Yes - similar to Small C :-) I've written several compilers/interpreters over the years, so thought I'd give the PDP 8 a go. Its challenging because of the lack of stack operations - but there are workarounds.

On 2/25/2023 9:52 AM, silvercreekvalley--- via cctalk wrote: Suggestions on how to make it more accessible are welcome. (Kyle and I are the maintainers, with some help from Doug and others.) Basically, the idea is to take the great many existing (RK05) images, and convert them to work over a serial port to a server running in a generic POSIX environment. There are several parts to this: There's some documentation (SerialDisk/docs) There are system and non-system handlers for the new device (handler). There are a couple of choices for short bootstrap programs to get things rolling (bootloader). There are a few provided RK05 images (disks). There's the POSIX program to serve the disk images over the serial line (server). There are various conversion schemes to get your RK05 image converted to the new drivers. The one I use most uses the SIMH simulator to automate the ritual of uninstalling RK05 drivers and replacing them with SerialDisk driver. The Makefile there will actually convert a disk image, and you could just swipe the default one once it's been made. Once the disk image(s) are converted to the new drivers, they boot OS/8 in a simple way using one of the short bootstraps. The other things you need (besides a PDP-8) are a PC/RasPi/whatever to run the server on, and as fast a serial port as you can manage. There are also example config files for running the whole thing under SIMH on your POSIX box. If you're more concerned about getting software developed than about the retro experience, that's also a great way to speed things along. TBH, when I'm in a hurry I often just use SIMH with the cross assembler, and skip OS/8 altogether until I'm ready to verify that things work on real hardware. You'll need more interaction with OS/8 to open files and such, so that might not work for you. It is a great relief to use a proper editor, rather than the one in OS/8. You might also want to hunt down the emacs-like editor that was done for OS/8 a couple of years ago, if you're going to be working native a lot. Vince

I always wanted to build a tss8 system! Ed#  hiding out at smecc Sent from AOL on Android On Sat, Feb 25, 2023 at 11:34 AM, Vincent Slyngstad via cctalk<cctalk(a)classiccmp.org> wrote: On 2/25/2023 9:52 AM, silvercreekvalley--- via cctalk wrote: Suggestions on how to make it more accessible are welcome.  (Kyle and I are the maintainers, with some help from Doug and others.) Basically, the idea is to take the great many existing (RK05) images, and convert them to work over a serial port to a server running in a generic POSIX environment.  There are several parts to this: There's some documentation (SerialDisk/docs) There are system and non-system handlers for the new device (handler). There are a couple of choices for short bootstrap programs to get things rolling (bootloader). There are a few provided RK05 images (disks). There's the POSIX program to serve the disk images over the serial line (server). There are various conversion schemes to get your RK05 image converted to the new drivers. The one I use most uses the SIMH simulator to automate the ritual of uninstalling RK05 drivers and replacing them with SerialDisk driver. The Makefile there will actually convert a disk image, and you could just swipe the default one once it's been made. Once the disk image(s) are converted to the new drivers, they boot OS/8 in a simple way using one of the short bootstraps. The other things you need (besides a PDP-8) are a PC/RasPi/whatever to run the server on, and as fast a serial port as you can manage. There are also example config files for running the whole thing under SIMH on your POSIX box.  If you're more concerned about getting software developed than about the retro experience, that's also a great way to speed things along. TBH, when I'm in a hurry I often just use SIMH with the cross assembler, and skip OS/8 altogether until I'm ready to verify that things work on real hardware.  You'll need more interaction with OS/8 to open files and such, so that might not work for you. It is a great relief to use a proper editor, rather than the one in OS/8.  You might also want to hunt down the emacs-like editor that was done for OS/8 a couple of years ago, if you're going to be working native a lot.     Vince

On 2/25/2023 10:46 PM, Tom Hunter via cctalk wrote: Ah, I seem to have misspoken. Inspecting the Makefile in .../installer, it still runs the executable "handler_installer", which is the older tool that more or less just hammers a new system handler and bootstrap into place. The script for the new installer is "sdsk". This sets up a path for the tools, copies the newest driver binaries into the exploded disk image (while padding them to an even block length), then imploding the exploded directory and making a new bootable RK05 image. Then "runbuild" is invokes, which in turn invokes SIMH and sends a series of commands to BUILD.SV to remove various other drivers and install the "SDSKSY" and "SDSKNS" drivers copied above. Finally, the "BOOT" command is given to write the new second stage (and third stage) bootstrap. The self-modified BOOT.SV is also saved as BUILT.SV. Sorry for the confusion. If you don't want to use diagpack2, SDSK.md is helpful, but basically you need to edit the line: image=diagpack2 in the sdsk script, and to have previously copied in your starting image, and run ../tools/os8xplode on it. SDSK.md also contains information about the need for "perl", "socat", and "expect" and to get them (on Linux, anyway). Oh, and the SIMH initialization file (pdp8.ini) is also firing up the server program, so you need to tweak that to make your "foo.new" available to the SIMH serial emulation. That script used to also require a functioning xterm, but it doesn't any more. If you have xterm and want to see the server log in real time, there are lines there to invoke ../server/server the other way, and you can un-comment that and comment out the bash line. Once you've done all the above, it suffices to just type "pdp8" in that directory and your new system should boot in SIMH. Hope that helps, Vince

Here's a link to a distribution for the E8 emacs-like editor I mentioned: https://tangentsoft.com/e8/wiki?name=Home Also, tangentsoft has a bunch of information about software development related to the PiDP effort, including their version of the C compiler. https://tangentsoft.com/pidp8i/wiki?name=Home https://tangentsoft.com/pidp8i/doc/442d9cff50/src/cc8/README.md The PiDP project has generated quite a bit of attention to the PDP-8 over the last several years. Vince

Attached are two bash scripts (tested on Raspberry Pi OS) for serial disk. run each script with sudo ./<scriptname>. If an existing os8diskserver directory is present it will be renamed and a new copy installed. getos8diskserver -- installs all of the necessary files to build the os8diskserver and the os8diskserver source itself. makeos8diskserver -- builds os8diskserver, installs a local copy of simh and builds a serial disk bootable image as well as several other rk05 images and creates a script to start os8serial disk with the default RK05 images. Feel free to fold, spindle and mutilate this. Please let me know if you find any bugs, have any suggestions or make any generic modifications. I'm hoping to get this into the os8diskserver github as soon as I get a chance to create a pull request Note: This is also being cross posted to the VCF DEC forum. Please post this to any other forums or mailing lists you feel are appropriate. Mike

On 25.02.23 18:52, silvercreekvalley--- via cctalk wrote: This is probably not the right place to link the youtube video of the guy that destroyed a FPP8/A set to recover the gold on the PCB-fingers... I do have a rather complete 8/A, with RL01 ad FPP8/A. Alas no sockets inside so the PROMs cannot easily be read out. Jos

On 2/26/2023 9:47 AM, Vincent Slyngstad via cctalk wrote: The source code for the FPP8-A is available - I retyped it while trying to debug the SIMH FPP code. Getting that source into a binary is a exercise for the reader. :) Apparently the compiler for this ran on TOPS-10. The maintenance docs for the FPP8-A give enough information to reverse engineer the programmable bits.     -Rick