I have one for the collective conscious alien brain of perkiset.org:

How come i cant xfr even with a 10/100 crossover factory or
self crimped wire more than like 5mbps.

Anyone ever like, ftp'ed something on a local

net

 work breaking
the initial four wire CAT 5 claimed 10mbps?

Going way back on that one, but doesn't 4 wire make the connection half duplex? I think you need all 8 to make it full duplex 10 or 100

You need only four wires for cat 5 10mbps cable, eight for 10/100.

I just have never personally or have ever heard of anyone actually
reaching ten, but this bogus 54.0mbps connection speed on my wireless
card makes me wonder.... cause nobody ever REALLY hit 56k on a kflex
modem either, cause that shit is sampled and digitzed at the phone switch on the street
at 36.6 to fit in the csu/dsu/t1 thingy and you only get 44k per slice
in that box... so anything over that was like 'unzipped' on the other side
and the calculation then done on the realized transfer speed.

I set up an ftp server on one

mac

 hine, and using gftp on the other
tried both uploading and downloading a file on a static ip

net

 work
of just those two computers connected with a 10/100 crossover
cable and still only hit like 5000k .... 5mbps pretty much.

Of course. But if you only use the 4 wires and you sent a packet and handshake it in 10M then you are effectively only transferring at 5M in each direction. It's 10M but since both packets are coming across the same 2 pairs then you cut your total transfer in half. I'm only about 75% sure on that one tho. Perhaps a streaming protocol would do better since there is no handshake like TCP.

The Perkster is correct re: half duplex and the slowdown caused by packet exchange. Connectionless protocols can achieve full rated speeds on half duplex connections afaik, but I have never tried it either.