Didn't study Computer Science but I really wish I had someone like Sir Dr Mike Pound at my university. Those folk at Nottingham are lucky

He seems like he enjoy what he is doing and that is reflected on his presentations.

Some days I spend 8 hours listening to the playlist of only him. He makes everything he says make sense, even if you don't get it the first timr

To me it seems like he's explaining from a design perspective, that is "how the protocol addresses the problem", and has the knowledge to back it up. I think it's the correct level of abstraction to explain the problem clearly with only the important details.

He's got passion for computer science...

if you search for numberphile cryptography you ll find the whole playlist in this topic

I believe these are all of them, chronologically. LMK if I missed any. Apr 18, 2014 | Heartbleed, Running the Code Jul 22, 2014 | Public Key Cryptography Oct 23, 2015 | Man in the Middle Attacks & Superfish Mar 22, 2016 | Secure Web Browsing Mar 30, 2017 | End to End Encryption (E2EE) Apr 11, 2017 | SHA: Secure Hashing Algorithm Dec 15, 2017 | Secret Key Exchange (Diffie-Hellman) Dec 29, 2017 | Key Exchange Problems Jan 16, 2018 | Elliptic Curves Aug 14, 2019 | Almost All Web Encryption Works Like This (SP Networks) Nov 20, 2019 | One Encryption Standard to Rule Them All! Nov 22, 2019 | AES Explained (Advanced Encryption Standard) Oct 23, 2020 | Transport Layer Security (TLS)

@@WilliamAndrea best comment ever, thanks

@@WilliamAndrea Thank you so much. If a blank sheet of paper wants to learn this, would watching the series in release chronological order be appropriate?

@@fanllawf I guess so, yeah. You could probably skip the first one.

I agree. Insane domain of the topic.

Yeah its excellent, its a guy who knows what he is talking about in a simple way to an IT tech who always gets overwhelmed by encryption

@Dr Mike - same request.

ESNI is being replaced with ECH, both still very young protocols

I THOUGHT EVERYONE STARTed conversations that way?

@@swine13 lol 😂

yeah such a sudden portion of british accent :D

@@danieljaszczyszczykoeczews2616 Yup ^^

Class clown are we? 🤡

I-

That sounds so homosexual. (Not that there's anything wrong with that.)

giggity

@@WmSrite-pi8ck What do you mean? Getting Pounded isn't sexual.

@@jackc3727 Maybe not where you're from.

not even just product placement. It has to be full on "paid programming" level of advertisement. But wait, there's more! Comment today and we'll throw in this free sub to a channel you'll never watch again!

youtube-dl is still available.

UKposts-dl

you just reinvented the "info-mercial" :)

UKposts Premium

He really should wear a microphone

it was for TLS 1.2

@@credence7777777 He discusses TLS 1.3 towards the end of the video.

I miss TechTV. 😭

Say hi to Matt for me. :)

ACK (dammit you beat me to it by 1 hour)

The internet is full of syn.

@@Twisted_Code You missed the SYN/ACK. :P

@@Acorn_Anomaly dammit well what do I know. I have no practical experience with handshakes (yet). I only know how to reply at all because college courses LOL TBH, said inexperience is part of the reason I clicked this video... That and of course I was going to make the handshake joke if nobody had already

ACK ACK ACK, said the Martian.

But it's still susceptible to MitM attacks that simply communicate with both in TLS, if you can't verify that the public key you get actually belongs to the party you thing you're speaking to. Which you can't for sure, unless you trust a CA or meet in person and compare keys.

RSA encryption is too slow for on-the-fly encryption (and in massive amounts for the server) like web traffic...

@@leogama3422 True, however: - Both the client and the server only have to use it once per connection. - They are already using it anyway since the server signs stuff, and as far as I know, encrypting and signing are equivalent operations in RSA. The server currently signs doing PKCS#1, powering to 'd'. The client checks the signature powering to 'e'. This is equivalent to encryption, but in reverse order, isn't it?

To anyone that might read this, I just got it. The handshake I proposed is called "RSA key-exchange". Its problem is that it does not offer forward secrecy, so all previous traffic, if recorded, could be decrypted if the server's RSA key is compromised. ECDHE prevents this: previous communications will never be decrypted, even though future ones could.

that should not explain it. there's something else going on which triggers the reload. otherwise you would just look at the page as you loaded it last.

+1 to that. I was looking for videos on it about a month back and wasn’t happy with any of them really. Happy to have Dr. Mike Pounder showing us how it’s done

Google also tracks and read minds.

@@qzbnyv I could have really used this at the end of August when I was writing a paper on this 😂

@@lakshminarasimmanv Well, I know UKposts creators often make their videos based around general internet search terms frequency as a way of being timely and catching wind from trends. I haven’t looked into what the UKposts creator dashboard looks like these days, but I wonder if Google has an API that could easily display “People who watch Computerphile regularly have been searching for” stats.

UKposts use QUIC, not TLS.

@@TotalImmort7l It says TLS 1.3 on the padlock tho

@@NeunEinser which device are you using? On a phone, it shows QUIC.

@@TotalImmort7l Desktop, Win 10, Firefox

🤣🤣🤣

They are using Elliptic-curve Diffie-Hellman when they use their public/private key pairs. ECDH is fairly fast, but session keys use even faster cryptography like AES. If the cryptography is faster, loading the website would also be faster.

Another problem with using it is that if someone takes over the private key, they can decipher all previous communication, but with Diffie Helmand a new key is generated each session

Perfect Forward Secrecy is the keyword here

Asymmetric cryptography is great, but slow and demanding. One operation or RSA is much more complicated than one AES operation for example. Symmetric crypto is MUCH more efficient and also has some bonus features, like automatically included integrity checking. Therefore it makes sense to use public key crypto for exchanging a key, which is then used in symmetric crypto shenaningans. This way, we make use of the best of both worlds and get great security as well as great performance.

I came to the comments looking for this hahah

Mike may have not described it clearly but network latency is measured by Round Trip Time. So his 400ms is correct.