Definitely not ready for prime-time, this project is under heavy development and may change or stop working without notice. Not much is working at the moment anyway, but you are welcome to poke around and try things out if you are brave and have lots of time.
After a four year hiatus I am starting work on this again. I'm no longer sure
what my original goals are, but I recently got my PiDP 8I running again on a
Modle A Plus. I'm looking for a system that lets me re-live my youth working
with a simple PDP 8e teleprinter and paper tape. The simd software
available is great, but overkill for me.
Most of the prerequisites are included as submodules of this repository,
but you will need pthreads installed for Google Test and the multi-threaded
parts of the program.
Since the git repository contains the source code for the parser, you don't need to build it. But if you want to experiment with the assembler syntax you will need the antlr 4.7 complete JAR to compile Asm.g4 into the parser source files. Get if from here:
curl -O https://www.antlr.org/download/antlr-4.7-complete.jar
and move it to a standard location. I used /usr/local/lib but you can use
a different location and update the CMakeLists.txt file.
You will need a modern build environment. First you will need a C++
compiler conformant with C++17. I recommend gcc-8.1.0. Follow the
instructions found at Raspberry Pi - Install GCC 8 and compile C++17 programs
Then you will need an up-to-date version of cmake. I suggest you follow
the instructions at Installing CMake and get version 3.12.0
I'm not a fan of continuous integration (CI) per-se. But there is one feature of CI that I believe is essential for any project that may be used by more than one person; that is building from scratch -- and testing -- on a previously unconfigured system. This helps prevent problems that arise when the prerequisites are not fully documented. For this project I am using Travis CI. Travis CI integrates nicely with GitHub.
I am using a test driven design methodology on this project. This methodology suffers without a clear understanding of the code coverage those tests are achieving. For code coverage analysis I am using Codecov.io which also integrates nicely with GitHub and Travis CI.
See the change log.