On Sat, 5 Nov 2011, Matthew Hudson wrote: There's still a VAX 6000 at CSU Bakersfield and the head of the CS department isn't letting it go anytime soon. -- David Griffith dgriffi at cs.csubak.edu A: Because it fouls the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail?

On Sun, Nov 6, 2011 at 4:03 PM, Gene Buckle <geneb at deltasoft.com> wrote: Nice rig. I still have my 8300 in the basement that I rescued when my employer closed in the mid 1990s (we bought it as a bare 8200, second-hand for $13000, then I nabbed a compatible CPU module for around $100-$200. For disk, it has one RA81 and one MDA ESDI-SDI disk box. I wish I could get the DWBUA working on it, but I've gotten stumped by what looks like self-test failures talking to the UET module at the far end of the Unibus. I have multiple UETs, but until I find a way to test them outside of the DWBUA, I have no confidence any of them work (or might not require an ECO specifically to work with a DWBUA). The machine runs fine - just no Unibus. -ethan

On 11/06/2011 06:51 PM, Ethan Dicks wrote: I still have the microVAXII that was under my desk at DEC still running VMS-5.4. That and a bunch of 3100s (m76s and some others) all running and a few mV2000 running and as formatters. The BA123 uVAXII has a RD54 and CMD200 with 2 RZ56s. The 3100s have 1gb baracudas and a few odd 4.3gb and some 400mb SCSIs internal along with external BA42 storage boxes with 2 RZ56s each. Storage is not an issue but space, power and heat tends to pop up if I run them all. When all running they are the AIRPRT: LAVC cluster of 10 nodes. My collection is anything CP/M, PDP-8e, All manner of Qbus PDP-11 and all those VAXen. There are a smaller collection of SBCs (SC/MP, ELF, 6800D1, Ti9900(technico), Super starter kit 6100, TK80, SDK85, and IMSAI IMP48 and a few others). In the oddball collection is my EELF (Spare time gizmos) Embedded elf fully loaded with disk and IO in a case with a CRT and accepts a PS2 keyboard. Allison

On Mon, Nov 7, 2011 at 9:32 AM, Gene Buckle <geneb at deltasoft.com> wrote: Yes. The 8200 and 8250 have one CPU, the 8300 and 8350 have two. Unlike later models, these CPUs sit right on the VAXBI bus. When these machines came out, DEC provided the most obvious native peripherals (KDB50 disk, DEBNI ethernet, DMB32 serial mux, CI interface, etc) but there were so many oddball Unibus cards out there (including the ones we sold) that another peripheral was the DWBUA BI-to-Unibus adapter. Some models were ordered and shipped with one BI section and one DD11DK and a DWBUA with short cables in the BA32 chassis, but what we had was the more common version - two BI sections in the main BA32 and long cables over to a BA11 in the next rack over. In my case, it was all working in its former home, but something happened to it when I moved it and the DWBUA has never worked since. The BI portion all works - memory, disk, dual CPUs, etc., but the amber LED on the DWBUA card stays off and the Unibus isn't seen by VMS. I have the maintenance manual for the DWBUA - I've been through the register poking at the chevron prompt - it appears to be a failure of one of the many tests the DWBUA performs before coming online of a kind where it is reporting some not-as-well-documented fault with the UET module on the far end of the Unibus. I could be having a cable fault, a card fault, or a backplane fault. In the end, the DWBUA performs some reads and writes and is not getting the answers it's looking for. I have more than one UET module - ISTR different ones give me different results from the DWBUA self-test, so I'm willing to entertain one or more are broken. What would help is being able to test and repair a UET in some other box, but that's a poorly documented module. Here's a thread from a couple of years back with some detail... http://www.classiccmp.org/pipermail/cctech/2009-April/064428.html -ethan

On Mon, Nov 7, 2011 at 11:16 AM, Gene Buckle <geneb at deltasoft.com> wrote: I have so many Unibus boards that it would be nice to get it working, especially since I can drop in multiple COMBOARD I and COMBOARD II boards along with one or two COMBOARD BI boards (the COMBOARD I only had 32K of SRAM, so it's limited to a HASP or 3780 emulator image; the COMBOARD II has 128K of DRAM but nobody ever ported the older images to it so it only runs SNA as a PU.2; the COMBOARD BI has 256K to 2MB (via socketed 44C256s) but customer demand for SNA on our products was already waning by the early 1990s, so I only ever got HASP and 3780 working on it. Interesting. I have either a DEBNI or a DEBNT and a cab kit. I haven't spent much time trying to get it working (since I'm running VMS 5.x and didn't have TCP/IP for it and I didn't have DECnet running on Ethernet anywhere else - only DECnet over DDCMP/sync serial). We ran Ultrix on 11/750s and 11/730s in the same room, but we only ever ran VMS on this box. It was purchased to finish our COMBOARD BI development, and after that, we mostly kept it powered off since all of our storage and everything else was on our 11/750 (half on SDI disk and half on SMD disks w/SI9900 controller), and by that time, we had less than 10 users, so the 11/750 was our main machine from our peak of 70-some employees down to the last days of the company. I still liked the 8300, but if it weren't for specifically needing a BI machine for product development, we never would have purchased it. Same goes for the 8530 (that I had to leave behind because I had no room left). That one was bought to fix a bug with Nautilus-bus machines where using our product could take down the entire machine (turned out to be bad to issue a bus STOP packet on an NI machine - was safe on BI and XMI machines, so the fix was literally commenting out one line of code in the driver). Those were fun days. -ethan