This implement the change discussed in #20 because it's very easy to do, and a net API generalization with no known harmful side-effect. But depending on how the discussion in #23 goes, this PR may be scrapped entirely in favor of a more ambitious API redesign that resolves the pointer alignment problem once and for all.
For convenience reasons, this PR is based on #22, #24 and #25. Reviewing these PRs first is recommended, but if you want to review this one in isolation, please consider doing so commit-by-commit in order to get a clean diff.
This implement the change discussed in #20 because it's very easy to do, and a net API generalization with no known harmful side-effect. But depending on how the discussion in #23 goes, this PR may be scrapped entirely in favor of a more ambitious API redesign that resolves the pointer alignment problem once and for all.
For convenience reasons, this PR is based on #22, #24 and #25. Reviewing these PRs first is recommended, but if you want to review this one in isolation, please consider doing so commit-by-commit in order to get a clean diff.
Fixes #20 .