2012/4/9 Milutin Jovanović <[email protected]>: > > Hi Jakob, > > Let me digress a little. I saw Notch's new baby, and it does seem very > interesting. I thought of doing exactly what you did as well. But then I > realised one thing. DCPU is a very limited, both in memory and instructions. > Also Notch started talking about instrustion cost. This is all starting to > sound to me that heavy optimization will be the name of the game. Do you > feel that you can introduce Instruction cost based optimization into tcc? > This is the main reason I decided not to go this route.
TCC is very simple and pretty much just compiles C directly without doing much. And there the problem of DCPU not really being made with C mind. Tho there are options for improving the output from tcc. > > A big plus on the other hand is that this would allow me to get to know > innards of tcc much better, which I think would be big plus. > > Your thought, if any? Any help would be greatly appreciated. I think TCC is still a viable option as it has a lot of inbuilt helper tools, like it can emit and link objects. And is full SDK in 150k of binary. Cheers, Jakob. _______________________________________________ Tinycc-devel mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/tinycc-devel
