#vi talks computers

[virescent-phosphor]

Wonder if I should be a good noodle and do my MCSA homework, or a naughty noodle and take apart a vintage Mac I need to fix :^)

[virescent-phosphor]

Just bought a used Nook at a yard sale to download PDFs and ePubs onto; realized/found out support might be being dropped for this model. (But I’ll figure something out if that’s true anyway.)

Still considering buying a 90s digital camera to take traveling despite how goddamn impractical it is.

[virescent-phosphor]

Everyone who programs likes to kinkshame me for saying I’m gonna learn Pascal

[virescent-phosphor]

@yoddel replied to your post “I have a big exam coming up does anyone want to help my study/learning...”

So when you have two computers using the same router to connect to the interwebz, how does the router know where to send the interwebz it gets back?

So my answer will hopefully mostly answer this well/correctly...

How to Interwebz Back to the Right Computer: A Summary

When you connect to the Internet through a residential gateway (fancy proper name for the home integrated router/modem/etc. you have, corrected from incorrect term) you have two addresses: a private address, which is not routable to the Internet, and an external address which shows to the public web.

Private addresses are in certain blocks; for IPv4, these would be addresses like 192.168.1.10, 172.16.0.3, and 10.0.0.1. They cannot be routed out to the internet. Your public address is assigned in blocks by an authority, depending on the area (IANA is the big org,) and you probably don’t actually own your own, only lease from an Internet Service Provider (ISP) Your private address is used within your local network to connect with others on the same network.

Generally speaking, your ISP has a big pool of public addresses which it leases out to users. When you make a request to an IP address outside of your private network, or LAN, the router will translate your private IP address to a public IP which is assigned for your usage. This is called NAT, or Network Address Translation.

But you’re sharing a single address with multiple hosts, and using multiple services. How will you handle sending that information to the correct hosts?

This is where PAT, or Port Address Translation, is needed. PAT is sometimes called “NAT overload,” as it’s meant for connecting several internal devices/addresses to the “outside” with a single (or limited number) public address.

To track all of these requests, a source port number or a query ID (ICMP) is generated. This is your source because you are originating the request. If there is another host using the same source port, PAT will methodically change the port number to ensure uniqueness. Each of these will be unique to identify an individual session.

The transport layer of the TCP/IP protocol suite is what actually handles the transfer of data/sessions. That is what TCP or UDP does, which is probably out of the scope of your question. The IP layer is associated with IP addresses, as the transport layer is associated with ports. Transport will identify the individual conversations taking place between applications and ensures the data stream is properly segmented (broken into pieces for reassembly.)

For example, one of the best known ports is port 80, the HTTP port. When you request a (non-secured) webpage, such as Google’s homepage (http://google.com,) you send a request to port 80 on one of Google’s servers. That would be your destination port. The reply will not be sent to port 80 on your end, because you have a dynamically generated port address!

There’s a lot more to the process of establishing a session, but I think this is all you need, plus some extra! To properly understand all of this, it helps to understand what a protocol stack is, particularly TCP/IP. You’d also asked about OSI, so I might talk about that next.

So, more or less, using these source ports and matching them to your internal IP address (the combined identifiers are called a socket pair,) your router is able to differentiate the requests you’re making to the outside network from one another. :)

[virescent-phosphor]

@rudescience replied to your post “I have a big exam coming up does anyone want to help my study/learning...”

I am curious about ethernet cables. Why are there so many? What do the individual colours indicate? Is this a good question?

I’m not sure what you mean by “why are there so many?” Do you mean types or number? CATs?

There are a ton of categories of UTP, or unshielded twisted pair, cabling. CAT5 is the lowest category Ethernet uses; CAT3 is used for voice calling on POTS, I believe. CAT5e and CAT6 are two higher standards, and today, most would say to start at 5e, not 5.

“Fast Ethernet” is a standard which starts at 100 Mbps, and is the lowest standard people use today. Gigabit Ethernet is common, and we have 10 and 100 Gbps as well.

Shielded Twisted Pair (STP) or fiber optic cabling are also used, but UTP is the most common and recognizable, and terminates with this familiar little guy:

So for colors, those are cable pairings: remember that “pair” in “twisted pair”.

Before they hit the plug, in the jacket, this is what your cable looks like:

These are twisted together to minimize crosstalk. But why do the colors matter? Well...

There are two main pinouts based on the cable colors. In straight-through cabling, the ends of the cables must both be the same type. In crossover, one end has to be T568A, and the other must be T568B.

Crossover is used to interconnect similar devices: switch to switch, or host to host. Straight-through is used to connect dissimilar devices. However, thanks to Auto-MDIX, which most modern equipment can have, you can generally use either type and have the equipment configure itself accordingly for the cable you plug in.

I’m unfortunately not familiar with what each pin individually does, but they would stand for TX+/TX- (TX=transmit), RX+/RX- (RX=receive), etc. Physical layer is admittedly not my forte, although I do remember researching a bit on X.25′s cabling for an assignment.

Jacket colors, on the other hand, are not standardized in regular use. There are apparently standards, but doubtful anyone follows them; instead, users and businesses may color code them by type of cable or what they are used for to keep things organized. My lab uses orange crossover cables, for example.

(I should review the course I took with more info on physical stuff when I have the time!)

[virescent-phosphor]

I have a big exam coming up does anyone want to help my study/learning experience by asking questions about basic networking?

(I’m not totally being facetious; I retain info through discussion and teaching others)

Ever wondered what the hell the OSI model is? How do cables actually work? Addressing? There you go!

[virescent-phosphor]

I didn’t remember you’re supposed to call a wireless interface connector for USB a wireless adapter, so I thought “NIC dongle” and then, “let us eat at NIC dongle”

time to phase “dongle” out of my vocabulary before I use it irl in a school or work environment

[virescent-phosphor]

Gonna start appending "am I right, ladies?" to my hashtag relatable tech posts and complaining tech posts to see how people react

Such as: "CDP is a bunch of bullshit, am I right, ladies?"