Zane Healy writes: I started getting serious about collecting vintage computers when I was about 20, so not much older. That was in 1993, and I was scrounging PDP-11/23s and VT100s and VT220s, so most of what I was finding was 10-15 years old at that point. The more things change, the more they stay the same! -Seth

First access was when I was 16 and used the remains of 3 Teletype 33's to make one working unit, the same neighbor who got me the machines help get and fix a modem, I had access to the the HP 2000C at the local college from home. (guess the year). First oldie computer was a then not oldie Burroughs B720, I worked at the plant where they were made and got one that was freight damaged. About 1,000 14+16 pin DIP 7400 TTL packages and ran at 2 Mhz. Next was a Data General Nova (no suffix the original) with a Teletype 33 ASR, And a high speed tape reader. Still wish I had it now. Love being able to turn it off and back one without loosing data other then what was in the registers. One day I got a call, "do you want some minicomputers ?". Stupid question. "Come down to Veterans Stadium (Philly) with a truck", the caller was a friend who was an electrician. The company that won the bid to replace the game board at the stadium had the remains of the old setup piled up ready to go out for scrap. I filled the truck with General Automation SPC-16's and I/O boards, cables, etc. enough to control over 100 individual 120V lights. The big regret .. 1975, get a call "do you want an IBM mainframe" .. but was moving 3,000 west thee weeks after the call. Was a whole computer room 360/45, disks, tapes, etc. everything except the 1403 Model 5, but the Model 2 was in the offer. If I had not accepted the job and was not moving I actually had free use of a warehouse. After moving, 1981 I paid $100 for a 4 Racks of DEC, Rack 1 PDP 11/45, Rack 2, RP03 controller, Rack 3 4 x RK05s, Rack 4 a TU?? two RP03's, VT100's, etc, etc. Filled 1/2 of the basement. I ran RSX11M on it. The last big box at home was a SUN M5000, but its now gone to a good home. The last offers I've turned down are a Sun/Fujitsu M9000 and a V890, My employer would not let me or anyone have the IBM Z/?? and it went out as E-scrap a couple months ago. In storage there are twp DEC Alpha DS20E's and a older DS20 that I'll do something with or find a home for. On Wed, Mar 30, 2016 at 8:16 AM, <ethan at 757.org> wrote:

Wonderful video. On Tue, Mar 29, 2016 at 5:42 PM, Evan Koblentz <cctalk at snarc.net> wrote:

Sent from my iPhone I got to see that DG just after he picked it up in Akron. He stopped by the Large Scale Systems Museum in New Kensington PA and we gave it a look over and also discussed the docs he got with it and what to tackle first. Now if only it had come with some of those cool blue/green DG terminals :) I still remover the day he sat down in front of the HP 2116B, he had never touched a blinking light front panel, he asked if Dave or I minded if he erased the program in memory, we said no, he powered it on and 2 min later was reading the programming docs for it and wrote a quick blinking lights program in like 15 min. Connor is a good guy!

OK, I have to admit, that video is impressive, he held the audience attention better than people twice his age, and the sheer audacity has to be commended. Sadly, my zOS days are long over, I don't even remember how to configure LU2 or LU6.2 anymore, and I got paid to do that. That said, I do not regret getting into "home" computers :-) Man, those machines are heavy and bulky. Anyone have the zOS machine that was in a PC case (or is my memory off on that)? Jim

On 3/29/2016 9:54 PM, Mike Ross wrote: I think that was it. I heard about it when I worked at Compuware in the early 90's. The presentation took me back to those days, not so long ago, when it was expensive to put TCP/IP on OS/390, so much so that Compuware developed "HCI", which was a TCP/IP stack for the platform (not sure if they sold it, bundled it with apps, or it was just Andy Falvey's pet project. I remember he borrowed my Comer books so he could implement TCP/IP... Jim

Compared to waterbeds or pianos, computers aren't all that heavy. Not even mainframes. DEC's old headquarters (the "Mill" in Maynard) had prominent signs everywhere saying the floor load limit was 100 pounds per square foot, or something like that (perhaps less). But they just planted their computers all over anyway, and I don't remember ever seeing floor reinforcements for that. The stated limit was probably quite conservative. After all, that building used to contain spinning and weaving machinery -- big hunks of cast iron. paul

On 03/30/2016 10:51 AM, William Donzelli wrote: Yes. Well, the problem with the 370's were peripherals and power. Although the lower level 370's had integrated channels, you still needed control units and drives. To do any real work, you'd need at least a DASD controller and drives and a com controller. Probably at least a tape controller and one drive to do backups and load software. That won't fit in the bedroom of a 2-bedroom house, which was where this thing was! Second was the 370s used a motor alternator set to convert 50/60 Hz power to 415 Hz power. To keep compatibility, they used the same size MG set for a range of machines. So, the 145 had a 17 KVA output alternator with a 20 Hp 208V 3-phase motor. We tried to build a phase converter to run the motor off the guy's 60 A 240 V service, but the reactive current was over 60 A and would blow the main breaker. The main logic supply for the /145 was 390 A at +1.3 and -3 V, so about 1.5 KW. We should have just got a bunch of big 5V switcher supplies and tuned them down to the appropriate voltage. But, that would have likely cost more than he paid for the machines, maybe even surplus. The 145 was not a real high performance machine. Although the memory was 64 bits+parity wide (for ECC) the main data path was only 16-bits wide, and the microcode word was rather vertical, so while 360/50 and 360/65 could execute simple instructions in about 4 microcode cycles, the /145 took something like 7+. A 32 bit add register to register was 1.4 us, add memory to register was 2.4 us. Floating point divide was 28 us (short) or 88 us (long). Jon

On Wed, Mar 30, 2016 at 03:54:40PM +1300, Mike Ross wrote: Dang, I suspected they were hobbled, but that's painful. mcl

On Wed, Mar 30, 2016 at 08:34:00PM +1300, Mike Ross wrote: I guess I didn't highlight the joke. Thirty-*one* bits? mcl

Long time I don't laugh so hard! :D 2016-03-30 7:48 GMT-03:00 Raymond Wiker <rwiker at gmail.com>:

There's an IRC channel? What IRC network is it on?