Alan Hudson wrote:
> Paul Sandoz wrote:
>
>> Hi,
>>
>> With all these features added the performance has dropped but i have
>> tried to counter-act this with additional performance enhancements
>> (reducing method calls and field access). Performance results show
>> that i have made up ground but it is probably about 5%-15% slower.
>>
> I'll rebench mark Xj3D with the latest. Just got back from travel, so
> it will be wed before I have new numbers.
>
I managed to claw back a little bit more by implementing UTF-8 decoding
of content more efficiently. So, on average we are similar to the
results previously presented, sometimes a little better sometimes a
little worse.
I suspect that performance fix will not make much difference to binary
X3D documents since attribute data will be binary encoded.
So probably the biggest impact will be duplicate attribute checking. I
can add a feature to switch this off if it is a problem for you. The
implementation is quite efficient and should for the most part performed
in O(1) for each attribute.
Paul.
--
| ? + ? = To question
----------------\
Paul Sandoz
x38109
+33-4-76188109