Resurrecting a PDP-12 in Duluth

It is with a deep sense of loss that we report that Warren Stearns, the PDP wizard who came from Rapid City, South Dakota to help us restore our PDP-12, died on Sunday, August 6th 2017 following a massive heart attack. We waited to write this post until his obituary was posted ; visitation is tonight and the funeral is tomorrow, Thursday, August 17th, in Hartford, SD (details at the above link). Warren was well known in the PDP-8 community and spent a considerable amount of time traveling to various vintage computing gatherings and helping others restore their vintage machines. We first got in touch with Warren through the Rhode Island Computer Museum's (RICM) Michael Thompson, who had (with others) worked alongside Warren to restore their own PDP-12 starting in 2015. Michael told me that Warren might be interested in coming to help us. Warren called me a day or two later and said that he could be there in two days. Warren was so generous to us with his time -- he spent e

Photo courtesy Bob King and the Duluth News Tribune Our demo on Friday was a big success! It's often hard to get people to show up on campus in the summer, but thanks to some university announcements and an events on  Perfect Duluth Day's  online  event calendar  and Facebook , Heller Hall 306 was standing room only! My favorite part was when we asked for volunteers to come play Spacewar! and two kids shot up their hands. Their dad (a former UMD CS grad) drove up from the Twin Cities just to see the demo. The Duluth News Tribune wrote a nice article about the demo and Fox 21 (KQDS) did a little video piece on the demo. We appreciate everyone's interest and support and we definitely plan on doing more demos in the coming school year -- so don't be too disappointed if you missed this one. (We just need to find a bigger room on campus that also satisfies the constraint of being on a pathway navigable by a PDP-12!)

An article was recently written about the restoration process and was featured on the UMD home page. Here's the link: https://news.d.umn.edu/news-center/news/computer-restoration Tomorrow, Friday August 11th at 11am we will be giving a public demonstration of the computer in Heller Hall 306, 1114 Kirby Dr., Duluth, MN 55812 . More details are in the article . Feel free to contact us with questions!

Over the last week or so I have been busy with the other half of my UROP, the benchmarking. I will be benchmarking the PDP-12 and comparing its results to other machines. So far for test machines besides the PDP-12 I have a Commodore 64, a 286 PC, and a 486 PC. For a while I was trying to rewrite Dhrystone to run in Fortran IV. After spending a few days on it I realized that doing so would likely alter how Dhrystone runs. Dhrystone was written in Ada originally, and then ported to C and Pascal. All three of these languages support pointers and structs/records. Fortran IV does not. Not having pointers makes translating the code very difficult. There is Pascal for the PDP, but it requires 28k of core. We have 8k. Whetstone seems to be a better fit. It was originally written in ALGOL 60, and there is a Fortran version available (Fortran 90, but I could rewrite it to IV). There is an ALGOL compiler for the PDP-8 (DECUS 8-213 4k ALGOL) called ALGOL 8. It is a subset of ALGOL 60. I hav

This is a bit overdue of an update, but... We received a package in the mail from a collector. They donated a 4k core and a M216 Flip Chip. We were borrowing one of Warrens extra cores before. We needed the M216 for the TC12-F option which allows us to read DECtapes as well as LINCtapes!   The 4k core was mostly working. Warrens Tune program said there was a bad row and column. Using a multimeter I tested all the diodes on the core. All of them seemed fine. One was a little flakey so we checked them again. Pushing on the diode caused it to snap in half. This was one of our problem diodes. The other one was a bit harder to find. We narrowed it down to a section of 8 and then used a power supply with current limited to 300mA to test it a bit more. The diodes normally would draw just around 1V. One of them was taking about 1.5V. I pushed on it a little and it cracked. This was the other problem diode. Both diodes were replaced and then I put the core in to test it. The memory che