I'm working, or was working, on it too.
I'm almost positive I really do have the environment, size, and attribute issues fixed. (I wrote a simple driver to change the console to random sizes, and dump text with random colors which so far seems to "just work" and returns the console to the original state--minus the actual text.)
I think I've patched a couple of other API bugs I noticed. (It is hard to be sure because I use a very thin US locale Windows XP environment.)
I've separated the interpreter logic, the various "mini-basic" expansions and expectations, the editor logic, and the actual platform logic into separate chunks. (A few separate classes with simple abstractions. It certainly isn't my best work, but it is by far and away better than the original.)
I've added the command "ALIAS" to the interpreter to support variable names. (Mechanically backed by a different approach to individual operations based on the simple fact that "simplify_expression"--replacing "EvalExpr" and various other bits--is always called before the line is otherwise parsed.)
I've actually setup a few hooks so that it actually supports extensions without knowing how absolutely everything works. (I wanted to keep it reasonably simple; this too is by far from my best work, but you could actually add overloads and extensions without, more or less, having to rewrite everything.)
I've also giving the source a general "spit and polish" dropping the crazy use of jumps, adding some targeted exceptions, and simplifying a few of the functions with extra support routines.
What I've not done is added *all* of Sebastiani and matsp work back into the mix--just bits and pieces.
Unfortunately, a friend of mine, a DBA monkey, is in a living nightmare of a recovery and I've lent him the time I had been spending on this.
That said, you can stop finding reasons to bump this thread; what time I've spent on it was a game--I'm sure the others have similar reasons. If we want to stop working on it, we will; bumping the thread will not change that. If we don't want to stop working on it, we aren't going to forget about it. So... relax. You could, instead of pointlessly bumping the thread, actually help by rewriting the documentation--which according to you is simply wrong.
Also, you two, I have something to discuss with you, but I've just run out of time... and in a truly bashworthy way. I'll try to get around to it later... with your permission of course.