Boost logo

Boost Users :

From: Noah Roberts (roberts.noah_at_[hidden])
Date: 2007-05-22 19:00:20


Noah Roberts wrote:
> I believe I have managed to use breadth first on a reverse graph to get
> a list of vertices I need. However, I can't seem to create a subgraph.
> I'm apparently not using an edge mutable graph? Anyway, it pukes up a
> static assert when I try to create one.

I can't believe how difficult using this library has turned out to be.
I'm very near just giving up on it entirely. After hours and hours of
excruciating pain and suffering I have managed to get everything working
that I need except for one thing. Now that I have changed to using
subgraphs, as required for being able to retrieve a subgraph at all, I
have run into a brick wall.

I can't write graphviz output anymore. The function that now gets
called when I attempt to requires a graph_name property that just plain
isn't there. I can find how to retrieve it by looking in the code that
thinks it should be there...but there is nothing anywhere that helps me
figure out how to set the damn thing; not in the docs and not anywhere
in the code that I can see. So I'm at an impasse unless someone can
tell me exactly where in the docs I'm supposed to be able to find this
information.

Not only that. I know for a fact this thing has a memory leak. The
subgraph setup creates a subgraph somewhere inside the parent that I
can't get to. I can create subgraph after subgraph but there's no way
to destroy them! Looking at the code that creates them it is quite
obvious that they are created on the heap and my repeated creation of
them is going to pose a problem rather quickly. Quite frankly, I can't
understand how this API exists without a method to kill subgraphs. This
is a major flaw.

Boost is great, but I tell you....some of these libraries could REALLY
use some work on the documentation side. I'm going bald trying to
figure this damn thing out.


Boost-users list run by williamkempf at hotmail.com, kalb at libertysoft.com, bjorn.karlsson at readsoft.com, gregod at cs.rpi.edu, wekempf at cox.net