https://git.schokokeks.org/derivepassphrase.git/tree/6e2d2728b76c22a0d2280c34a55dba45885084cb 2024-12-07T08:50:03+01:00 tag:gitlist.org,2012:commit/6e2d2728b76c22a0d2280c34a55dba45885084cb 2024-12-07T08:50:03+01:00 Marco Ricci software@the13thletter.info <pre>We expect to add new option categories that apply to multiple super- and subcommands, not just the "vault" subcommand. &lt;/pre&gt; tag:gitlist.org,2012:commit/10c7604bf033455d3e777bdeb028ee7d034a9bb7 2024-12-07T08:50:03+01:00 Marco Ricci software@the13thletter.info <pre>Our previous method of dispatching to subcommands was not tying the supercommand and the subcommand together properly; each command essentially ran in its own isolated context. This lead to unnecessary double-parsing of the command-line arguments, and to `click` becoming confused as to what the full command-line actually was; in particular, the help and usage messages gave a wrong command summary. Instead, we now use custom `click.Group` objects as the supercommands which default to the "vault" subcommand (with a deprecation warning) if no explicit subcommand could be matched, rather than throwing an error; this all happens internally using the normal mechanism for subcommand dispatch. Additionally, if the supercommand callback actually is invoked -- which only happens if the supercommand is given without any arguments -- we emit a deprecation warning and dispatch manually to the "vault" subcommand using the same code path as the `click` implementation would otherwise do. We assert that the old and new code paths work via new tests as necessary, particularly for the "supercommand without arguments" case. This implementation necessarily messes with some `click` internals, duplicating the code as it currently stands. Being a non-public API, there is of course no guarantee that this will work with future versions of `click`. However, we use this solely to implement a compatibility interface, which should vanish with `derivepassphrase` v1.0 anyway. The alternative -- rewriting `derivepassphrase` for another command-line framework, or rolling our own command-line interface from scratch -- seems even more expensive/wasteful than this solution. &lt;/pre&gt; tag:gitlist.org,2012:commit/f2624c769d0c1bcf8596488f550e78e438e849e0 2024-11-28T21:30:03+01:00 Marco Ricci software@the13thletter.info <pre>In particular, in online documentation builds, the URLs are slightly different, so the CSS rules weren't triggering correctly. &lt;/pre&gt; tag:gitlist.org,2012:commit/21da667f0ed633647c3d406f00318ba3aaaf29a1 2024-11-28T13:47:41+01:00 Marco Ricci software@the13thletter.info <pre>&lt;/pre&gt; tag:gitlist.org,2012:commit/c63aa90be7c4b8f801638c4d8c734be4d902d325 2024-11-28T13:38:43+01:00 Marco Ricci software@the13thletter.info <pre>* t/033-pre-release-copyediting: Copyedit tutorials, how-tos and reference pages structure Copyedit and reclassify changelog entry for deterministic DSA signatures &lt;/pre&gt; tag:gitlist.org,2012:commit/9518a764d9429c661603f00bc4dfb5bd63afffbf 2024-11-28T13:34:17+01:00 Marco Ricci software@the13thletter.info <pre>Includes * index pages for the tutorials and how-tos sections * adding a "man page" section in the global navigation block * dropping the rendundant "tutorial" in the tutorial titles &lt;/pre&gt; tag:gitlist.org,2012:commit/9176ab7d9a7cf9d32fcc9b98a31a5805703ca349 2024-11-28T13:08:34+01:00 Marco Ricci software@the13thletter.info <pre>Since the behavior is backward-compatible, the changes are API "additions" and "fixes", not "changes". Also fix minor spacing errors and typos. &lt;/pre&gt; tag:gitlist.org,2012:commit/c7d52b780ebed8175ef8238a2c811082b29403e6 2024-11-28T11:32:33+01:00 Marco Ricci software@the13thletter.info <pre>This includes the 'pageant-deterministic-signatures' topic branch as well. * t/ssh-key-howto: Add missing tests for `SSHAgentClient.query_extensions` runtime errors Retire the use of symlinks for the README and the changelog Use canonical test keys in all SSH agent key listings Unmark icon- and badge-type external links as external Document that `is_suitable_ssh_key` now takes an optional agent client in the changelog Add an ecdsa-sha2-nistp521 SSH test key Publish polished `is_suitable_ssh_key` and `has_deterministic_dsa_signatures` interfaces Let the `running_ssh_agent` test fixture report the agent type Decouple deterministic signatures from general SSH agent detection Fix test suite to actually test deterministic signature support Fix spurious overloaded signature mismatch Indicate external links in non-API documentation as well Split the SSH key how-to into how-to and reference documents Make suitable SSH key listing easier to distinguish Add changelog entry for deterministic DSA/ECDSA signature support Add how-to for setting up an SSH key for `derivepassphrase vault` Purge the info badges for current Python or derivepassphrase versions Support the "all signatures are deterministic" feature of some SSH agents Support one-off SSH agent client child contexts &lt;/pre&gt; tag:gitlist.org,2012:commit/b725e5f10e6031cf0781c8593fe854b2046a3bae 2024-11-27T20:23:06+01:00 Marco Ricci software@the13thletter.info <pre>Currently, this is still a white-box, maliciously compliant design that relies heavily on implementation details of the function under test. That said, the failure cases being triggered here are all SSH agent protocol violations, which would otherwise be difficult to trigger in "real" SSH agents. &lt;/pre&gt; tag:gitlist.org,2012:commit/70535b833256e31c5913e3decffc1706bb4ecd38 2024-11-27T18:21:17+01:00 Marco Ricci software@the13thletter.info <pre>Symlinks interact very poorly with online git repository viewers, and while they are offline viewable on Unixish systems, readers only see the source view, not the rendered view. For the documentation frontpage `docs/index.md`, use the Snippets extension (Material for MkDocs/Python-Markdownx) to include the contents of `README.md`. This will allow future customization of the frontpage by either appending further content or only embedding parts of the README. For the changelog, because we use `scriv` to maintain it semi-automatically and thus want the changelog snippets to stay out of the top-level directory, we do not want `docs/changelog.md` to symlink to or to snippet-include the top-level `CHANGELOG.md`. Instead, we let `CHANGELOG.md` point to the `docs` directory (in prose), and purge the symlinks to the SVG badges. The original intent was to make the changelog renderable at the top-level, without the MkDocs machinery, but this is only really feasible if the changelog is handwritten, which hasn't been the case anymore for quite some time already. &lt;/pre&gt;