Re[2]: [Gnash-commit] gnash backend/render_handler.h backend/render_h...

[Udo Giacomozzi]

Hello strk,

Thursday, March 1, 2007, 11:48:23 AM, you wrote:
s> My own style so far has been to use CamelCase for new things.
s> I'm sure many will disagree...

You choose CamelCase for some good reason?

I'm happy with any of them, but think it's a bad idea to mix both
styles.


>> s> How about renaming this to shouldMerge() or similar ?
>> s> And, does the 'inline' part make any sense in a templated class ?
>> Why not? How are these two related?
s> I'd think templated functions would always be inlined, but I might be wrong.

AFAIK templates just duplicate the code for whatever type you need.
If templates would be always inlined then this would produce a huge
AGG backend. "inline" just means that the function is not built as a
real function (no stack and such), AFAIK

>> I'm avoiding abs() since it depends on variable type.

s> std::abs() should be an overloaded one, did you try ?

My docs say abs() is for int, fabs() for float etc. But I could have
looked the wrong place.

>> s> I think we want contains(const Range2d<T>& ) too..
>> 
>> You are free to change this.

s> I'll do when approaching existing testcases.


Patience...


Udo