Re: [cdt-dev] Scanner configuration correctness for CDT 3.0 feature spec

["Alain Magloire" <alain@xxxxxxx>]

> 
> Ed,
> You are right in your analysis. 
> There were three factors (at least that I know of) that were slowing down 
> the processing of output and error streams:
> 1. Extra layer of pipes through use of PipedOutputStream and 
> PipedInputStream objects in CommandLauncher.java. This has already been 
> addressed.
> 2. Error parsers that parse the build output for compilation and other 
> errors.
> 3. SC console parser that parses for SC affecting compiler options.
> 
> Certainly, we will be looking into decoupling SC discovery related 
> processing from build as much as possible, either through background 
> processing or some other way.
> 

Vlad, take a look at the new ErrorParser framework document
in cdt-core-home(Unfortunately because of lack of time, it is
not part of 2.1) make sure you push any requirements you have.