home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.sgi
- Path: sparky!uunet!ukma!wupost!decwrl!sgi!rhyolite!vjs
- From: vjs@rhyolite.wpd.sgi.com (Vernon Schryver)
- Subject: Re: DECSERVERS with UNIX
- Message-ID: <op4u6so@rhyolite.wpd.sgi.com>
- Organization: Silicon Graphics, Inc. Mountain View, CA
- References: <86489@netnews.upenn.edu> <19AUG199210035775@jhuvms.hcf.jhu.edu>
- Date: Wed, 19 Aug 1992 14:56:12 GMT
- Lines: 30
-
- In article <19AUG199210035775@jhuvms.hcf.jhu.edu>, ecf_stbo@jhuvms.hcf.jhu.edu (it takes a bigger man to laugh at that man...) writes:
- > ...
- >We are running LAT (from SGI) on an SGI box. I'm not directly involved so I
- >don't know intimate details, but there were some initial problems. They seem to
- >have fixed most of them. We have LAT/TCP-IP terminal servers and running tcp-ip
- > they were busying our ethernet with lots of little packets. LAT multiplexes a
- > bunch of users keystrokes into the same packet. The downside is that SGI's LAT
- > daemon has to look at every packet, but with four processors we have cpu to
- > burn.
-
-
- I'm glad our LAT implementation is working for you. The people who
- worked on it worked hard. (not me.) Please continue complaining to
- the TAC when it doesn't work.
-
- However, the old story about LAT being significantly more effecient on
- the wire than Telnet was false back when DEC started it. The number of
- packets from interactive traffic just doesn't amount to a big part of
- an ethernet when that ethernet is used for data moving. Perhaps LAT
- was needed by DEC back when VAX's could not discard 100 p/s. (Recall
- the stories of what dogfight used to do to corporate networks.) The
- only reason to run LAT is because you need to talk to existing hardware
- or software, or because you can get much better deals on LAT terminal
- servers than TCP/IP terminal servers. There are a bunch of reasons not
- to use LAT if those reasons do not apply, starting with the fact that
- LAT is not "route-able." As an abstract protocol, LAT is widely used
- among protocol designers as an example of what not to do.
-
-
- Vernon Schryver, vjs@sgi.com
-