- Great collection! Interestingly, the only explicit MacApp book is for C++.
- This is incredible. How does it work?
EDIT: ok, here is some description: https://lisagui.com/info.html Apparently it's a re-implementation in JS, not using the original Lisa source code. Impressive in any case.
- Interesting article; I remember seeing DR DOS in the eighties and that everyone told me that it was superior to MS DOS, but I never cared enough, and it's great that now we can learn the details of all the ploys at the time.
- > seems to be an empty ZIP file
Same here. But the converted files are ok.
> some stuff seems to be unimplemented
As far as I understand, the provided files don't represent the latest compiler version. So maybe a later version had a better optimizer. But it's good to have the sources anyway, so I can use it to clarify language issues for my forthcoming Lisa Pascal compiler/transpiler.
- Cool project. Is the discussed p-system still 16 bit, or does it support a larger address space?
- You're limited to 64k on the AppleII, but one of the extensions to standard Pascal that Apple Pascal supports is "segments", where you can break a larger program into multiple code segments, which can replace each other in memory.
- Turbo Pascal had a similar method where code segments could be swapped in and out via .OVL files that were created at compile/link time.
https://secondboyet.com/articles/publishedarticles/theslithy...
- Apple Pascal supported > 64k of RAM on suitable machines, so it probably already did?
- Apple Pascal on the Apple II was restricted to 64 K memory; this limitation was due to the 6502's 16-bit address bus. The Apple III with it's 6502A could address 256 K, but only by bank switching; pointers remained 16-bit. The later Pascal versions (after Apple III) did no longer use P-code, but compiled to native code.
- There was a 128k version of Apple Pascal available on the Apple II. My question, applicable to both this and the Apple III version, is: Was the bank switching managed by the developer or by the Pascal runtime? If it's managed by the runtime, the addresses must have been encodable with >16 bits.
- All versions of Apple Pascal use 16-bit pointers, so data was always limited to a 64 KB address space. On the 128K Apple II, p-code was located in aux memory, and data was located in main memory.
Apple III Pascal had similar limitations, with separate 64K address spaces for p-code and data. On machines with more than 128 KB of RAM, there were assembly routines available for allocating additional memory and swapping data memory.
- The beginning of MIDI workstations; remembering Lengeling's and Steinberg's first programs; these were great times for musicians.
- Delphi was a significant achievement, and it's still relevant.
- It's an interesting interview, actually recorded on 2022, even before CHM published the source code, and it's only about 30 minutes long.
- And here is a tool to browse and navigate it: https://github.com/rochus-keller/LisaPascal
- Amazing, also that it's implemented in C#.
Star was apparently implemented in Mesa. Does anyone know whether the source code of Star was published?
- More
The Lisaem Emulator indeed does this. Lisa is a pretty complex system with an asynchronous bus which requires cycle-accurate simulation to make the original software work. I spend some time to experiment with simpler (and faster) emulators like QEMU, but this would only work with significant changes to all drivers (the Lisa source code was published by the CHM). So for me personally, the re-implementation of the Lisa GUI as demonstrated is a very wellcome solution to appreciate the innovations Lisa brought us without all the hassles of a cycle accurate emulator.