Little Smalltalk, version 4.0
Timothy A. Budd, [email protected]
Oregon State University
I first started developing Little Smalltalk back in 1984, and to tell the truth hadn't touched it for several years now. However, when I attended OOPSLA in Portland a few months ago, several things struck me
- Addison-Wesley was still selling my book ``A Little Smalltalk'' (despite it being many years old now and very very dated), and people were still buying it (don't ask my why).
- There was a heck of a lot of interest in Smalltalk on the vendor floor.
- People kept asking me about Little Smalltalk.
First and foremost, let me say at the beginning in as LOUD a voice as I can that Little Smalltalk is NOT a smalltalk-80 system. If you want a real Smalltalk system go buy one from ParcPlace, Digitalk, IBM, or whomever.
Little Smalltalk is a sort of fun experiment in building a minimal smalltalk system. The source code for the interpreter is less than 1800 lines of code. The image consists of less than 4000 objects. It runs in almost no memory. In short, it's small, it's reasonably fast, it's easy to understand, and easy to modify.
As you might expect, in order to get Smalltalk to run in something this small, I've had to make lots of changes. Here are the most notable:
- No graphics.
- The user interface is a more conventional real-eval-print loop (like most Lisp systems), not a browser. (But it is all written in Smalltalk itself, so you can modify it if you wish).
- Primitives and cascades are handled as they were in earlier Little Smalltalk systems, which is not how they are handed in Smalltalk-80.
- The standard class hierarchy differs considerably from Smalltalk-80.
- My basic collections are List, Tree, Dictionary, Array and String
- The only numbers in the standard image are large and small integers (adding fractions and floats might be an interesting exercise, however).
Anyway, if after this you still want to play with the system,
here are the details. You can ftp from
ftp.cs.orst.edu.
The file
you are reading is found in
/users/b/budd/little
on that system, and the
file names I'm going to mention in a moment are found in that directory.
Or, if you are using a www browser, you can set the option to copy sources to disk, and read the following files.
- distr.tar contains the entire distribution as a tar file. Alternatively, you can get it piece by piece.
- The interpreter is contained in three files: main.c (the main driver), interp.c (the bytecode interpreter), memory.c (the garbage collection system).
- The standard image is found in the binary file named LittleSmalltalkImage
- Various test cases are found in a shar file named Tests.
To run the test cases use the fileIn: command (see ``documentation'' below).
I've only compiled the system on a sys-v based Unix system -- but it uses nothing out of the ordinary so it may work on other systems. Let me know what problems you run into.
Documentation is, as usual, rather spare. In my spare time when I'm not hacking the system I am revising the Little Smalltalk book, and a few sample chapters are available in postscript form in book.ps.
Have fun. Maybe this will be a real system some day. But I didn't think I should hold up distribution until that point.
Little Smalltalk can be freely and widely redistributed if not for profit and as long as the markings indicating origination are retained.
The system is not supported (it's only a hobby for me), and is, of course, not warrentied as suitable for any use whatsoever.