Re: [boost] Boost.Graph Metagraph project

Subject: Re: [boost] Boost.Graph Metagraph project
From: Joel Falcou (joel.falcou_at_[hidden])
Date: 2011-01-24 02:17:51

On 16/01/11 05:07, Gordon Woodhull wrote:
> On Jan 15, 2011, at 5:01 PM, Gordon Woodhull wrote:
>
>> I'm pleased to announce that the proposed MPL.Graph will be included in release 1.46 as a sublibrary of MSM!  msm::mpl_graph
>
> Forgot to mention that mpl_graph is being used by MSM to find regions / connected components in state machines, and to find unreachable states.
>
> I think that work may lead me to implement a forest for efficient disjoint set tracking.  I also want to abstract out the connected components algo - Christophe is using DFS directly for now.

Jumping again here. What are th eimpact of said algorithm on compile 
time ? Is there a way we can handle DAG in a meaningful way ? My
experiences with rpoto lead me to think that we need some CT-DAG to get 
more out of our proverbial proto AST.