Talk:SUN workstation

Discussion
This was from the Sun 1 page where it was incorrect, Sun never used multibus memory as far a I know. I am stashing it here so it does not get lost:

RAM could be increased by installing two Chrislin CI-8086 128 or 512 KB memory expansion boards

Robert.harker (talk) 09:06, 25 April 2008 (UTC)


 * The CPU board had memory on it. Should mention clearly each workstation had three boards: CPU (including memory), graphics, and Ethernet. So this was an irony of the project: a standard bus was used, but ended up being all custom boards. In fact an anecdote I remember from the Stanford Office of Technology Licensing, was using the standard bus as an argument that there was no "innovation". Should mention the legal battle with Cisco, of course. I was part of the team, but cannot rely on first-person memory of course. If I have the time I will try to dig up the refs. W Nowicki (talk) 03:08, 21 April 2009 (UTC)

Also some misleading information here. Ralph Gorin was an educator with a technical background, but not a hardware designer. He had an important role, which was mostly in providing requirements, getting funding, etc. There are a couple sources, so will try to rectify. Also misleading comment about "fellow graduate of Stanford" since Andy graduated from University of Technology Munich and CMU. Eventually he might have gotten an Engineer degree I thought, but never finished his PhD. W Nowicki (talk) 00:40, 11 April 2011 (UTC)


 * "The SUN workstation is a modular personal computer system designed for use in an Ethernet-type local network. A SUN workstation provides a single user with significant local computing power, a high resolution graphical display, graphical input, and network communication [Bechtolsheim]. The SUN station capabilities can be realized at a cost of approximately $10,000 per station, using commercially available VLSI components."


 * "The S-1 design system was constructed to support the development of the S-1 processor. The S-l design system consists of: SCALD (Structured Computer Aided Logic Design) System — a hierarchically structured logic design system. SUDS (Stanford University Drawing System) — a graphics system, used to edit, input, and output to/from SCALD. A timing verifier — based on SCALD and designed to analyze and verify timing specifications and constraints in a digital system. The physical design system — consists of both placement tools (chip placement) and wirewrap technology tools."

These quotes seem out of place and redundant with the body. Just paraphrase would read better. And indeed, Andy never graduated from Stanford. Will work on it a bit today. W Nowicki (talk) 22:40, 1 May 2011 (UTC)

Terminal server description is misleading
The sentence about the TIP terminal server configuration is misleading, but I'm not familiar enough with the configuration to know how. It says:

"For example, the CPU board combined with a multi-port serial controller created a terminal server (called a TIP) which connected many terminals to the Digital Equipment Corporation time-sharing systems at Stanford or anywhere on the Internet."

The question is what was the backhaul path to the DEC minicomputer? If it was prototype Ethernet at 2.93 Mbps, the sentenced fails to mention the Ethernet board along with the CPU and multiport serial board. If the backhaul to the DEC was one or more serial ports, the sentence is misleading in suggesting a TIP user could connect directly anywhere on the Internet -- they would be connecting to a DEC minicomputer, and only indirectly using telnet on to other Internet hosts.

My guess is, as with later terminal server designs, Ethernet was the backhaul, but I can't rule out a high-speed serial backhaul, particularly if the TIP terminal-facing serial ports were connected to modems, which were 300 bps in those days.

184.78.246.47 (talk) 12:00, 26 November 2012 (UTC)