cpockrandt

@h-2 Done. Here are the bug and meta-bug issues: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=24639 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84299

The remaining warnings are the ones that I reported in the bug issue above (triggered by the same line). Since the issue is not closed yet, I guess we've to...

Couldn't get cmake working with clang.exe on our build machines. Therefore postponed until further notice.

Hi @ilya-kolpakov, 1. The FMIndex does not have such an interface and since we're currently working on SeqAn3 there is unfortunately no time left for new features in SeqAn2. But...

> How can one extract (initialize) a monodirectional iterator from a bidirectional iterator (index). If `it` is a bidirectional iterator, you can access the unidirectional iterators which are member variables,...

I forget to make an important remark: Using a unidirectional iterator that belongs to a bidirectional iterator, will break the bidirectional iterator (you will get false behavior since both unidirectional...

As long as you use TopDown-Iterators (without ParentLinks, which store the entire history of going down in the tree), you should be fine in terms of space.

@h-2 Is this issue really supposed to be assigned to milestone 2.2.0? I haven't heard from David in months and couldn't reach him. Maybe this should go to Backlog?

For 32bit and 64bit we're already using the fastest popcounts. I did some calculations & tests and came to the conclusion that If we could use block sizes > 64...

I just read into this recently published paper [1]. They give an implementation of a popcount on 256 bits using AVX2 intrinsics which they claim to be twice as fast...