Standard graph API?

[Jeremy Bowers]

On Mon, 23 Aug 2004 20:41:53 +0000, Magnus Lie Hetland wrote:

> In article <pan.2004.08.23.16.22.34.771887 at jerf.org>, Jeremy Bowers wrote:
> [snip]
> 
> As I tried to state in the original post (I probably wasn't clear enough)
> I'm not talking about a standard *implementation*, just a standard *API*,
> like the DB-API. This could easily cover all kinds of strange beasts such
> as directed or undirected, weighted or unweighted (etc.) graphs;
> multigraphs, chain graphs, hypergraphs, who knows.

Point conceded about API and not a library, but I'm not sure that changes
my point much. Your API is going to assume something about how edges are
represented (which will conflict with somebody), *or* it will be so vague
as to not have any particular advantage over the nothing we have now. And
so on for most of the other dimensions.