Re: Re: Prothon's parser is too naive

Mark Hahn wrote:

> "Lenard Lindstrom" <[email protected]> wrote
> 
> 
>>Remember that Prothon has no equivalent of Python's compiled files.
> 
> Modules
> 
>>are compiled every time a program is run.
> 
> 
> I will offer pre-compiled programs some day, but they will be quite
> different than Python.  All the objects will be pickled for example.

Python modules are basically marshalled code objects. Marshal is just an 
older, simpler, probably more efficient form of pickle.

>>Unless that changes one must
>>consider compiler performance. And given the dynamic nature of prototypes
>>optimizations that work in a typed language will not apply to Prothon.
> 
> Prothon will never do static optimizations in the compiler.  It is a dynamic
> language through and through.

Maybe you mean static *type-based* optimizations.

If not: are you saying that if I look at a series of Prothon's byte 
codes I will be unable to find any inefficiencies? i.e. byte code 
sequences that could be optimized to other byte code sequences without 
changing the semantics at all?

If there are no such inefficiencies then I'm amazed.

If there are such inefficiencies then why would you argue against 
optimizing them away at compile time? (anyhow, I don't see why you would 
decide in advance never, ever, to take advantage of type optimizations 
either. If the code says "3+5" why shouldn't the compiler replace it 
with "8")

  Paul Prescod