick 1 as it's an easy option, now let's go for something in the middle so 5 looks good, next we'll go for 7, oh no we can't pick 1 again so let's go for 2 and next we'll take 9. This may look random but in fact there's a good chance you won't pick the same number twice when in fact a random pick has more than a 50% chance of duplicate numbers. You'll also go for easy options like at the start, end or middle and ignore some numbers. This is in fact a number that only appears random but has a recognisable pattern so is really a pseudo random number.
 

bwana

MyBroadband
Super Moderator
Joined
Feb 23, 2005
Messages
82,105
Sounds like something stolen from Tom Scott's recent video:
I'm confused - who stole what from whom?

Pseudo random they mean as lava lamps aren't random. Doesn't sound like top security as radio static for instance would give.

But it's not just the lava lamps bubbles but also the "noise" in the photographs that they're using, how is this different than radio static?
 
Last edited:

Mr Scratch

Expert Member
Joined
May 15, 2013
Messages
4,819
Tom Scott did a video on this not long before this article was published, guess the writer is a Tom Scott fan, which is a good thing because his videos are always quality.
 
Top