Dan Schult

We already had a version of the code that generated the maximal s-metric graph for a given degree distribution according to the method in that paper. Based on [the "blame"...

[Here is a reference paper ](https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=152129) following up on the 2005 s-metric paper. It was written in 2008. It points out that the algorithm in the 2005 paper for finding...

I also like the `method` kwarg approach. @GuyAglionby do you have a sense for whether and why we should include multiple methods? For example, why not just implement the Wu...

Arg -- yes, I see the problem. Maybe we make two private helper functions: the current edges function and the current multiedges functoin. Then we make a new public version...

I think the tests are failing because there isn't any return value from `draw_networkx_edges`. It should return the drawn objects and probably now doesn't return anything -- or maybe there...

See #5466 and #5514. Perhaps we need a warning when `arrows is True` and G is undirected **and** a warning when connectionstyle is specified and G is undirected but arrows...

Would changing the name of `arrows` to `fancy_edges` alleviate this confusion? When `fancy_edges` is False, the arrowstyle and connectionstyle are ignored. Directed graphs default to `fancy_edges=True`, Undirected graphs default to...

I think it would be fine to work on this -- but you should realize that the final API decisions haven't been made. It might be helpful to see what...

Closed due to being fixed in #6098

I think the aspect ratio comes into play anytime you take ratios of x-values and y-values. Think of it as translating from x-units to y-units in each of these ratios....