Steffan Sølvsten

With regards to _complement_ edges: The semantics of _attribute edges_ for Zero-suppressed decision diagrams as given in Section 2.5 of [Zero-suppressed BDDs and their applications](https://link.springer.com/article/10.1007/s100090100038) by Shin-ichi Minato: > Minato[16]...

I've been taking a look at the `zdd` branch, and here are some comments of mine. You should take into account, that I primarily think ZDDs in terms of _families...

I still have to read the Master Thesis of Hajighasemi in detail, but a quick glance at it repeats the same idea of a more BDD-like view on most algorithms....

I just fell over this figure in Minato's paper on ZDDs . It shows the intended rule for the _attribute edge_ - I'll leave it here, assuming it could potentially...

Here is a list of not yet implemented features that were not mentioned in the comments of *sylvan/sylvan_zdd.h*. - [ ] C++ support - [ ] Boolean Polynomials as used...

It should also be noted, if *C++* support is added, you almost for free get multiple benchmarks ready to evaluate Sylvan's performance against CUDD. Specifically, the following three combinatorial problems...

The maximum variable of what? The maximum variable supported by Sylvan or the maximum variable of a single BDD (i.e. the deepest variable)?

Sylvan always picks for the table and cache size the largest power of two that fits into the given amount of memory (that also satisfies the given cache-to-nodes ratio). So,...