Re: Re: Re: [Findbugs-core] toward ASM 4.0?

On 08/21/2011 08:29 PM, Brian Goetz wrote:
>> as I said previously, the code ensuring backward comptability will
>> only add a cost for old code. Up to date code will run as fast as
>> with the current API (or even faster since invokevirtual may be
>> faster than invokeinterface).
>
> You might be right, but I don't think you don't have remotely enough 
> data to back up this performance assertion.  Taking what is now a 
> simple code path which might inline easily, and turning it into a 
> complicated code path with lots of conditional logic and where some of 
> the "dead" conditional branches point off into native code (e.g., 
> reflection), may very well change the inlining and other optimization 
> characteristics.
>
> Again, I think this is an awful lot to consider between an RC release 
> and a final release.  RC releases are supposed to mean "unless we find 
> serious problems with new functionality, it's ready to go."
>
>

I agree with Brian, we will have plenty of time to think/test/polish a 
new API *after*
the release of ASM 4.

Rémi

-- 
You receive this message as a subscriber of the [email protected] mailing list.
To unsubscribe: mailto:[email protected]
For general help: mailto:[email protected]?subject=help
OW2 mailing lists service home page: http://www.ow2.org/wws