IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
- This aligns the subcommand with the others which emit armored data
by default.
- If --binary is given, require the user to provide a prefix.
Previously, we used the input file name as prefix, and falling
back to a static prefix. However, this leads to surprising
results, such as writing the shards to a surprising location or
clobbering existing data.
- If neither --binary nor --prefix is given, a single stream with
multiple armor blocks is emitted. Currently, sq packet join can
not operate on this, we'll fix that shortly.
- See #51.
- First, update to Debian Trixie. Second, set CARGO_TARGET_DIR to a
fresh directory in /tmp, as the one in the source copy might be
clobbered by a symlink.
- This is the less useful and more dangerous variant of sq keyring
merge. The former does not merge two copies of the same
certificate into one, and is likely not what the user wants or
expects, and the resulting keyring is more likely to cause
problems when consumed by implementations.
- Further, we have to explain the difference between join and merge,
hurting ergonomics.
- Once, sq keyring join had the advantage of not buffering
certificates, but we now do that in order to produce the right
labels, so even that advantage was gone.
- Remove it. Developers who want to produce keyrings with multiple
copies of the same cert for testing purposes can do so with sq
packet join.
- Adapt the subplot tests. Notably, as sq keyring merge does not
preserve the order of certs as encountered in the inputs (should
it?), but orders them by fingerprint, the order in the output is
not predictable. To work around, just test listing one key.
