Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!cs.utexas.edu!uunet!zephyr.ens.tek.com!gvgpsa!gold!grege From: grege@gold.GVG.TEK.COM (Gregory Ebert) Newsgroups: comp.sys.nsc.32k Subject: Re: Call for discussion: alt.hack.system Message-ID: <872@gold.GVG.TEK.COM> Date: 22 Mar 90 22:02:09 GMT References: <1341@carroll1.cc.edu> <865@gold.GVG.TEK.COM> <1349@carroll1.cc.edu> <604@unicorn.WWU.EDU> Organization: Grass Valley Group, Grass Valley, CA Lines: 79 In article <604@unicorn.WWU.EDU> bobh@unicorn.WWU.EDU (Bob Hayes) writes: >OK, OK Now there's only 3 solder splatters out here! Ok. I'll take some initiative. The first step is to define our systems. Since most of us hackers are on a limited budget, we should concentrate on using as much as we can of what's already out there or in our junk boxes. The hardware aspects should be kept as bare-bones as possible, yet leave provisions for future expansion. Stuff like DMA, 32 serial ports, SCSI, etc is nice, but for the purposes of just getting something running, all we need is a serial port and a monitor program. A few years ago, I bought a 32032 designer's kit from Jameco, which had a 6Mhz set of the CPU, MMU, FPU, TCU, and ICU. There were also some ROMS which have an assembler. I gave up on using the ROMs because I had no source code, and there was no detailed documentation on what kind of hardware was expected. Thus began my endeavor : to build an expandable 32032 system. I started by redefining the S-100 bus so that I could use my 8-bit cards with the new system. That simplified matters immensely, because all I needed then was to wire-wrap a CPU and RAM/ROM card. Why the memory card ?? The 32032 needs a 32 bit bus. I chose a very simple arrangement with 12 28-pin sockets which can use 8Kx8 or 32Kx8 static RAMs or EPROMs, mapped from address 0. The boot ROM will need to be copied to RAM, and deselected in future revisions to allow software to change the interrupt vectors. The CPU card follows the schematic in the data book, except I added the ICU and ran the INTx lines to the VIx lines on the S-100 bus. Thats as far as I've gotten. We need to get some software going. I propose that we use the stuff in beowulf, and whatever is available from the Free Software Foundation. Although I am not affiliated with them, I think that we should contribute to their cause if we decide to use their software. For the meantime, we will need to get a VERY simple monitor up and running, before anything else can be done. ACTION ITEMS ------------ I will undertake the following tasks: 1. See what is available from beowulf. Some people in the past have had trouble getting stuff back because the netland mailers sometimes screw-up the return path. Don't blame beowulf! I will report waht happens. If I have trouble, will ask someone to give it a try, and that person can then be the individual through whom we coordinate software requests. I DON'T want everyone pounding on the server there, and tying it up, because we could lose the privelege. 2. Dust-off my notes/schematics for the 32032. Get stuff online in particular the S-100 bus retrofit for others to see. 3. Find and adopt a P-D assembler for the 32032. I would greatly appreciate it if someone could help with this. It would be ideal if we had 'C' source program which would create ASCII-hex files which I could then dump onto our PROM burner. FUTURE TOPICS ------------- 1. The monitor. I want to write it in assembly-source code for the above-described assembler. It MUST have drivers which can be easily modified for ANY hardware configuration. This entire project is gigantic waste of time if we cant get some sort of agreement upon hardware-independence. I don't want a fiasco like the IBM PC !! 2. Remember: This project can only survive if YOU contribute. The 32032 is in my opinion the best thought-out processor ever made, and we will all have a great time working with it.