Hi bigriverguy,
I realize that there are probably a lot of reasons why the project is constructed the way it is. It just popped into my head that there would be a lot less translation to do going from QB64 to Freebasic rather than to C.
I am not suggesting that it should be done, just that it was a curiosity question for me, so I thought I would ask it.
You're wrong on 1 point:
"
there are probably a lot of reasons why the project is constructed the way it is".
Without any intention to hurt anyone and especially Galleon, to my point of view, the sole and unique reason of this construction is that at the beginning, this project was made by only one person without any real plan where to go. The integration of the editor into the compiler (and vice versa) plus the lack of comment is the cruel evidence of the lack of an initial real design. Yes it works, but I've never ever seen such a spaghetti code and I really admire the few guys who maintain it.
You're right on 1 point:
Yes, it should be done.
Both in term of maintenance and evolution, it would be for the real benefit of both the compiler and the IDE to have two different and separated parts: one, the compiler (the language) and two, the IDE.
Further, this would allow to use other IDEs (i.e. CodeBlocks, Eclipse, etc) and/or to develop a graphical brand new one "a la visual studio" including the real time debugger vWATCH64 from Fellippe.
Now, on the compiler part, this would allow to have a very smaller executable that could even be integrated into the Gnu compiler collection while being usable from any terminal session such as GCC or G++.
So I know it's a big job to do (due to the way the code is now), but for the future and the safety of this language (and especially the compiler), I don't see any other route to drive.
This would also help and motivate other new devs to focus exclusively on the development of the language itself, even by adding a new "object" layer (as did PHP years ago when moving from version 4 to 5).
Just my two cents.