On 30-Nov-2016, Daniel Kahn Gillmor wrote: > Thanks for resolving #835598 by switching to a python wrapper of > gpgme. Alas, the ecosystem for using gpg from python is a cluttered > one and it's easy to settle on a problematic choice.
Thanks for wading into the clutter :-) > The "gpg" python module is maintained by the upstream maintainers of > GPGME and is now released with each new version of GPGME. Is this the library installed as ‘pyme’? (A terrible name, IMO, since it needlessly tells the Python programmer that they're using a Python library, and omits the most important part: that this is a library for GnuPG.) If the library is renamed ‘gpg’ that must be quite recent. Where can I read more about that? > It makes more sense to rely in a consolidated way on the active > upstream maintainers where possible. I agree, though I took the ‘pyme’ name as a pretty poor sign of fit with the Python ecosystem. When I looked at that library the interface was terribly low-level – a thin wrapper around the C-style library – and the ‘gpgme’ library was a much more Pythonic approach. If that has changed, I was unaware of it last month. > The attached cleanup/migration patch is mostly cleanup of the very > extensive test suite to more closely match how python-gpg maps to the > GPGME interface. Thank you for working with the code base and making sure to update the tests. The ‘python-gpg’ package doesn't seem to be in Debian Stretch yet, so I'll need to wait before assessing this. How new is that package? -- \ “From the moment I picked your book up until I laid it down I | `\ was convulsed with laughter. Someday I intend reading it.” | _o__) —Groucho Marx | Ben Finney <bign...@debian.org>
signature.asc
Description: PGP signature
