another project

[Skaperen]

i don't have a clue if this is a small project or a large project.  i suspect it is a massively huge monstrosity.

i want to create a stack oriented, interpreted, language that has data types, classes, and other things just like python and make use of the python library.

[micseydel]

What's the motivation for this project?

[Skaperen]

the project is an old one for me ... about 30 years old.

i started working on it and came up with a data model very much like Python.  i encountered Pike around 1996 while still programming in LPC for MUD games.  i was in the midst of my transition from mainframes to Unix (later Linux) and from assembly to C.  Pike had a similar data model as the unnamed language ideas i had come up with.  i wanted to mostly create the data model and create two types of simple languages for it, one a classic procedure oriented language, and the other, a stack oriented language, and that by having a common data model (implementation) they could work together.  Pike was half of this ... i just needed to fine tune some details.  then i would add on a stack-oriented language on the same data model.

then Python came along.  it has a similar data model and a new syntax that i learned over time to be one with some nice advantages.  i decided that this data model was a little bit better than what i had come up with.  it also has one really great advantage (that Pike also has) ... it has an existing implementation.  so i dropped the idea of using the Pike data model and picked up using the Python data model, instead.

so now the overall big project is one data model and three languages, or languages syntaxes (or dialects), to work with that data model.  the stack-oriented language is now the one to implement.

the motivation is doing that

[Larz60+]

Just wanted to throw this in as I love stack oriented languages.

I used forth as a language to control a scanning monochrometer back in the 80's. It was totally stack oriented.
It is still used in some microcontroller applications, but pretty much died out.

The reason, I believe, was that it had a create ... does> command which had a compile time action and a run time action.
Essentially making the compiler totally expandable. This, although extremely useful, had the effect of making every forth application
a separate language, and making core language updates almost impossible.

---

I Was surprised that the forth interest group stil has a website here