As a 20-year veteran of working as a network engineer, currently on the backbone for Super Huge Major Telecom, Inc., I heartily endorse this video. It is as succinct, accurate, and accessible as anything I've ever seen.
As someone who's been in the IT industry in one form or another for 25 years I completely agree. I still don't understand network routing over large numbers of nodes though - no matter how many times it's explained to me :-).
Yes indeed, I too found this obvious mistake a bit peculiar. But I guess it's an approximation that comes with the fact that they didn't have time to talk about private adresses. This series is still absolutely brilliant imho
Two Professors with PhDs in computer science spent 6 hours in lectures trying to explain this in my Master's lectures. You guys did it in 12 mins, phenomenal!
The video is totally worth watching! Now i get it: IP - is simply what and where to send UDP - on top of IP provides ports and packaging but does not ensure the delivery, just informs when the data is corrupted TCP - on top of IP also provides ports and packaging, but insures the delivery (thus increasing the network load)
I am now 38 years old and I am learning Python. I tried to learn programming early, but despite all my efforts this opportunity was not given to me in my country. Despite all the obstacles I have never given up on this dream and now I am more motivated than ever to learn from the precious help of the internet. Luckily there are many bright people here who kindly help anyone looking to learn computing. Although very difficult, I hope that one day I will become a programmer.
Moysés Gurgel So sorry that you didn't get the opportunity earlier; I was fortunate enough to have a hobbyist dad who pushed me into programming at age eight. This was in the Java 1 days.
Moysés, congratulations on your choice to learn (what seems to me) the most useful language of these years! I can only warmly recommend the Socratica channel for their Python introductory course, it's short and well explained! (In particular, I can sadly no longer recommend LPTHW which isn't available for free online any longer...)
Same here, wanted to learn computer science since 14 but such courses weren't available in my country. I couldn't afford to go abroad to learn CS and had to be good at math which I wasn't. Self taught all the way, 32 now and I am getting closer.
@@rosellecarmen2224 You mean computer science courses at age of 14 or after ? I'm actually in computer science bachelor 1rst year in my country. It's extremely accessible (by education) here.. Don't worry that almost everyone are struggling in analysis course, even mathematicians.
I know absolutely nothing about computers besides clicking the thingy to get the stuff and thought it was just magic but I love this series and have binge watched it all. Now I actually have a basic understanding on how all the components do relatively simple jobs to make everything happen and why things are the way they are. Now I just think it's magic that it was thought up by somebody in the first place. Awesome show and please keep them coming
I've been computer science since childhood (for about 25 years now) and I haven't seen such a great explanation of Internet and especially ISO/OSI standards. However I had take exams of these during my university years. Thanks for making this video and please make more videos like of this. :)
Wow, so much work has been put into this video, and all for the purpose of spreading knowledge didactically! You guys are awesome, please keep up the excellent work!
This video sums up my entire networking course in a thousandth of the time. Truly amazing stuff, I absolutely love the level of detail that goes into these videos!
I'm taking CCNA (Cisco Certified Networking Associate) and this video would have really been helpful as an introduction to the course. The Cisco site really should link to your video.
Agreed. When I started working at the Cisco TAC in 1996, I was told to get a particular book and read it cover to cover. That was my training in networking (I'd been educated as a programmer - not even close.) Two years later, I was a CCIE and was proctoring the lab test - working in TAC kinda had that effect back then. I can still name that book - "Internetworking with TCP/IP", volume 1, 3rd edition, by Douglas Comer. I still have my copy somewhere around here.... Some say, though, that in later editions Comer "sold out" to Microsoft by introducing some spurious elements to the OSI model. Sad.
Russ Emerson These days I point people to ISP Essentials by Barry Raverndran Greene and Philip Smith on Cisco Press. It’s got most of the practical stuff in it without a lot of fluff.
"Hi, I'd like to hear a TCP joke." "Hello, would you like to hear a TCP joke?" "Yes, I'd like to hear a TCP joke." "Ok, I'll tell you a TCP joke." "Ok, I will hear a TCP joke." "Are you ready to hear a TCP joke?" "Yes, I am ready to hear a TCP joke." "Ok, I am about to send the TCP joke. It will last 10 seconds, it has two characters, it does not have a setting, it ends with a punchline." "Ok, I am ready to get your TCP joke that will last 10 seconds, has two characters, does not have an explicit setting, and ends with a punchline." "I'm sorry, your connection has timed out. Hello, would you like to hear a TCP joke?"
Carrie Anne talks very fast! She also knows her stuff. I’ll need to watch these lovely well constructed videos several times to really take on board the information but I’ve love every second! Thanks for producing such clear concise and highly enjoyable videos. Magic. 👍
I love this series, thankyou! I'm really hoping at some point you'll took about encryption techniques, and also compression techniques, because I'm learning about both in my own time and they're really interesting.
UDP does not apply the checksum across the data payload. It only applies to the port numbers and parts of the IP header to make sure the packet didn't go to the wrong place.
I love you, you are simply excellent, there are many concepts so hard to understand about the computers, but you can explain them so clearly. Congratulations, this is the best channel I've ever found in youtube.
This is exceptionally well done. Brilliant animations and love the discussion on abstraction (used a lot in CS topics and something that can add confusion at first but eventually becomes super helpful and you get why its done so much!). Gonna watch all of your vids now. Cheers.
You forgot Layer 8: the User layer. If an network support guy says someone had a Layer 8 issue, that's a nice way of saying the user was being an idiot. The idea can be extended with the organisation the user was a part of being a ninth layer, and the laws of the country it operates in being a tenth.
4:14 No, it's cleverer than that. The uppermost bits overflow _into_ the lower bits: it's called "ones' complement" or "end-around carry". This is because the values to be summed are not the individual octets but rather _pairs_ of octets treated as 16-bit numbers. Thanks to this clever "ones' complement" hack, it doesn't matter whether your computer is a 386 that thinks of them as little-endian numbers, or an m68k that thinks they're big-endian: each octet carries into the other octet - it's a circle that has no ends to be big or little. There's also _another_ clever hack: the checksum isn't just the sum of the packet data, instead it's _minus_ the sum of the packet data. This means that if you want to validate the checksum on an incoming packet, you just add up the _whole_ packet, including the checksum, and if it's valid you should get *zero*. (This turns out to be a big deal for network encapsulation - where your UDP payload is another entire packet with its own IP, UDP etc. headers - because it means you can calculate the outer checksum just by looking at the outer headers, since you know the inner packet with a correct checksum adds up to zero. This is called Local Checksum Offload, and was invented by, erm, _me_. But I think the people who originally designed the Internet Protocol checksum are the ones who really deserve the credit, because it's their brilliant design that made it possible.) 7:29 The main reason for using UDP instead of TCP isn't the ACKs and the bandwidth they use, it's that latency-sensitive applications generally want each individual packet delivered as soon as possible - they don't want new data to be buffered up in the OS just because it's still waiting for some old data that's supposed to come before it but maybe isn't even relevant any more. This rationale applies to everything from video chat (I already displayed that frame with low resolution and glitches, it's too late now to tell me what it should have shown, I want the next video frame ASAP) to financial trading (give me the most recent stock price, I don't care what the price 10 microseconds ago was!). Lag is caused not by ACKs but by waiting for retransmissions and enforcing ordering constraints that - at the application layer - may be totally irrelevant. 9:11 Technically (and yes I'm being pedantic now), the very top of the tree are the _root name servers_, 'a.root-servers.net' through 'm.root-servers.net'. 10:05 Oh no, not the OSI 7-layer burrito. Real-world networking implementations don't look anything like that over-engineered heap (does anyone really understand wtf the Presentation Layer corresponds to in the Internet protocol stack?) The standard protocol stack today combines the physical and data-link layers into Ethernet (attempts to separate them out conceptually founder on the way the data-link part of it gets waaaay too intimate with the physical part). Similarly, some of the network layer's functions are handled by 'layer 4' protocols like TCP or UDP, and some functions that OSI puts in the transport layer are done either by IP or by other layer-3 protocols that go straight on top of Ethernet rather than being wrapped in IP datagrams. Sometimes the layering even differs between IPv4 and IPv6: ARP, which is considered a link layer function, is layered on top of Ethernet, making it network layer; its IPv6 equivalent, Neighbour Discovery, is part of ICMPv6, sent inside IPv6 datagrams, making it technically a transport layer protocol! Certainly the Internet protocol suite _is_ layered, but the trouble with the OSI model is that it's so neat and theoretically elaborated that it _sounds_ like it should be a strict and precise thing, whereas real, working networks are inevitably messy and don't follow neat layering boundaries. See also RFC3439. While the OSI model does give us a few nice names ("Layer 3", "Layer 4", and also (though it doesn't actually define it itself) "Layer 8" for the "user's intentions and behaviour" layer), it shouldn't be taught without a warning that it's basically a cross between numerology and stamp-collecting.
The Cheaterman The Cheaterman Use Google, he did actually invent it... And sure, they couldn't go into crazy detail, but the video definitely had flaws that wouldn't have been hard to fix. The TCP/UDP reasoning is one of them and the OSI thing didn't make a ton of sense to even include
I don't think that you are really the intended audience for this mate. "... it's cleverer than that" - sure. But if you want to have a crash course in how this works, but don't actually intend to work with this stuff it doesn't really matter in detail how a checksum works.
+Simon T Yeah I know I'm not the intended audience, +The Cheaterman I realise they couldn't go into this much depth. I'm trying to provide _additional_ information in the comments for anyone who does want a bit more depth, and also maybe give them ideas for if they ever do a "compsci 201" series like they did with CC World History. I should probably have started my comment with a disclaimer like "notwithstanding the following nitpicks, this was a well-made educational video".
You are an amazing person sharing this details! Thank you so much for such a great presentation...! If you have this as an online course I am totally in!
Hmm... I tried doing traceroot for dftba.com from Serbia. It took me only 8 hops and my latency was lower than Carrie Anne’s. Didn’t expect I had such a good connection.
why not make an alternate TCP where the transmitter just sends the whole thing and the receiver simply sums which packets didnt come in and rerequest those until the whole thing is complete. I know this wouldnt work with all applications, but most will work just fine without clogging the bandwidth with so many requests and acknowledgements
Jokes on you- I didn’t stream this. I sent someone to download this video, burn it on a disk, then play it for me off an old portable DVD player. Now he’s writing this comment for me. Checkmate, computer lady.
I hate to say it, but that first line in the output from traceroute is not *her* computer, but the address of the gateway it's connected to though it's local ethernet (or the Wifi hotspot).
Here's a practical question...why is it not recommended to type sensitive information (passwords, card numbers) on public networks (coffee shop or airport), like how exactly can someone view that info?
The OSI model doesn't really apply to the TCP/IP stack, hence it having it's own model. The session layer an above are just considered "Application layer" because how they are used and implemented is *very* application specific. DNS, for example, has no session layer.
Yes it does. TCP is layer 4 and 5. IP is layer 3. DNS has no session layer because it is an application layer protocol. It uses UDP as its transport. UDP doesn't have a strict session management. I'm sure you know all this because the OSI is intended to model and make it easier to understand where everything fits together, and isn't a protocol or even meant to be an exact representation of what's going on. Stop trying to snipe. No one uses the TCP model.
I am assuredly behind the times on this observation, but her use of "the Doopleydoo" is by far the most apparent John thing I've experienced this week.
Ok, a question, how do we avoid ip address clashes? The IPv4 protocol can contain only 4228250625 unique adresses, but there are way more physical devices, connected to the internet, and we can't assign every one of them a unique ip address. The solution I can think of is to have the top level routers split up the network into smaller bits (where the user count is below 4 billion) and then assign unique ip addresses, but then we run into the issue of having to know two (or more) ip addresses per device, one for the router, one for the device. Also, those routers would have to be monsterous to handle that amount of data and supply an adequate amount of latency. How do we deal with this issue in the real world?
There are a couple of approaches to this problem. Look them up for a lot more info. 1. IPv6. Instead of using a puny 32 bits for the address, use 128 bits, allowing for a much larger number of devices to have unique addresses. 2. NAT. Using Network Address Translation, a router receives a unique IPv4 address for its physical interface to the Internet. The router then hands out a unique address to all the devices attached to it's other side, the private LAN. The trick is that the addresses that the router hands out are duplicated behind every router. i.e. Router A hands out addresses 192.0.0.1 - 192.0.0.255, and so does Router B and C, etc. When a device behind the router sends a packet, the router re-writes the packet to use the router's IP address as the sender, then transmits onward. And returning packets are re-written and re-transmitted by the router to the device. The router keeps a translation table between its actual Internet address, and port, to the local (LAN) destination device and port. Theoretically, each router could give out an entire Internet's worth of IPv4 addresses, treating them as a private LAN, and doing packet address translation every time a packet enters or leaves the private LAN.
is this the last episode in this great series on Computing Science? Its been good viewing them, look forward to others in your series eg the history one
CrashCourse was started by John and Hank Green. John has presented multiple series about World History and Literature, while hank has presented series on Biology, Ecology, Chemistry and Philosophy.
As a 20-year veteran of working as a network engineer, currently on the backbone for Super Huge Major Telecom, Inc., I heartily endorse this video. It is as succinct, accurate, and accessible as anything I've ever seen.
As someone who's been in the IT industry in one form or another for 25 years I completely agree. I still don't understand network routing over large numbers of nodes though - no matter how many times it's explained to me :-).
same
same
🐢
🐢
🐢
Yes indeed, I too found this obvious mistake a bit peculiar. But I guess it's an approximation that comes with the fact that they didn't have time to talk about private adresses. This series is still absolutely brilliant imho
would you mind elaborating about the mishaps, so that those who are not beginners could get the right information too. TIA
Two Professors with PhDs in computer science spent 6 hours in lectures trying to explain this in my Master's lectures. You guys did it in 12 mins, phenomenal!
The video is totally worth watching!
Now i get it:
IP - is simply what and where to send
UDP - on top of IP provides ports and packaging but does not ensure the delivery, just informs when the data is corrupted
TCP - on top of IP also provides ports and packaging, but insures the delivery (thus increasing the network load)
I am now 38 years old and I am learning Python. I tried to learn programming early, but despite all my efforts this opportunity was not given to me in my country. Despite all the obstacles I have never given up on this dream and now I am more motivated than ever to learn from the precious help of the internet. Luckily there are many bright people here who kindly help anyone looking to learn computing. Although very difficult, I hope that one day I will become a programmer.
Moysés Gurgel Keep on there friend. Do what your heart desire. You will get there.
Moysés Gurgel So sorry that you didn't get the opportunity earlier; I was fortunate enough to have a hobbyist dad who pushed me into programming at age eight. This was in the Java 1 days.
Moysés, congratulations on your choice to learn (what seems to me) the most useful language of these years! I can only warmly recommend the Socratica channel for their Python introductory course, it's short and well explained! (In particular, I can sadly no longer recommend LPTHW which isn't available for free online any longer...)
Same here, wanted to learn computer science since 14 but such courses weren't available in my country. I couldn't afford to go abroad to learn CS and had to be good at math which I wasn't. Self taught all the way, 32 now and I am getting closer.
@@rosellecarmen2224 You mean computer science courses at age of 14 or after ? I'm actually in computer science bachelor 1rst year in my country. It's extremely accessible (by education) here.. Don't worry that almost everyone are struggling in analysis course, even mathematicians.
I know absolutely nothing about computers besides clicking the thingy to get the stuff and thought it was just magic but I love this series and have binge watched it all. Now I actually have a basic understanding on how all the components do relatively simple jobs to make everything happen and why things are the way they are. Now I just think it's magic that it was thought up by somebody in the first place. Awesome show and please keep them coming
I've been computer science since childhood (for about 25 years now) and I haven't seen such a great explanation of Internet and especially ISO/OSI standards. However I had take exams of these during my university years. Thanks for making this video and please make more videos like of this. :)
HEILLL NAAA www.studytonight.com/computer-networks/complete-osi-model
Just google it :)
Wow, so much work has been put into this video, and all for the purpose of spreading knowledge didactically! You guys are awesome, please keep up the excellent work!
+
If you turn off your ad block, you'll see there's another nefarious sinister reason.
Everyone involved in this : thank you ! This series is AMAZING !
This video single-handedly made me understand more about networking than a whole year of my university module.
As a non-cumputer person, thank you very much for this whole series !
lol cumputer
This video sums up my entire networking course in a thousandth of the time. Truly amazing stuff, I absolutely love the level of detail that goes into these videos!
But they have to lengthen the course to get the most money out of you. See crash course Economics 😁
Who knew so much thought went into distributing cat videos?
Best distributed by cat5 cables. Make sure you use your mouse to click on it 😁🐈🐀
I'm taking CCNA (Cisco Certified Networking Associate) and this video would have really been helpful as an introduction to the course.
The Cisco site really should link to your video.
Barnard Rabenold you have your work cut out for you if you think this would have been helpful for getting your CCNA.
I would hope you would have enough of a grasp on networking to shoot for CCNA to not have this be anything but a review. This is Network+ level stuff.
That's basically what I said. This video pretty much was a review for me, but would have been very helpful when I first started.
Agreed. When I started working at the Cisco TAC in 1996, I was told to get a particular book and read it cover to cover. That was my training in networking (I'd been educated as a programmer - not even close.) Two years later, I was a CCIE and was proctoring the lab test - working in TAC kinda had that effect back then.
I can still name that book - "Internetworking with TCP/IP", volume 1, 3rd edition, by Douglas Comer. I still have my copy somewhere around here....
Some say, though, that in later editions Comer "sold out" to Microsoft by introducing some spurious elements to the OSI model. Sad.
Russ Emerson These days I point people to ISP Essentials by Barry Raverndran Greene and Philip Smith on Cisco Press. It’s got most of the practical stuff in it without a lot of fluff.
I could tell you a joke about UDP, but you might not get it.
I didn't get it.
Can you replay?
Wokka wokka!
Would you like to hear a TCP joke?
"Hi, I'd like to hear a TCP joke."
"Hello, would you like to hear a TCP joke?"
"Yes, I'd like to hear a TCP joke."
"Ok, I'll tell you a TCP joke."
"Ok, I will hear a TCP joke."
"Are you ready to hear a TCP joke?"
"Yes, I am ready to hear a TCP joke."
"Ok, I am about to send the TCP joke. It will last 10 seconds, it has two characters, it does not have a setting, it ends with a punchline."
"Ok, I am ready to get your TCP joke that will last 10 seconds, has two characters, does not have an explicit setting, and ends with a punchline."
"I'm sorry, your connection has timed out. Hello, would you like to hear a TCP joke?"
I UDP the could of joke out you be tell, but a order might about response.
3:57 "Yo! :)" (in ASCII/UTF8)
The production of this is amazing, explained so well from scratch. I am honestly impressed, thank you so much for this!
can't say I've heard of it
it's like a net but with inter
This series has been extremely informative and super easy to process.
Informative as always, I love these videos, I learned more here than I learned in my entire 12 years of school.
But I thought the Internet was a black box with a red light on top. MY LIFE IS A LIE.
Oh that reference brings back memories.
DreadKyller it is. She is.lying
+
IT CROWD HOORAY
It is. Everything in this video takes place in a black box at the top of Big Ben
Thank you very much! This lesson is very usefull for my class (I'm a IT teacher).
Carrie Anne talks very fast! She also knows her stuff. I’ll need to watch these lovely well constructed videos several times to really take on board the information but I’ve love every second! Thanks for producing such clear concise and highly enjoyable videos. Magic. 👍
Nobody:
Carrie Anne: *casually sharing her IP address on a 13.7 million subscriber channel*
/s
I love this series, thankyou!
I'm really hoping at some point you'll took about encryption techniques, and also compression techniques, because I'm learning about both in my own time and they're really interesting.
Conor O'Neill
She already talked about data compression a couple of weeks ago.
Easily understood such a complex topic! Many thanks.These videos are gold.
This series is absolutely fantastic!
Wow so awesome! Learning this for my internship for computer networking :)
God Bless Crash Course. I'm forever grateful of their services to humankind.
This series is really good.
It is amazing that how you can explain such a complex topic like computer networks clearly and in such a short time!!!
UDP does not apply the checksum across the data payload. It only applies to the port numbers and parts of the IP header to make sure the packet didn't go to the wrong place.
Doing the world a great service with all you offer CrashCourse! You have my support.
Thank you!
I love you, you are simply excellent, there are many concepts so hard to understand about the computers, but you can explain them so clearly. Congratulations, this is the best channel I've ever found in youtube.
Thanks!
This is exceptionally well done. Brilliant animations and love the discussion on abstraction (used a lot in CS topics and something that can add confusion at first but eventually becomes super helpful and you get why its done so much!). Gonna watch all of your vids now. Cheers.
You forgot Layer 8: the User layer.
If an network support guy says someone had a Layer 8 issue, that's a nice way of saying the user was being an idiot.
The idea can be extended with the organisation the user was a part of being a ninth layer, and the laws of the country it operates in being a tenth.
I can't believe I had a class on computer networks in college that didn't teach me anything nearly as efficiently as two 10-minutes youtube video.
ISP...
ISP . . .
_ISP_ . . .
*Isp*
*specific impulse*
*KERBAL SPACE PROGRAM*
4:14 No, it's cleverer than that. The uppermost bits overflow _into_ the lower bits: it's called "ones' complement" or "end-around carry". This is because the values to be summed are not the individual octets but rather _pairs_ of octets treated as 16-bit numbers. Thanks to this clever "ones' complement" hack, it doesn't matter whether your computer is a 386 that thinks of them as little-endian numbers, or an m68k that thinks they're big-endian: each octet carries into the other octet - it's a circle that has no ends to be big or little.
There's also _another_ clever hack: the checksum isn't just the sum of the packet data, instead it's _minus_ the sum of the packet data. This means that if you want to validate the checksum on an incoming packet, you just add up the _whole_ packet, including the checksum, and if it's valid you should get *zero*. (This turns out to be a big deal for network encapsulation - where your UDP payload is another entire packet with its own IP, UDP etc. headers - because it means you can calculate the outer checksum just by looking at the outer headers, since you know the inner packet with a correct checksum adds up to zero. This is called Local Checksum Offload, and was invented by, erm, _me_. But I think the people who originally designed the Internet Protocol checksum are the ones who really deserve the credit, because it's their brilliant design that made it possible.)
7:29 The main reason for using UDP instead of TCP isn't the ACKs and the bandwidth they use, it's that latency-sensitive applications generally want each individual packet delivered as soon as possible - they don't want new data to be buffered up in the OS just because it's still waiting for some old data that's supposed to come before it but maybe isn't even relevant any more. This rationale applies to everything from video chat (I already displayed that frame with low resolution and glitches, it's too late now to tell me what it should have shown, I want the next video frame ASAP) to financial trading (give me the most recent stock price, I don't care what the price 10 microseconds ago was!). Lag is caused not by ACKs but by waiting for retransmissions and enforcing ordering constraints that - at the application layer - may be totally irrelevant.
9:11 Technically (and yes I'm being pedantic now), the very top of the tree are the _root name servers_, 'a.root-servers.net' through 'm.root-servers.net'.
10:05 Oh no, not the OSI 7-layer burrito. Real-world networking implementations don't look anything like that over-engineered heap (does anyone really understand wtf the Presentation Layer corresponds to in the Internet protocol stack?) The standard protocol stack today combines the physical and data-link layers into Ethernet (attempts to separate them out conceptually founder on the way the data-link part of it gets waaaay too intimate with the physical part). Similarly, some of the network layer's functions are handled by 'layer 4' protocols like TCP or UDP, and some functions that OSI puts in the transport layer are done either by IP or by other layer-3 protocols that go straight on top of Ethernet rather than being wrapped in IP datagrams. Sometimes the layering even differs between IPv4 and IPv6: ARP, which is considered a link layer function, is layered on top of Ethernet, making it network layer; its IPv6 equivalent, Neighbour Discovery, is part of ICMPv6, sent inside IPv6 datagrams, making it technically a transport layer protocol!
Certainly the Internet protocol suite _is_ layered, but the trouble with the OSI model is that it's so neat and theoretically elaborated that it _sounds_ like it should be a strict and precise thing, whereas real, working networks are inevitably messy and don't follow neat layering boundaries. See also RFC3439. While the OSI model does give us a few nice names ("Layer 3", "Layer 4", and also (though it doesn't actually define it itself) "Layer 8" for the "user's intentions and behaviour" layer), it shouldn't be taught without a warning that it's basically a cross between numerology and stamp-collecting.
"invented by me"? Otherwise, great comment, although I'm sure you realize they couldn't go into such depths in the video. :-)
The Cheaterman The Cheaterman Use Google, he did actually invent it... And sure, they couldn't go into crazy detail, but the video definitely had flaws that wouldn't have been hard to fix. The TCP/UDP reasoning is one of them and the OSI thing didn't make a ton of sense to even include
7:29 You say it isn't the ACKs and then go on to explain why it is the ACKs.
9:11 You didn't just start being pedantic here.
I don't think that you are really the intended audience for this mate. "... it's cleverer than that" - sure. But if you want to have a crash course in how this works, but don't actually intend to work with this stuff it doesn't really matter in detail how a checksum works.
+Simon T Yeah I know I'm not the intended audience, +The Cheaterman I realise they couldn't go into this much depth. I'm trying to provide _additional_ information in the comments for anyone who does want a bit more depth, and also maybe give them ideas for if they ever do a "compsci 201" series like they did with CC World History. I should probably have started my comment with a disclaimer like "notwithstanding the following nitpicks, this was a well-made educational video".
Samuel D. Reynolds.
Wow, this video from Crash Course is very easy to understand. As a beginner, this course will help me a lot. Thank you!
“You computer may not know which application to give the data to, Skype or call of duty? “That is an insult to Reddit
Excellent performance
and then you start working with websockets which add another layer of packets on top of the application layer packets.
Better explanation than the one I got from my professor at the University of Minnesota lol
Anthony Phan LOL
You are an amazing person sharing this details! Thank you so much for such a great presentation...! If you have this as an online course I am totally in!
Hmm... I tried doing traceroot for dftba.com from Serbia. It took me only 8 hops and my latency was lower than Carrie Anne’s. Didn’t expect I had such a good connection.
Thank you very much.
What a weird coincidence, my networking class just went over this today.
Might be worth noting that your home "router" is actually a switch, a router, and wireless AP combined.
why not make an alternate TCP where the transmitter just sends the whole thing and the receiver simply sums which packets didnt come in and rerequest those until the whole thing is complete. I know this wouldnt work with all applications, but most will work just fine without clogging the bandwidth with so many requests and acknowledgements
Network fundamentals in a nutshell. Definitely going to refer this vid to my pals interested in network administration.
Jokes on you- I didn’t stream this. I sent someone to download this video, burn it on a disk, then play it for me off an old portable DVD player. Now he’s writing this comment for me. Checkmate, computer lady.
I hate to say it, but that first line in the output from traceroute is not *her* computer, but the address of the gateway it's connected to though it's local ethernet (or the Wifi hotspot).
This series is amazing! Thanks a lot
0:15
8.4M views -> 17 thumbs down. Well that's being optimistic.
As an IT guy I got it, and it was good for me to see it again, but for the lay person I think they would say "Whaa??"
man i had problems understanding the OSI system and this video explains it in the best way possible in 10 minuets!! double like
Here's a practical question...why is it not recommended to type sensitive information (passwords, card numbers) on public networks (coffee shop or airport), like how exactly can someone view that info?
If it is not a secure HTTPS session, eavesdropping devices inserted into the network will reveal unencrypted information.
LEARNING.. SO.. MUCH...! Also, overwhelmed af.
The OSI model doesn't really apply to the TCP/IP stack, hence it having it's own model. The session layer an above are just considered "Application layer" because how they are used and implemented is *very* application specific. DNS, for example, has no session layer.
Yes it does. TCP is layer 4 and 5. IP is layer 3.
DNS has no session layer because it is an application layer protocol. It uses UDP as its transport. UDP doesn't have a strict session management.
I'm sure you know all this because the OSI is intended to model and make it easier to understand where everything fits together, and isn't a protocol or even meant to be an exact representation of what's going on.
Stop trying to snipe. No one uses the TCP model.
I love this , thank you🙏🏼 , all of you
Again, the Internet is really, really great.
Thank you soo much for this video. Never really understood the concept of UDP and TCP in college..You explained it really well.
Best teaching approach I have found. Thank You.
Mind-blowing Course
5:41 sounded like rapping for a second
Hope to see a knowledge graph of this entire series
Thank you Carrie Anne!! ❤❤
I love this video, but the transition noise was a bit distracting. You go, girl xo
I'm pretty sure that the Internet is not a big truck you can just dump stuff on. It's a series of tubes!
Thank you
The TCP packet walks into a bar, 'I would like a beer' he says. 'You would like a beer?' The bartender replies. 'I would like a beer' the packet says.
Great episode!! We just learned about UDP / TCP and other network protocols in our IT apprenticeship class.. :)
I like your use of the 8 x 12 form of the Windows built-in Raster Command Line Font for these videos (because I just think that font is cool).
This was published at the perfect time because we're learning about this in class right now
I am assuredly behind the times on this observation, but her use of "the Doopleydoo" is by far the most apparent John thing I've experienced this week.
Does this UDP-checksum perform an end-around carry?
...So did anyone else think of Timothy Leary's 8 circuit model of consciousness towards the end there?
Wow, thank you! Subscribed and liked immediately!!! :)
Amazing video, thanks guys!
Way to go Carrie Anne!
Ok, a question, how do we avoid ip address clashes?
The IPv4 protocol can contain only 4228250625 unique adresses, but there are way more physical devices, connected to the internet, and we can't assign every one of them a unique ip address. The solution I can think of is to have the top level routers split up the network into smaller bits (where the user count is below 4 billion) and then assign unique ip addresses, but then we run into the issue of having to know two (or more) ip addresses per device, one for the router, one for the device. Also, those routers would have to be monsterous to handle that amount of data and supply an adequate amount of latency.
How do we deal with this issue in the real world?
Heyyyy, you just reinvented network address translation. Go look up NAT on wikipedia.
HyperWind the latest crack at a permanent solution to this was to try to make IPv6 the standard across the board. I'll have to look up this NAT thing
Private IP addresses. For example 192.168.0.1-192.168.255.254
There are a couple of approaches to this problem. Look them up for a lot more info. 1. IPv6. Instead of using a puny 32 bits for the address, use 128 bits, allowing for a much larger number of devices to have unique addresses. 2. NAT. Using Network Address Translation, a router receives a unique IPv4 address for its physical interface to the Internet. The router then hands out a unique address to all the devices attached to it's other side, the private LAN. The trick is that the addresses that the router hands out are duplicated behind every router. i.e. Router A hands out addresses 192.0.0.1 - 192.0.0.255, and so does Router B and C, etc. When a device behind the router sends a packet, the router re-writes the packet to use the router's IP address as the sender, then transmits onward. And returning packets are re-written and re-transmitted by the router to the device. The router keeps a translation table between its actual Internet address, and port, to the local (LAN) destination device and port. Theoretically, each router could give out an entire Internet's worth of IPv4 addresses, treating them as a private LAN, and doing packet address translation every time a packet enters or leaves the private LAN.
Slight correction on the address range, it uses a 16-bit netmask (not 8) and starts at 192.168.
is this the last episode in this great series on Computing Science? Its been good viewing them, look forward to others in your series eg the history one
Is there a crash course like this for hypervisor engineering?
8:19 "It's like the phone book for the Internet"
Zoomers: Phone book? What is that?
Awesome.
Why did I go to college when I could’ve just watched RUclips videos which I’m doing anyway?
Lol now I must watch all these 12 min videos for my project of my English course
So good. I learn so crazy much
*mac oss*
It's like gooey all over again...
1:40 - uh oh.. prepare for some spam packets.
How'd it work in 90's? when you just pluged into the wall? freedns?
Thank you for all the knowledge. Learned a lot.
Wish these were around 10 years ago!
Could you make a course on statistics as well? It would be really useful.
Why do you have an oscilloscope for a computer science series?
Awesome job explaining this 👏
Incredible this helped me a lot
2:20
Bold of you to assume that I live there.
"dobledoo" wasn't that from John Green in Vlogbrothers? lol
CrashCourse was started by John and Hank Green. John has presented multiple series about World History and Literature, while hank has presented series on Biology, Ecology, Chemistry and Philosophy.
WheezyWaiter started that. He is friends with John and Hank. John and Hank created CrashCourse