BLADE 0.2Alpha: Vector operations with mixins, expression templates,

[Pragma]

Dan wrote:
> Pragma Wrote:
>> In short: no.  But it's really an unintentionally loaded question: there are some rather severe limitations as to what 
>> kind of data structures you can create at compile time.  You're basically limited to tuples and strings, each of which 
>> have drawbacks of their own.
>>
>> You can create a tree using tuples, but then you run into problems with over-running the limitations on identifier 
>> length built into the compiler.  Strings are no where near as flexible out of the box, but pose no storage limitations, 
>> which gives a slight edge to string-based representations of data.  With some clever coding, strings can be made to 
>> store just about any structure imaginable (even if it does make for some ugly code).
> 
> Well said.  Once the AST is reflected at compile time, I'm sure the code can be made vastly more refined; and perhaps additional code generators can be developed and *hopefully* integrated into D.

Thanks.  But evaluating the AST will only take things so far.  There is a huge gulf of functionality that stands between 
where we are now (at compile time) and what D would be like as a full-on scripting language (not that I'm advocating 
that it be pushed that far).  Much of what sits in the middle (namely structs, multi-dimensional arrays and maps) can go 
a long way towards *analyzing* that tree for code generation and more.

> 
> Particular generators that spark my interest tend to have to do with x87, SSE extensions, and x86-64.  Most compilers to date don't properly use these functionalities.  Having them integrated into D Agner Fog-optimally would hugely replace alot of C++ gaming, video, rendering and graphics engine code.
> 

Anyway, I'm grinning too.  I can't wait to see what comes up next.  So far, converting Enki to a compile-time generator 
is proving to be a huge challenge, but workable.

-- 
- EricAnderton at yahoo