Maximilian Knespel

Thanks for your suggestion. It seems that libarchive has support for WARC since [2014](https://github.com/libarchive/libarchive/pull/81), but when I tried to mount it with archivemount or fuse-archive, the mount point was empty....

While working on #109 / #130, I have a state that can mount WARC files with libarchive. Without doing any special treatment, the file hierarchy for [hello-world.warc](https://github.com/iipc/warc-specifications/blob/master/primers/web-archive-formats/hello-world.warc) provided by libarchive...

You are right. After writing the issue I noticed that the RFC explicitly restates the ABNF rules that "T" is case insensitive. So, it might be an issue when you...

I'm doing `__shfl_down` reduction using `warpSize`. I guess it's very CUDA specific anyway ``` /* reduce per warp (warpSize == 32 assumed) */ int constexpr cWarpSize = 32; assert( cWarpSize...

I wasted time with this problem when trying to create a Conda patch, although I'm using the absolutely uniquely named and searchable "build" Python module. I tried to open an...

> You might wish to look into migrating to `pyproject.toml`-based configuration, which wasn't available at the time when this issue was opened and which setuptools won't rewrite. Thanks for the...

Format Outline: ``` Index Format Outline: - Magic Bytes: Don't have to be super short, it's negligible compared to a window anyway. - Format Version - Flags - Offset Format:...

Kinda fixed with gztool index format support. There still might be yet another index format that works cross-compression and fixes some small issues with gztool indexes.

> A generic solution should add a third blockfinder that looks for gzip stream header magic bytes, which should be sufficiently fast. It might be even easier to simply skip...

Simply allowing the final block bit to be set, almost halves the block finder performance! Because of this, looking for gzip headers might be faster then even if it requires...