kicad-developers team mailing list archive

[Dick Hollenbeck <dick@xxxxxxxxxxx>]

On 09/05/2013 09:38 AM, Lorenzo Marcantonio wrote:
> On Thu, Sep 05, 2013 at 07:12:40AM -0500, Dick Hollenbeck wrote:
>> That looks excellent.
> 
> Yep, quite.
> 
>> I was also thinking bit set.  The fact that std::bitset<> takes a size suggests it is
>> putting the bits in the instance block, not in a separate block of memory.
> 
> I think the issue is not the 'best' kind of bit vector (his
> implementation is more or less like the bitset, there are also
> std::vector<bool>, or even std::set<LAYER_NUM> which are not
> compile-time bound in size). 


Says you.   Says me:  having the the bits in the instance block matter.


I guess if what I have to say is going to be labelled as irrelevant right out the box, I'd
be wasting my time continuing.

I don't have it to waste.




Of course you can't have an 'unbounded'
> fixed size structure (there are tricks, like, keep the first 32 often
> used bits in a local word and allocate a block if other ones are needed,
> but this is an implementation optimization).
> 
> I think the question is if to keep the current fixed-in-code layer
> stackup or moving toward a more dynamic layer structure.
>