25 Feb
2011
25 Feb
'11
11:12 a.m.
On Fri, Feb 25, 2011 at 1:50 PM, Shoppa, Tim <tshoppa at wmata.com> wrote:
I came from the academic side where Ethernet
networking had been standard for some time by the mid-90's and getting networks of
VAXen and Suns to interoperate was a breeze.
When you were doing that, I was in an environment of VAXes running VMS
and Ultrix linked with async lines (lots of Kermit going on ;-) and
sync lines (some DDCMP/DECnet and *lots* of Bisync/HASP) since we
manufactured Bisync products and had a shelf of sync modem eliminators
in our comms rack. When the company flared out in 1993, there wasn't
a single Ethernet NIC nor an inch of coax or twisted pair anywhere to
be seen.
Coming from that, I was happy to jump into an environment where TCP/IP
could be found.
In contrast the mid-90's era focus by consumer
OS's (read that as "Windows and bolt on products") on the modem as the
lynchpin in networking, was simply bizarre to me.
Yeah... in the late 1980s, my modem was busy pumping UUCP up and down
to my Amiga (I helped debug one of the issues with the HDB UUCP port).
By the mid-1990s, I had my first TCP/IP network going between the
Amiga (via A2065 Zorro-II Ethernet NIC) and a surplussed SPARCstation
1. The modem was still doing UUCP duties until 1996 when I found a
device driver for AmigaDOS (telser.device) that emulated a modem over
a TCP/IP socket (we had a local "club" that had set up a UUCP calling
pyramid, and it was still handy to be a part of that).
A very different networking world from the stuff that Byte magazine
was covering.
-ethan