From: SMTP%"RELAY-INFO-VAX@CRVAX.SRI.COM" 7-AUG-1993 12:20:16.28 To: EVERHART CC: Subj: PDP-10/DEC-20 emulator status From: Mark Crispin X-Newsgroups: comp.os.vms Subject: PDP-10/DEC-20 emulator status Date: Mon, 2 Aug 1993 22:11:51 -0700 (PDT) Organization: University of Washington Lines: 57 Message-Id: Nntp-Posting-Host: ikkoku-kan.panda.com Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII To: Info-VAX@kl.sri.com X-Gateway-Source-Info: USENET I was warned that I was mentioned in this newsgroup in conjunction with the C PDP-10 emulator...sure enough, the messages are coming in. Maybe if I post an answer here, I won't have to answer 69 additional times... :-) There are no less than three independent efforts in progress to preserve PDP-10 (also known as DEC-10 and DEC-20) operating systems and software. First, XKL Systems in Redmond, WA is a startup founded by many of the people who started cisco (the terminal server makers). They are building a new PDP-10 machine. Try: Means@Mathom.XKL.COM and/or Bosack@Mathom.XKL.COM. I have had a few contacts with Dave Means but am not directly involved with this project. Second, (Ms) Pat Tressel at the Locke Center of the University of Washington is working on a PDP-10 emulator for the Alpha in assembly language. There are a bunch of crazy TOPS-10 hackers over at the Locke Center. Unfortunately, Pat is getting laid off this month and was looking for new work. One possibility is to get funding (= salary and the loan of an Alpha) from DEC, but I haven't heard from Pat in about a month so I don't know what the status of that is. Pat can be reached as Pat@Locke.HS.Washington.EDU. Third, Ken Harrenstein has written a KS10 emulator in C which runs either ITS or TOPS-20 (you have to decide which one you want when you build the emulator since the microcode was different on a real KS10). Ken wrote it for the SPACE using gcc, and I have ported it to the NeXT (trivial) and the Mac (not so trivial). I have sent all my changes back to Ken, who is working on a KL emulating version. Ken says that on the SPARC it actually runs faster than a real KS10. My experience on a NeXT (68040, 25MHz) is that it is about 2.5 times slower. In the Mac world, a Quadra (68040, 33MHz) is about 4 times slower (polling for I/O *does* cost!), a PowerBook-165c (68030, 33MHz) is about 12 times slower, and bringing up the rear, a PowerBook-100 (68000, 16MHz) is about 65 times slower! A 486 (33MHz) running BSDI is about 1.5 times slower (Peter Lothberg in Sweden did the BSDI port). These speed tests are the time it takes to compile/load/execute a MACRO program to output ``Hello, world!'' and return to the @ prompt. Ken's e-mail address is klh@us.oracle.com. I do not know when he will be ready to release. He is working on the KL emulating version and has undoubtably fixed many bugs. I won't give out copies of the version I have without explicit permission from Ken. Also, the TOPS-20 list is now located at Panda.COM, since SIMTEL20 is scheduled to bite the dust at the end of September. The request address is TOPS-20-REQUEST@Panda.COM (or just MRC@Panda.COM if you feel lazy). Hmm...since this is the VMS newsgroup. I wonder if the VMSites are starting to feel the way we did back in 1982 with everyone claiming that our favorite operating system is on the way out... If so, I wonder how long it will be before there is a VAX emulator. I suspect it would be a lot harder, since it is fairly easy to make a fast implementation of the PDP-10 instruction set and not so easy to make a fast implementation of the VAX instruction set. Fortunately, a lot of your stuff is written in BLISS, so perhaps you aren't so tied to the VAX as we were to the PDP-10......