[email protected] (Paul Gilmartin) writes:
> Why did that fail?  Just too little, too late?  NIH?

re:
http://www.garlic.com/~lynn/2014b.html#39 Resistance to Java
http://www.garlic.com/~lynn/2014b.html#44 Resistance to Java

internal network was larger than the arpanet/internet from just about
the beginning until sometime late '85 or early '86.  Part of the reason
was that the (vm370 vnet-based) internal network had a form of gateway
in everynode (which didn't exist in the arpanet host protocol, sna,
and/or osi model). Arpanet/internet got internetworking protcol as part
of the great change-over to tcp/ip on 1jan1983.

this "ibm-main" discussion group originated on univ. "bitnet" which used
technology similar to the internal network ... and about the same time
the communication group was forcing the internal network to convert to
SNA ... bitnet was converting to tcp/ip for bitnet2 ... which is what
the internal network should have done also.
http://en.wikipedia.org/wiki/BITNET

i had project i called hsdt with T1 (1.5mbit/sec) and faster speed
links ... started doing T1 in 1980s.
http://www.garlic.com/~lynn/subnetwork.html#hsdt

the initial mainframe tcp/ip product (mentioned upthread, implemented
in vs/pascal) has some performance issues .... getting about
44kbytes/sec effective using nearly full 3090 processor. I did the
enhancements to support RFC1044 and in some tuning tests at cray
research got 1mbyte/sec channel sustained throughput between
4341 and cray using only modest amount of 4341 processor (possibly
500 times improvement in bytes moved per instruction executed)
http://www.garlic.com/~lynn/subnetwork.html#1044

standard mainframe 37xx product only supported up to 56kbits/sec.
communication group prepared a report for corporate hdqtrs that
customers didn't want much more than 56kbits/sec and wouldn't need T1
(1.5mbits/sec) until well into the 90s. At the same time, HSDT didn't
superficial customer survey and found 200 customer T1 links connected to
IBM mainframes (but using non-ibm controllers).

The communication group kept up the facade for a time ... but eventually
was forced to came out with the rube-goldberg 3737 hack supporting
T1. SNA/VTAM had a separate problem with latency handling in
communication links. It was unable to get only a very small fraction of
T1 throughput capacity.  To get around it the 3737 supported a CTCA link
to another local mainframe. Inside the 3737 it had a bunch of processing
and enormous amount of buffering (including four 68k, 100k lines of
code) ... it would simulate ACK on data RUs to the local VTAM (as if the
data had already arrived at the remote end) trying to compenstate for
the lack of latency compensation in SNA & VTAM. some past posts
detailing 3737
http://www.garlic.com/~lynn/2011g.html#75 We list every company in the world 
that has a mainframe computer
http://www.garlic.com/~lynn/2011g.html#77 Is the magic and romance killed by 
Windows (and Linux)?
with these old email
http://www.garlic.oom/~lynn/2011g.html#email880130
http://www.garlic.oom/~lynn/2011g.html#email880606
http://www.garlic.oom/~lynn/2011g.html#email881005

the claims were the 3737 could just barely support 1.5mbit/sec
throughput ... even though full-duplex T1 is 3mbits/sec aggregate
(1.5mbit/sec concurrent in both direction) and european T1 is 4mbits/sec
aggregate

other trivia ... I had been blamed for online computer conferencing on
the internal network in the late 70s and early 80s. folklore is that
when the corporate executive committee was informed of computer
conferencing (and the internal network), 5of6 wanted to fire me.
http://www.garlic.com/~lynn/subnetwork.html#cmc

old email from person charged with setting up EARN (bitnet in europe)
looking for network apps
http://www.garlic.com/~lynn/2001h.html#email840320
posts mentioning bitnet
http://www.garlic.com/~lynn/subnetwork.html#bitnet

as part of HSDT, we was also working with NSF and NSF supercomputer
centers.  We were suppose to get $20M from NSF to tie together the
supercomputer centers, then congress cut the budget and several other
things happened.  Finally NSF releases an RFP ... but internal politics
(large amount from communication group) prevent us from bidding. The
director of NSF tries to help by writting the company a letter
(co-signed by some other agency CTOs) ... but that just makes the
internal politics worse (as does comments that what we already had
running was at least five years ahead of all bid submissions).

-- 
virtualization experience starting Jan1968, online at home since Mar1970

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to