Idle question

Can anyone offer an explanation for the ages-back decision to use two
different pin alignments for CAT-5/RJ45 networking plugs/sockets? You know
the way you need a crossover when the alignments are identical and a
straight through when they're mirrored?

Could it have been as simple as someone just imagining that all cables
would be straight-throughs because no one would deviate from the standard
pattern of router->[switch/hub]->Host?

I still remember my CCNA tutor having to stop and think when I pointed out
that if all sockets had the same pinout and all cables were crossovers
then there'd never be any problem..

On further thought, I can see that even under my supposed 'common sense'
approach, you'd still sometimes want straight-throughs for extension
purposes. Hmm, actually, no, you wouldn't would you, just make the inline
connectors also all crossover. So an extended link would be
[crossover-cable]-[crossover-inline]-[crossover-cable] and it would still
work perfectly.

So, what was the thinking? Does anyone have an idea?

Dave J.

 

No real answer, but I have no problem with the setup because I come from
the DCE/DTE era (Data Communications Equipment - eg modem - and Data
Terminal Equipment - eg teletype [yes, I know teletypes also used 80v
signalling for some versions]) where the interfaces were designed for
straight though cables as required by the ISO (also EIA and ECMA)
standards

In networking, the switch (or hub) is DCE and the computer DTE, so logic
would suggest a continuation of the practise.

The more confusing issue IMHO is the cross-over cable crosses two pairs
and leaves two pairs straight through :-(

--
PeeGee

The reply address is a spam trap. All mail is reported as spam.
"Nothing should be able to load itself onto a computer without the
knowledge or consent of the computer user. Software should also be able
to be removed from a computer easily."
Peter Cullen, Microsoft Chief Privacy Strategist (Computing 18 Aug 05)

 

No real answer, but I have no problem with the setup because I come from

Ah the good old days.....never went anywhere without a null modem....

 

I reckon you probably invented firewire without knowing (which, btw,
work great for networking under XP.)

 

I take it Firewire is "all cables crossed" then. Not computer-related, but
SCART is like this too (for the audio in/out and CVBS in/out).

Back to Ethernet, links made between punch-down blocks would need to be
crossed too. That's fine if one end is in a wiring cupboard (or whatever)
and the other inside a wall-mounted socket in an office, but is that always
the case? (I have no idea, but if not and you forget which way you wired the
first end...)

Alex

 

Yes, IEEE1394 cables are all cross-wired:
http://www.technick.net/public/code/cp_dpage.php?aiocp_dp=pinconbus_ieee_1394
(scroll down to Cables and connectors, if interested.)

 

That makes sense. Still doesn't quite make it correct, but it does make
sense.

Oh? Does the DCE/DTE standard have more than the data cables reversed?

Dave J.

 

DCE/DTE, as I used the terms above, are part of the ISO/EIA/ECMA
standards (the latter are virtually identical) and the standards defined
the pin-outs for the equipment. IIRC, apart from the physical
obstructions, the connectors were defined so that the two units would
plug in directly - the cable, therefore, was straight-through extension
lead.

The ethernet cross-over patch lead is confusing because the two data
pairs for 10 and 100Mbps are crossed over, the other pairs are not. If
you use such a cable with Gigbit ethernet, two data pairs are crossed,
two are not :-(

Thankfully, unless you are using old equipment (especially hubs) the
interface is auto-sensing and the patch lead is no longer a factor.
Cross-over leads are largely redundant unless you wish to connect two
computers directly, when auto-sensing is (or is that was?) not common
for NICs.

--
PeeGee

The reply address is a spam trap. All mail is reported as spam.
"Nothing should be able to load itself onto a computer without the
knowledge or consent of the computer user. Software should also be able
to be removed from a computer easily."
Peter Cullen, Microsoft Chief Privacy Strategist (Computing 18 Aug 05)

 

IIUYC, that had, umm.. crossed my mind..

My thought was that the standard could cope with that by defining
plug<->plug as always crosover but (plug or device or socket) <-> (socket
or panel) as always straight through.

IOW as opposed to straight-through/crossover there would be 'connector
cable' or 'extension cable' - one defined as for connecting two standardly
wired sockets together and the other as extending a socket, the former
being always crossover and the latter being always straight-through. No
difficulty telling one from the other where it's cables and easy enough
for technicians to get used to as the wiring from devices to sockets or
patch panels would be identical.

I dunno, I meant it as 'idle question', just me reflecting my musings and
wondering if anyone else had thought similarly.

Dave J.

 

Firewire's something I know little about, for some reason (don't quite
know why) it's always come across to me as one of those gimmicky things
that's implemented on a few devices but is superseded within months or a
couple of years. Or, in some examples, before it's even released..

Dave J.