568A or 568B?

A on one end, B on the other. That way you get the benefits of both. :^)

>>57966656
It doesn't matter as long as the ends are the same.

Makes no difference. Just do it the same as the rest of the patch panel.

Solid is for in-wall, stranded is for standalone cables. You can use solid for patch cables, but you'll ruin a few ends.

>>57966656
WHY THE HELL ARE THERE TWO DIFFERENT WAYS?

>>57966740
Both ends the same (either A or B) - standard cable, to connect a switch type device to a client device.
One end of each type - crossover cable, to connect two switches or two client devices directly.
Less critical nowadays now everything has auto MDI/MDIX

>>57966656
It really, really doesn't matter. You could have a 4 port LACP uplink with 3 type A cables and one type B and it wouldn't give a shit.

Crossover is legacy shit these days, straight through will work just fine unless you are doing CCNA stuff. Computers are smart enough to flip the wires if they need it.

If you are going to put copper in the wall and have the money, go for future proofing if you think you will still be there when 10g is cheaper.


Im not 100% sure, but I would assume that stronger wire is used for in wall installations, you can use whatever inside, you can just swap it out later if you need to.

A-B between 2 computers
A-A between a switch and a computer
A-B between 2 switches

Some (most new) NICs can 'twist' them themselves so it doesn't matter that much anymore, but when they can't above the only correct way to do it.

>>57966656
T568B because this is what we use in France.

>>57966656
I personally 568B for my straight through cables, but it doesn't technically matter as long as both ends are the same. It's best to stick with whatever convention they are using at the site that you are wiring, switching shit around on them will really piss people off.

Solid is for cable runs between patch panels & jacks, stranded is for patch cables. You need solid for anything that gets punched down, stranded punched in patch panels or modular jacks will not make as solid a connection and cause problems down the line. Stranded is supposedly better for patch cables as it is more flexible and less likely to break due to fatigue. however unless you are hooking up your paint mixer to the internet, I doubt you will ever have problems using solid. If I need to make a patch cable, I just use solid, because there is no point in buying a second roll of stranded just to make patch cables when you should just buy pre-made patch cables instead. As someone who has been wiring this shit up since the 10BaseT days (with motherfucking hubs, not switches), spend the extra to get the latest standard of cable you can for your runs between patch panels & your jacks, and just buy patch cables in whatever your using today. When the time comes to upgrade to 10G or beyond, you'll need to upgrade the NIC / switch at that time, and can just put a new patch cable in then, but you sure as hell don't want to have to pull a new cable run if you don't have to.

>>57966661
That's a crossover cable which you would need to connect two NIC/router or two hub/switch ports where auto-MDIX is not supported. The question is which one to use for straight-through cables.

>>57966663
I know it doesn't, but you should pick one for consistency (let alone OCD/autism etc.).

>>57966998
>You need solid for anything that gets punched down
Really? If you want to make your own cable from UDP cable bought by the meter, you necessarily have to use solid? Is stranded only available as pre-made patch cables?

>>57967139
Punched down != crimped, retard.

>>57967061
>but you should pick one for consistency
Why? It literally doesn't matter.

>>57967278
Because when you are working on site it's very rare that you can see both ends of the cable at the same time, and when you are trying to sort out some cabling gremlin, you are looking for anything you can find that is out of the ordinary.

Of course that doesn't work WHEN THERE IS NO FUCKING ORDINARY TO BEGIN WITH!

Use some fucking consistency, savages.

>>57967364
>diagnosing things based on assumptions

>>57967478
There is no other way. You are assuming your diagnostic equipment is functioning correctly, and no one else has left their own tone on whatever cable bundle you are probing.

>>57967364
This. If you see a cable end wired the other way you're using as as standard for straight-through, you immediately know it's a crossover cable. If you don't have a standard, it could be either.

What about gigabit crossover? The A vs B standards only swap the 2nd and 3rd pair used by (fast) ethernet, but gigabit uses all four pairs. Is it necessary for a crossover gigabit-compliant cable to swap pairs 1 and 4 on one end as well? Or do all gigabit interfaces support auto-mdi/mdix so that actual gigabit crossover isn't ever necessary?

>>57966998
It would be nice to put cat7 everywhere, though in real life often it's still kinda hard to even overcome the "cat5e oughtta be enuff for anybody" attitude when the budget is being defined. I think that 5e won't be phased out anytime before 10gigabit starts becoming widespread.

Why the fuck is there two types

What's the point?

>>57969441
Apparently it's mainly historical reasons, that's one thing (but unfortunately authors of sources both printed and internet are in great disagreement about the details). The other thing is crossover cables, for 10/100 Ethernet you have to swap pairs 2 and 3 if auto-mdi/mdix isn't available.

>>57967048
you've got a little autism there

>>57969441
A-A straight
A-B crossover
B-B don't, please.

Can european and american anons confirm which one they have seen implemented most?

Poll:
http://www.strawpoll.me/11863144
http://www.strawpoll.me/11863144
http://www.strawpoll.me/11863144

>>57969911
B-B and A-A are literally the same thing

The cables can be in any order so long as the order is the same on both sides

>The standard defines two alternative pinouts: T568A and T568B. The pinout definitions occupy merely 1 of the standard's 468 pages. Much attention is paid to them because cables do not function if the pinouts at their two ends aren't correctly matched.

>TIA/EIA-568 recommends the T568A pinout for horizontal cables. This pinout's advantage is that it is compatible with the 1-pair and 2-pair Universal Service Order Codes (USOC) pinouts. The U.S. Government requires it in federal contracts.

>The standard also allows the T568B pinout, as an alternative "if necessary to accommodate certain 8-pin cabling systems". This pinout matches the older AT&T 258A (Systimax) pinout. In the 1990s, when TIA/EIA-568 was published, 258A had the most widely installed UTP cabling infrastructure. Many organizations still use T568B out of inertia.

https://en.wikipedia.org/wiki/TIA/EIA-568

So it is encouraged to use the A variant for all new installations in America? What about Europe?

>>57969951
>B-B and A-A are literally the same thing

Electrically yes, but then there's the color coding which also exists for a reason (i.e. it functions as wire labeling). You should pick just one and stick with it, using the other only for crossover cables if necessary. Also see >>57969974.

>>57969951
Theoretically yes. But, now imagine if every cable manufacturer would use their own set of colors, and every network deployment technician wired them as they pleased. Would that be helpful, or rather not at all?

I use B. I remember that one better for some reason. It's the first I learned and I'm using it.

>>57970523
If you remember one "better" than the other, it probably means that you would struggle with making a crossover cable (as you apparently don't recall the simple pair-2-and-pair-3-swap principle that's literally the only difference between the two).

>>57969144
Someone please answer this? Do the blue and brown pairs need to be swapped too at one end to get a properly functioning gigabit crossover cable?

File: banana-crop.jpg (16KB, 700x320px) Image search: [Google]

16KB, 700x320px

Which end are you supposed to eat this thing from? Everywhere (both in books and on the internet) one tends to find conflicting information as to which one has been, is, and should be preferred, in which regions, and why(?) - literally every source seems to have a different idea of the above. Obviously it's good to decide on one, even if it shouldn't matter too much, but among the confusion the choice seems not easy.

>>57970775
If you read >>57969974, you would see that it does appear to matter.

>>57970807
are you doing contract work for the US Government? otherwise I don't think it really matters

for what its worth I do remember reading that T568B is the most common wiring type here in sweden, alltho I'm not sure where the book sourced that information

>>57970705
there is no "Gigabit Crossover". gigabit NICs simultaneously send and recive across all pairs at the same time and calculate the signal difference between the sent and recived to reconstruct the data sent by the opposite device. The paradigma "2pairs send, 2 pairs recieve" is not used anymore these days

>>57971544
Oh, ok, thanks for the explanation. Will a crossover (one and A, one end B) cable work at all with gigabit?

http://www.phonegeeks.com/56vs56whshyo.html

>If the installation is residential, choose T568A unless other conditions apply (see below). The two inner pairs of 568A are wired the same as a two-line phone jack.
>If there is pre-existing voice/data wiring (remodel, moves, adds, changes), duplicate this wiring scheme on any new connection.
>If project specifications are available, use the specified wiring configuration.
>If components used within the project are internally wired for either T568A or T568B, use that wiring scheme.
>Make sure both ends of a cable are wired the same way.